D E V E L O P M E N T, T R A D E , A N D T H E W T O A Handbook Bernard Hoekman, Aaditya Mattoo, and Philip English, editors THE WORLD BANK Washington, D.C. © 2002 The International Bank for Reconstruction and Development / The World Bank 1818 H Street, NW Washington, DC 20433 All rights reserved. Manufactured in the United States of America First printing June 2002 1 2 3 4 05 04 03 02 The findings, interpretations, and conclusions expressed here are those of the author(s) and do not necessar- ily reflect the views of the Board of Executive Directors of the World Bank or the governments they represent. The World Bank cannot guarantee the accuracy of the data included in this work. The boundaries, colors, denominations, and other information shown on any map in this work do not imply on the part of the World Bank any judgment of the legal status of any territory or the endorsement or acceptance of such boundaries. Rights and Permissions The material in this work is copyrighted. No part of this work may be reproduced or transmitted in any form or by any means, electronic or mechanical, including photocopying, recording, or inclusion in any informa- tion storage and retrieval system, without the prior written permission of the World Bank. The World Bank encourages dissemination of its work and will normally grant permission promptly. For permission to photocopy or reprint, please send a request with complete information to the Copyright Clearance Center, Inc., 222 Rosewood Drive, Danvers, MA 01923, USA, telephone 978-750-8400, fax 978- 750-4470, www.copyright.com. All other queries on rights and licenses, including subsidiary rights, should be addressed to the Office of the Publisher, World Bank, 1818 H Street NW, Washington, DC 20433, USA, fax 202-522-2422, e-mail pub- rights@worldbank.org. Cover and interior design by W. Drew Fasick, Serif Design Group, Washington, D.C. Cover photographs: background image, U.S. National Aeronatics and Space Administration. Inset images, clockwise from upper left: Edwin G. Huffman, the World Bank; Curt Carnemark, the World Bank; the World Bank; Edwin G. Huffman, the World Bank; PhotoDisc; Francis Dobbs, the World Bank. ISBN 0-8213-4997-X Library of Congress Cataloging-in-Publication Data has been applied for. Contents Foreword xi A Tribute to J. Michael Finger xiii Acknowledgments xix Contributors xxi Abbreviations xxiii Introduction xxvii Part I Trade Policy Reform in Context 1 1 Trade Policy Reform as Institutional Reform 3 Dani Rodrik 2 Economywide Dimensions of Trade Policy and Reform 11 Kym Anderson 3 Exchange Rate Overvaluation and Trade Protection 17 Howard J. Shatz and David G. Tarr 4 Fiscal Dimensions of Trade Liberalization 24 Liam Ebrill, Janet Stotsky, and Reint Gropp 5 Trade Policies for Poverty Alleviation 28 L. Alan Winters Part II The World Trade Organization 39 6 The WTO: Functions and Basic Principles 41 Bernard Hoekman 7 Reciprocity in the WTO 50 J. Michael Finger and L. Alan Winters 8 WTO Accession 61 Constantine Michalopoulos 9 Developing Countries and the WTO Dispute Settlement System 71 Valentina Delich 10 The Adequacy of WTO Dispute Settlement Remedies: A Developing Country Perspective 81 Robert E. Hudec Part III Selected Trade Policies Affecting Merchandise Trade 93 11 Market Access Issues and the WTO: An Overview 97 Sam Laird 12 Tariff Peaks and Preferences 105 Marcelo Olarreaga and Francis Ng iii D E V E L O P M E N T, T R A D E , A N D T H E W T O 13 Rules of Origin and Trade Preferences 114 Luis Jorge Garay S. and Rafael Cornejo 14 Nonpreferential Rules of Origin and the WTO Harmonization Program 122 Stefano Inama 15 Customs Valuation and Customs Reform 128 Vinod Rege 16 Trade Facilitation: Improving the Invisible Infrastructure 139 Brian Rankin Staples 17 Industrial Policy and Developing Countries 149 Mari Pangestu 18 Export Development Policies and Institutions 160 Philip English and Luc De Wulf 19 Trade-Related Investment Measures 171 Bijit Bora 20 Local Content Policies: Australia's Experience with Automobiles 179 Garry Pursell 21 Implementing the Agreement on Textiles and Clothing 186 Hanaa Kheir-El-Din 22 Safeguards: Making Sense of GATT/WTO Provisions Allowing for Import Restrictions 195 J. Michael Finger 23 Dealing with U.S. Trade Laws: Before, During, and After 206 Gary N. Horlick and Eleanor Shea Part IV Developing Countries and Negotiations on Trade in Services 213 24 Liberalization of Trade in Services in Developing Countries 221 James Hodge 25 The Magnitude of Flows of Global Trade in Services 235 A. Maurer and P. Chauvet 26 Quantifying Barriers to Trade in Services 247 Robert M. Stern 27 The GATS: Key Features and Sectors 259 Rudolf Adlung, Antonia Carzeniga, Bernard Hoekman, Masamichi Kono, Aaditya Mattoo, and Lee Tuthill 28 Negotiating Improved Market Access Commitments 280 Aaditya Mattoo 29 Domestic Regulations and Liberalization of Trade in Services 290 Carlo Gamberale and Aaditya Mattoo 30 Movement of Natural Persons and the GATS Major Trade Policy Impediments 304 Rupa Chanda 31 Electronic Commerce, the WTO, and Developing Countries 315 Catherine L. Mann 32 Completing the GATS Framework: Safeguards, Subsidies, and Government Procurement 326 Pierre Sauvé 33 Regional Liberalization of Trade in Services: Experiences in the Americas 336 Sherry M. Stephenson and Francisco Javier Prieto Part V Technology and Intellectual Property 347 34 International Technology Transfer and Economic Development 351 Kamal Saggi iv Contents 35 Implementing the TRIPS Agreement 359 Jayashree Watal 36 Benefiting from Intellectual Property Protection 369 Keith E. Maskus 37 Proprietary Protection of Genetic Resources and Traditional Knowledge 382 Arvind Subramanian 38 Cultural Industries and Intellectual Property Rights 390 Frank J. Penna and Coenraad J. Visser 39 Trademarks, Geographical Indications, and Developing Countries 403 Carsten Fink and Beata K. Smarzynska Part VI "Behind-the-Border" and Regulatory Issues 413 40 Multilateral Disciplines and Government Procurement 417 Simon J. Evenett 41 Standards, Regulation, and Trade: WTO Rules and Developing Country Concerns 428 John S. Wilson 42 Multilateral Disciplines and National Investment Policies 439 Bernard Hoekman and Kamal Saggi 43 Trade, Competition, and the WTO 447 Peter Holmes 44 Merger and Anticartel Policies in an Era of Integrating Markets 456 Simon J. Evenett 45 Trade and Labor: Text, Institutions, and Context 463 Simon Tay 46 Environmental Regulation and the WTO 472 Veena Jha Part VII The Trading System and Developing Countries 483 47 Developing Countries: Turning Participation into Influence 485 Diana Tussie and Miguel F. Lengyel 48 Implementation of WTO Commitments: The Development Challenge 493 J. Michael Finger and Philip Schuler 49 Special and Differential Treatment 504 T. Ademola Oyejide 50 Trade-Related Capacity Building for Enhanced African Participation in the Global Economy 509 David F. Luke Part VIII Trade Policy Questions and Guidelines 517 51 Rules of Thumb for Trade Policy 519 Constantine Michalopoulos, Maurice Schiff, and David G. Tarr 52 Arguments for and against Uniform Tariffs 526 David G. Tarr 53 Formula Approaches to Reciprocal Tariff Liberalization 535 Arvind Panagariya 54 Binding Tariffs: Why Do It? 540 Joseph F. Francois and Will Martin 55 Benefiting from Regional Integration 548 Bernard Hoekman and Maurice Schiff Appendixes 559 A Patterns of Trade and Protection: Selected Country Data 561 Francis Ng v D E V E L O P M E N T, T R A D E , A N D T H E W T O B Trade Indicators and Indices 585 Francis Ng C A Dataset on Trade and Production, 1976­99 589 Alessandro Nicita and Marcelo Olarreaga Glossary of Trade-Related Terms 593 Bibliography 605 Index 629 CD-ROMs 1 A Dataset on Trade and Production, 1976­99 Alessandro Nicita and Marcelo Olarreaga 2 Applied Trade Policy for Developing Countries: Outline,Content,and Readings for a Short Course Jaime de Melo and Marc Bacchetta Tables 3.1 Comparing "Great Depressions": Cameroon, Côte d'Ivoire, and the United States 21 4.1 Revenue Impact of Trade Liberalization 25 6.1 From GATT to WTO: Major Events 46 7.1 Control of Free-Riding in GATT Negotiations: U.S. Experience, 1947­67 53 7.2 U.S. Imports Covered by Kennedy Round Tariff Concessions (Reductions plus Bindings) as a Share of Total U.S. Imports from the Country Group 54 7.3 Tariff Concessions Received and Given at the Uruguay Round 57 8.1 Accessions to the World Trade Organization as of May 1, 2001 64 9.1 Number of Dispute Settlement Cases, 1995 through September 2000 76 11.1 Post­Uruguay Round Applied and Bound Rates of Industrial and Developing Economies by Major Product Group 98 11.2 Tariff Escalation on Products Imported by Industrial Economies from Developing Economies 99 11.3 Post­Uruguay Round Import-Weighted Applied and Bound Tariff Rates 99 11.4 Patterns of Protection in Manufacturing, 1995 102 12.1 Tariff Peaks and Imports, Quad Economies, 1999 108 12.2 Tariff Peaks and Preferential Duty Rates, Quad Economies, 1999 110 12.3 Effects of Granting Duty- and Quota-Free Access to Quad Markets to LDC Exporters 111 17.1 Evolution of Industrial Policies in East Asia, 1950s­1990s 153 17.2 Policies and Measures for Promoting Exports in Asia 155 21.1 Results of Integration Programs for ATC Stages 1 and 2 189 21.2 Number of Quotas Eliminated by Integration in ATC Stages 1 and 2 190 21.3 Restrained Trade Freed of Quotas, 1995­97 190 21.4 Expanded Market Access Attributable to Increases in Quotas (Stages 1 and 2 Combined) 190 22.1 Frequency of Use of GATT Provisions That Allow Trade Restrictions, 1948­94 196 IV.1 Summary of Selected GATS Negotiations and Domestic Policy Issues: Current Status and Desirable Outcomes 218 25.1 Inadequacies of Statistical Domains with Regard to Modes of Supply 237 25.2 Average Annual Growth of Services and Merchandise Trade, 1990­99 238 25.3 Trade in Commercial Services by Region, 1990­99 239 25.4 Share of Selected Economies in World Trade in Services, 1995 and 1998 239 25.5 Exports of Services of Selected Economies by Selected Partners, 1995 and 1998 240 25.6 Evolution of Transport, Travel, and Other Commercial Services, 1980­99 241 25.7 World Trade in Other Commercial Services by Category, 1996­98 241 vi Contents 25.8 Trade in Services by Mode of Supply, 1998 242 25.9 Shares in Inward and Outward FDI Stocks, Selected Country Groups, 1990 and 1999 243 25.10 Total U.S. Services Trade, 1994, 1996, and 1997 243 26.1 Constructed Ad Valorem Tariff Equivalent "Guesstimates" by One-Digit ISIC Services Sector, Selected Countries 250 26.2 FDI Restrictiveness Indices, Selected APEC Economies and Selected Services Sectors 250 26.3 Estimated Tariff Equivalents in Traded Services: Gravity Model­Based Regression Method 252 26.4 Average Gross Operating Margins of Firms Listed on National Stock Exchanges, 1994­96, by Economy or Region 253 26.5 Alternative Approaches to Modeling the Impact of Barriers to Trade and Investment 254 26.6 Welfare Effects of 33 Percent Reduction in Barriers to Trade in Agriculture, Manufactures, and Services 257 27.1 Main Provisions of the GATS 260 27.2 Format and Example of a Schedule of GATS Commitments 261 27.3 Number of Committed Services Sectors by Member, July 2000 263 27.4 Market Access Commitments on Insurance (Life and Nonlife) under the GATS 268 27.5 Market Access Commitments on Banking (Acceptance of Deposits and Lending of All Types) under the GATS 270 27.6 Current Status of GATS Commitments on Voice Telephone Services 273 28.1 Types of Market Access Restrictions on Commercial Presence in Services Sectors, All WTO Members, 2000 281 28.2 Developing Country Precommitments to Liberalize in Basic Telecommunications Negotiations, 1998 284 30.1 Commitments by Sector and Mode of Supply (Professional Services) 308 33.1 Specific Services Sectors Highlighted in the Subregional Integration Agreements of the Americas 342 V.1 Summary of Selected TRIPS Issues: Current Status and Possible Approaches from a Development Perspective 349 40.1 Performance Indicators for Government Procurement in Korea, 1993­98 424 41.1 Total Number of Countries That Have Established SPS Enquiry Points 435 44.1 Economies with Firms Convicted of Price Fixing by the United States or the European Commission during the 1990s 458 44.2 Cross-Border Mergers and Acquisitions in Latin America and Asia Pacific, 1991­98 459 48.1 Costs of World Bank Customs Reform Projects, Selected Countries 496 48.2 Costs of SPS-Related World Bank Projects 497 48.3 World Bank Projects Related to Intellectual Property Rights 499 54.1 Industrial Applied and Bound Tariff Rates 542 54.2 Unbound Tariff Lines by GATT Multilateral Tariff Negotiation (MTN) Category 543 A.1 Trends in Average Tariff Rates for Developing and Industrial Countries, 1980­99 562 A.2 Average Tariff Rates by Sector and FDI Inflows for All Countries in Recent Years 568 A.3 Tariff Escalation in Developing and Industrial Countries, 1994­2000 571 A.4 Pre­ and Post­Uruguay Round Nontariff Barriers for All Goods, Developing and Industrial Countries, 1984­93 573 A.5 Nontariff Barriers for All Products in Developing Countries, 1989­98 576 A.6 Tariff Peaks and Preference Margins, Canada, 1999 578 A.7 Tariff Peaks and Preference Margins, European Union, 1999 580 A.8 Tariff Peaks and Preference Margins, Japan, 1999 581 vii D E V E L O P M E N T, T R A D E , A N D T H E W T O A.9 Tariff Peaks and Preference Margins, United States, 1999 582 A.10 Comparison of MFN Applied Tariffs of Labor-Intensive Products, Selected Countries, 1997­99 583 Figures 11.1 Implications of a 40 Percent Reduction in Agricultural Trade Barriers 100 11.2 Share of Manufactures in Developing Country Merchandise Exports, Actual and Projected, 1965­2005 101 12.1 Projected Changes in LDC Export Revenues as a Result of Duty-Free Access, by Product and Importer 112 22.1 Renegotiations, Emergency Actions, and Voluntary Export Restraints (VERs), 1948­93 198 22.2 Renegotiations, Emergency Actions, Antidumping Actions, and Voluntary Export Restraints (VERs), 1948­93 200 27.1 Number of WTO Members Committing to at Least Part of an Aggregate Services Sector, July 2000 264 27.2 Market Access Commitments by WTO Members on Movement of Natural Persons (Services Mode 4) 265 29.1 Approaches toward Domestic Regulation at the Multilateral and National Levels 291 33.1 Matrix of Possible Elements in Services Trade Agreements 343 39.1 Global Demand for Trademarks and Patents 408 39.2 Nationality of Trademark and Patent Title Holders by Country Income Group, 1997 409 41.1 Number of Notifications under the SPS Agreement by Region, 1995­99 435 54.1 Implications of a Tariff Binding for the Applied Rate of Protection 545 54.2 Welfare Implications of Tariff Bindings for a Small Country 546 55.A1 A RIA Involving Small Countries 557 Boxes 4.1 Export Taxes 27 5.1 Key Questions for Determining the Poverty Impact of a Trade Reform 29 6.1 Transparency: Notification and Surveillance 45 9.1 Special and Differential Treatment and the DSU: Some Examples 73 9.2 The Advisory Centre on WTO Law 75 10.1 The Bananas Case 87 12.1 The European Union's "Everything but Arms" Initiative 106 13.1 The U.S. Trade and Development Act of 2000 115 15.1 The WTO Agreement on Preshipment Inspection 132 15.2 The International Convention on the Simplification and Harmonization of Customs Procedures: 1999 Revision of the Kyoto Convention 135 16.1 Increasing the Effectiveness of Preshipment Inspection Services 143 16.2 Improving Export Clearance in Jamaica 145 17.1 Nominal and Effective Rates of Protection 151 17.2 Instruments of Industrial and Export Promotion Policies Employed in the Republic of Korea and Japan 152 18.1 Matching Grant Schemes to Promote Exports 162 18.2 Duty Drawback for Direct Exporters: The Chilean Case 164 18.3 Subsidies, the WTO, and Developing Countries 168 21.1 The Cost of Trade Protection of Textiles and Clothing in the European Union 193 22.1 The Basics of WTO Antidumping Rules 201 22.2 The Flawed Economics of Basing Decisions on an Injury Investigation 203 24.1 Realizing the Gains from Financial Liberalization 223 viii Contents 24.2 Contribution of Inefficient Internal Transport Systems to the Concentration of China's Export Industries in Coastal Regions 225 24.3 Lessons from Reforming Argentina's Ports 227 24.4 Welfare Gains from Services Liberalization: The Case of Tunisia 229 24.5 Achieving Effective Competition in Telecommunications: The Asian Experience 233 25.1 Services Statistics and the GATS 236 25.2 Improving Data on Trade in Services 244 27.1 Financial Services: Market Access Commitments of Developing and Transition Economies 267 28.1 Services Exports by Developing Countries: Potential Gains and Current Barriers 286 29.1 Whose Regulations, and for What Purpose? Sticky Issues in Electronic Commerce 292 29.2 Assessing the WTO Telecommunications Reference Paper 293 29.3 Rules for Domestic Regulations: The Experience with Accountancy 295 29.4 Challenges in Implementing Procompetitive Regulation 297 29.5 Financial Sector Liberalization: The Need for Policy Coherence 299 29.6 Financial Liberalization and Access to Credit by the Poor 301 29.7 Pursuing Social Objectives in Competitive Telecommunications Markets 302 30.1 Immigration Patterns in OECD Countries 309 31.1 The Economics and Law of Duty-Free Electronic Commerce 319 32.1 ASEAN Proposal for a GATS Emergency Safeguard Mechanism 330 35.1 An Overview of Intellectual Property Rights 360 35.2 Combating Disease Worldwide: Fostering the Required R&D 366 36.1 Pharmaceutical Policies and the TRIPS Agreement 374 36.2 Differential Pricing 378 36.3 Health Care Systems and Essential Medicines 380 40.1 Accession to the GPA: Identifying Costs and Benefits 425 41.1 SPS Measures: The Kenya Fish Exports Case 430 41.2 Standards-Related Problems Encountered by Indian Exporters 433 41.3 Moving Forward: A Proposal 438 42.1 WTO Rules on FDI: A Positive View 441 43.1 Why Competition Policy Should Not Be Seen through a Market Access Lens 450 43.2 Free Trade and Competition Policy 452 45.1 What Are Core Labor Standards? 464 45.2 Trade and Labor Standards: Three Debates 466 46.1 Basic Principles of Environmental Regulation 474 47.1 Café au Lait Diplomacy 489 48.1 Costs of Implementing the WTO Agreements: The Jamaican Experience 502 51.1 Export Diversification in Africa: Elements of a "Business Plan" 523 55.1 Selected Major Regional Integration Agreements and Dates of Formation 549 ix blank Foreword T oday, trade policy is at the fore- markets should be more open to the goods of poor front of the development agenda, countries, and that developing countries should and it is a critical element of any strategy to fight open their markets as well as address a range of poverty. This renewed interest in trade liberaliza- institutional issues. tion does not come from dogma but instead is based The advance at Doha presents a unique opportuni- on a careful assessment of development experience ty for development, but it will require substantive over the last 50 years. participation from all countries to succeed. In partic- Developing countries that increased their integra- ular, each participating developing country will need tion into the world economy over the past two a thorough understanding of how trade liberaliza- decades achieved higher growth in incomes, longer tion can contribute to its national objectives of eco- life expectancy, and better schooling. These coun- nomic growth and poverty reduction. Such strategic tries, home to some 3 billion people, enjoyed an understanding will have to be supported by both the average 5 percent growth rate in income per capita trade negotiators and by civil society; at times, the in the 1990s compared to 2 percent in rich countries. medium-term goal of poverty reduction requires A common thread exists among these developing governments to challenge the interests of some par- countries that have been successful at generating ticular industries for short-term protection. greater growth and at lifting people out of poverty. In addition, many countries will have to break new They opened up their economies as part of a broad- ground. Today's trade issues go beyond the tradi- er development strategy that builds on two pillars: tional mechanisms of tariffs and quotas and include improving the investment climate for the private "behind-the-border" issues, such as the role of infra- sector to generate jobs and empowering poor peo- structure and governance in supporting a well-func- ple, so they can participate in growth. tioning trading economy. Many poor countries have This approach to development, with trade liberal- yet to create intellectual property regimes that make ization as one mechanism of improving the invest- traditional knowledge or cultural products into ment climate for private entrepreneurs, has gained negotiable and defensible assets; to identify options wide support among developing and industrial to upgrade and enforce national product, health, countries. All WTO member countries, including and safety standards; or to strengthen institutions those in the developing world, have reduced their for prudential and pro-competitive regulation of trade tariffs since the Uruguay Round. In its Every- services. Developing countries will have to acquire thing but Arms agreement, the European Union has quickly the needed expertise on these complex unilaterally lowered its trade barriers to the least issues, so they can negotiate more effectively and developed countries. The United States adopted the ensure that agreements serve their objective of African Growth and Opportunities Act. And in poverty reduction. November 2001, the members of the World Trade This Handbook is part of our efforts to prepare Organization launched a "Development Agenda" in developing countries to negotiate trade agreements. Doha. In doing so, they acknowledged that to make It is the product of a joint capacity-building effort progress in the fight against poverty, rich country involving a number of research institutes around xi D E V E L O P M E N T, T R A D E , A N D T H E W T O the globe. It aims to provide a summary of the eco- strong interest in, real-world trade policy making, nomics of sound trade policy and to be a guide to rather than the theory of international trade. Such many of the behind-the-border regulatory issues practitioners will be in ministries of trade, industry, that confront countries in the contexts of both and finance; parliaments; private sector associations domestic reform and international negotiations. such as chambers of commerce; consumer organi- Views and approaches to many of the issues that are zations, and policy institutes. dealt with in this volume differ substantially, and The diversity and pragmatism of the views repre- these differences are reflected here. All are motivat- sented contribute to the richness of this Handbook ed by the question of how the global trade architec- and make it a very worthwhile resource for all trade ture might be made more supportive of practitioners. It will help us "seize the moment" and development, and the question of how developing fulfill Doha's promise to focus on the need for trade countries can use international negotiations and to bring about greater growth and poverty reduction. cooperation as an instrument to further domestic reform and access to export markets. NICHOLAS STERN The Handbook is intended to be a source of SENIOR VICE PRESIDENT AND information and guidance for all practitioners, CHIEF ECONOMIST defined as those with either a responsibility for, or a THE WORLD BANK xii A Tribute to J. Michael Finger F ifteen years after the publication friends. This Handbook is dedicated to him both as of the World Bank's first handbook a practical tribute to his work and influence and in on trade policy and multilateral negotiations (Fin- the belief that his clear-sighted approach to trade ger and Olechowski 1987), the development dimen- policy will motivate researchers, analysts, and com- sions of trade policy and trade negotiations often mentators who have never had the opportunity to seem to be neglected. This is especially the case as meet him. regards trade agreements, where negotiations are Finger has noted that "trade theory is about iden- frequently driven by interest groups in high-income tifying whose hand is in whose pocket. Trade policy countries and where outcomes can have significant is about who should take it out" (Finger 1981). Both costs for developing countries, both in a monetary are important. Good policymaking requires a solid or resource-use sense and--since these obligations grounding in fact and analysis--an understanding may deflect attention and resources away from of the processes that are taking place--and a frank other, more important, tasks--in opportunity costs. recognition that, at least for international trade, The focus of much of the advice and assistance that there will be winners and losers from virtually any is offered to developing country policymakers cen- policy decision. Trade policy advice needs to pro- ters on enhancing their understanding of the rules vide this grounding, but it must also understand of the international trading game, as opposed to and internalize how the potential conflicts between determining what type of trade policy makes the winners and losers are played out in actual decision- most sense from a development perspective. The making institutions. Merely wringing one's hands latter is crucial, as only on that basis is a national and bemoaning the fact that policy advice is ignored "bottom-up" approach to the design of multilateral is not satisfactory; one needs to see why and ask rules possible. As noted by J. Michael Finger (1991a, how institutions can be designed to produce better 1991c), when it comes to the relationship between policy outcomes. the multilateral trading system and development, Five components of good trade policymaking can there is a widespread tendency to "think about be distinguished: economic analysis, information GATT only in the GATT way." Finger made this and data, political economy, operationalization of observation before the creation of the WTO and, policy advice, and a contestable market for policy characteristically, well before it became convention- research. Each is discussed below. al wisdom in the development community. Subse- quent experience has reinforced his insight.1 Economic Analysis This Handbook continues a series that Finger launched in 1987. Mike Finger, who retired from the Economic analysis is perhaps the most obvious and World Bank in 2001, has been a source of inspira- most easily provided input into the policy cocktail. tion, a guide, and a mentor to several generations of Many academic economists are active in this area, trade policy analysts. Many of the contributors to and theory is cheap. Getting concepts clear and this volume have been inspired by his writings, and showing how one thing implies another, and under a significant number have also been colleagues and what conditions, are necessary first steps toward any xiii D E V E L O P M E N T, T R A D E , A N D T H E W T O reliable policymaking. Finger's early work on the lizing support for disciplining the use of NTBs in usefulness of the concept of effective protection is the Uruguay Round. an example (Finger 1969). Such analysis might be Finger also produced the most useful summaries purely positive--describing relationships between of tariff concessions in the Uruguay Round (Finger, economic phenomena--or it might have normative Ingco, and Reincke 1996). They were useful partly dimensions in which conclusions about welfare are because he chose to measure concessions in a more inferred. Either way, it needs to be clear and logical- informative way than did national statisticians and ly watertight. the WTO. He generated the statistic, much cited in But although logical precision may be necessary, the mid-1990s, that trade reforms under IMF and it is far from sufficient for defining good policy World Bank programs liberalized more developing analysis. The other essentials--which are, regret- country trade than had the Uruguay Round. And tably, much scarcer--are that the analysis deal with recently he showed how developing countries a real problem that someone is interested in and received fewer Uruguay Round concessions (in a that it be couched in operational terms.2 Policy mercantilist sense) than they gave (Finger and research must ultimately refer to phenomena that Schuknecht 2001; Finger, Reincke, and Castro forth- can be observed--especially in terms of defining the coming). In each case, the secret was to base detailed circumstances in which the research can be applied work on the data on a well-specified question posed and the levels at which to set policy instruments-- in terms of clear operational concepts. This is, in and it must refer to policy levers that could, in prin- fact, not "mere" description but a sophisticated ciple, be manipulated. Unfortunately, all this makes marriage of theory and data that relies on simple such analysis less attractive to the editors of aca- statistics rather than high-technology econometric demic journals. Finger's work, and his publishing techniques.4 record, suggest, however, that policy analysis can be A further important dimension of good policy every bit as exciting intellectually as "pure" theory. research is to devise ways of helping policy analysts Some examples are early pathbreaking work assess- throughout the world to repeat and extend state of ing the relative impact of transport costs and tariffs the art analysis. Many theoretical analyses are easily (Finger and Yeats 1976) and assessments of the replicated, but once one begins to work with data, effects of the offshore assembly provisions in U.S. there can be high costs to replication even if there customs law (Finger 1976b). are no unforeseen difficulties to overcome. The first requirement for facilitating replication is to make data available. Great strides have been taken recent- Information and Data ly in this direction, but there are still many hurdles Great analysis will occasionally throw up universal to be overcome. For example, it is unfortunate that truths that do not require detailed empirical analy- members of the WTO continue to resist making the sis in order to be applied. More commonly, howev- Uruguay Round bound tariff rates available in a er, policy analysis is empty without a supporting convenient form for analysis.5 information base. The most obvious need in inter- The second component also focuses on conven- national trade is for data on international trade ience. Where complex operations or calculations are flows and policies, but information on the laws and required, it is essential to make tools and routines institutions that govern behavior is also essential. available to external researchers and so make it "Merely" collecting and presenting information can cheap and easy for them to carry out their own be spectacularly illuminating. For example, in the analysis. Although we tend to equate data with mid-1980s Finger inspired and managed the World numbers, it is equally important for policy analysts Bank's empirical work on nontariff barriers to have accurate representations of institutional and (NTBs); Nogués, Olechowski, and Winters (1986) is legal processes. Some of these may be summarized an example.3 At that time theory about the effects of numerically, as in, for example, Finger's compilation NTBs was not lacking, but there was almost no con- of tariff concessions granted and received in the cept of how pervasive these barriers were. Present- Uruguay Round (Finger, Ingco, and Reincke 1996). ing the data (and being clear about the adequacy or In other cases it is a matter of observing processes inadequacy of the measurements) took the debate and finding ways of distilling their essence into a much further and had a significant effect in mobi- few simple statements. The pioneering work by Fin- xiv A Tribute to J. Michael Finger ger, Hall, and Nelson (1982) on administrative pro- into an unbalanced game with a strongly biased tection (on which more below) is based on a careful outcome. The classic analysis by Finger, Hall, and institutional study of U.S. antidumping, counter- Nelson of antidumping actions makes clear how the vailing, and escape clause protection. technical process favors business interests and dis- Today, as we grapple with the subtleties and com- enfranchises consumers and users--in fact, the plexities of trade and development policies and the authors argue that this is precisely the political role of the WTO in supporting development, work function of the process. Only by covertly loading on information and data is prominent. For exam- the scales can a low-level (ostensibly nonpolitical) ple, many of the chapters on services in this Hand- process hope to resolve distributional struggles; if book deal with how to measure the progress of the loading is obvious, the matter tends to become liberalization in these areas and examine how gov- political and more difficult. Moreover, once estab- ernments have applied the agreements domestically. lished, such unbalanced processes tend to perpetu- ate themselves because the interest groups whose interests are served (including the people who man- Recognition of Political Economy age the processes) ensure that this happens. Finger, If entrepreneurs and workers respond to price sig- who has contributed massively and seminally to our nals and incentives in determining their behavior, it understanding of antidumping and safeguards over is hardly surprising that they respond to other eco- two decades, discusses the subject in Chapter 22 of nomic incentives, such as the opportunity to create this volume. economic rents by intervening in markets. Nor is it A second area in which political economy has surprising that politicians and bureaucrats also been and continues to be crucial is in the analysis of respond to incentives of various kinds. Political the GATT and the WTO. Finger (1974, 1976a) economy may be broadly thought of as exploring demonstrated the centrality of reciprocity to GATT the role of political activities and forces in shaping tariff negotiations: despite the requirement that all economic behavior. Policy analysis must take it into tariff cuts be extended to all GATT partners, nego- account not only in predicting the outcomes of par- tiators chose commodities in such a way that much ticular policy changes but also, and more particular- of the benefit of a cut went to the country request- ly, in designing institutions and policy regimes. ing it (the principal supplier), and nearly all of the Political economy and trade policy are closely benefit went to countries taking active part in the linked because every trade intervention potentially negotiations. In passing, Finger also showed that the creates conflict between winners and losers--pro- developing countries which made significant con- ducers and consumers, agriculture and industry, cessions in the multilateral negotiations received far skilled and unskilled labor. Because all political sys- more concessions on their exports than did passive tems find it difficult to deal with redistribution observers. explicitly, trade policy presents both problems and Finger (1979) was being relatively unfashionable opportunities. Supporting a trade liberalization in observing that the critical political balance of the while providing explicit redistribution to prevent GATT was internal to countries, as exporters seek- anyone from losing is a big political challenge, ing market access abroad pressed import-compet- whereas using trade intervention as a form of covert ing sectors to concede market access at home. When redistribution that is buried beneath a pile of arcane U.S. legislation provided an alternative route for technical detail often looks like an attractive way of exporters to open foreign markets, in the form of protecting or rewarding specific interests. Section 301, which authorized unilateral trade sanc- Opacity can make trade policy very attractive. tions against trade-restricting partners, the whole Finger (1981) and Finger, Hall, and Nelson (1982) dynamic of U.S. support for the multilateral system showed that the mechanisms involved can be changed (Finger 1991b). Finger was prominent extremely subtle. Making protection subject to legal among those who had argued that for developing requirements and establishing expert investigative countries the principal requirement was not to authorities to establish the "facts" of the case all open up others' markets, but to open up their own serve to remove the process from public scrutiny. By (Finger and Kreinin 1976) and that in this respect making procedures complex and expensive, one can the GATT approach (and GATT-think) based solely turn an apparently objective pseudo-legal process on reciprocity was not very useful. Developing xv D E V E L O P M E N T, T R A D E , A N D T H E W T O countries must address their own needs directly via forces and shining a light on them (Jagdish Bhag- their own policy and should not view the multilat- wati's "Dracula principle"); redressing the balance eral system as providing a shortcut to good trade of forces in trade debates to promote consumer policy or good trade outcomes; it just does not do interests; and making transparent the winners and that (Finger and Winters 1998). losers from any action (or inaction); see, for exam- These political-economy problems figure promi- ple, Finger (1982, 1986). It also illustrates the dan- nently in this Handbook, for it is only by recogniz- gers of complexity, suggesting a second aspect of ing their force that they can be overcome. Thus, for operationalization: the use of rules of thumb in pol- example, Chapter 52, by Tarr, stresses the political- icymaking. Among the rules of thumb advocated in economy advantages of uniform tariffs, which are this Handbook by some authors are the use of uni- much more robust to lobbying than tailor-made form tariffs as a robust antidote to sectoral special tariffs. Rodrik, in Chapter 1, emphasizes the impor- pleading and rent-seeking, and promotion of effec- tance of developing local institutions as local solu- tive competition as the single most important tions to local problems rather than adopting objective in services markets. uniform institutions imposed by the international community, and this analysis is consistent with Fin- A Contestable Market for Policy Research ger's point about the costs of certain Uruguay Round institutions (Finger and Schuler 2000). In An important dimension of such beneficial compe- Chapter 22, Finger explores the political economy tition relates to policy analysis. The social function of safeguard provisions, and in Chapter 7 Finger of such research is to improve policy outcomes by and Winters explore what reciprocity means in the basing them on the best possible understanding of current broad agenda of the WTO. (It has no "exter- the effects of policy. De facto, its political function is nal" definition; it is whatever deal the parties are to smooth the path of decisionmaking by ensuring willing to agree on.) that relatively minor issues do not destroy social consensus and impose huge costs in the form of strife. This second function is not unimportant (as Operationalization of Policy Advice Rodrik notes in Chapter 1), but it is often at vari- Policy analysis is ultimately sterile if it does not ance with the first. The tension between the two change behavior. How to present and package the roles of policy research is felt most immediately in findings of analysis in ways that both strike chords official policy research centers. If analysts there stick with decisionmakers and are (relatively) easy to to the objective side of their brief, they are ignored, apply is critical. Again, Mike Finger leads the way. abused for being irrelevant or obstructionist, and The clarity and directness of his writing is a model often, as happened to Finger's unit in the U.S. Trea- for all researchers. And it is substantially achievable sury, closed down. If they stress the political aspects, by them too, for while it certainly requires talent, it they discredit themselves and, ultimately, their insti- mostly relies on thinking hard--and with brutal tutions as purveyors of information; indeed, they objectivity--and on working hard (spending time may even discredit analysis itself. And by giving a refining one's prose). Finger also has a talent for the politically convenient compromise a gloss of spuri- memorable phrase or metaphor: "Antidumping is ous intellectual respectability, analysts can sow the ordinary protection with a great public relations seeds of further problems by establishing the wrong program" (Finger 1993); "Where the WTO got it basis for thinking about future decisions. The falla- wrong, it was perhaps because the World Bank did cy that trade liberalization creates jobs (perpetrated, not get it at all" (Finger and Nogués 2002, on the for example, in the debate on the North American inappropriateness of certain Uruguay Round out- Free Trade Agreement), and its refutation by experi- comes for development);"Half of domestic interests ence, have made rational trade policy more difficult have no chance to score" (on antidumping, in to achieve. The fallacy that reductions of tariffs on a Chapter 22 below, with a picture of a soccer field of developing country's exports are more important which only one end has a goal). than reductions of tariffs on its imports has led to The discussion of political economy in the pre- the waste of huge resources on instruments such as ceding section covered some aspects of operational- trade preferences, the Generalized System of Prefer- izing trade policy advice: recognizing reactionary ences (GSP), and the New International Economic xvi A Tribute to J. Michael Finger Order and, ultimately, to the false notion that the Notes GATT/WTO process and good trade policy are 1 Similar arguments are made in Finger and Kreinin (1976) and coterminous (see Finger 1975, 2001; Finger and Finger (1982). Kreinin 1976). 2 This is not to decry basic economic science but merely to place How can this tension be resolved? In his valedic- it outside the box of policy research. tory speech to the U.S. Treasury (Finger 1981), Mike Finger observed that "political responsibility is the 3 The actual collection was mostly (and continues to be) done ultimate intellectual vasectomy." What is the by the UNCTAD, but the presentation and use of the data for policy analysis was pursued more vigorously by the World answer? It is to ensure that the market for policy Bank. research is open and contestable. Governments and international organizations require research arms, 4 Another example was Finger and DeRosa (1980), which showed with the simplest of tools that the IMF's Commodity but it is vital that others, outside government, are Compensatory Fund might not have the desired effect of stabi- also able to participate fully in this market. Govern- lizing developing countries. ments thus have to make data and information eas- 5 A CD-ROM that replicates the country schedules is available, ily available publicly, accept criticism, and be but it is not an electronic file of data. prepared to justify distributive judgments and deci- sions. Finger has shown by example what type of analy- sis is necessary for better policy choices and out- comes. We hope that this Handbook, and the kind of collaborative, research capacity­building effort on which it draws, will help stimulate others to emulate the "Finger approach" to policy research and analysis. BERNARD HOEKMAN L. ALAN WINTERS xvii blank Acknowledgments T he chapters included in this Hand- (UNCTAD); Diana Tussie, LATN; Ademola Oyejide, book are a product of a collaborative University of Ibadan and AERC; Rajesh Chadha, research and capacity-building project that involved National Council of Applied Economic Research scholars in developing countries, international (NCAER), New Delhi; Hana' Al-Sagban, ERF; Thier- experts, and World Bank trade economists and that ry Noyelle, United Nations Department of Economic was designed to assess the costs and benefits of fur- and Social Affairs (UNDESA); Dominique Njinkeu, ther multilateral rule-making and liberalization. In AERC; Mari Pangestu, Center for Strategic and addition to researchers based in national research International Studies, Jakarta; Alan Winters, Sussex organizations, the project draws on the work of a University; and Stephen Yeo, Centre for Economic number of research networks, including the Latin Policy Research (CEPR), London. We are grateful as American Trade Network (LATN); the Economic well to Constantijn Claessens, Antonio Estache, Research Forum for the Arab Countries, Iran, and Carsten Fink, Faezeh Foroutan, John Hegarty, Turkey (ERF); the African Economic Research Con- Charles Kenny, Dorsati Madani, Amrita Narlikar, sortium (AERC); the Coordinated African Program Christina Neagu, Claudia Orozco, Randeep Rathin- of Assistance on Services (CAPAS); and the Trade dran, and Gomi Senadhira, who contributed part or Policy Forum of the Pacific Economic Cooperation all of the material for boxes on specific topics. Council (PECC). The views expressed are entirely Many of the papers that were prepared under the those of the contributors, and do not necessarily auspices of the research capacity­building project reflect the views of the World Bank Group, the insti- of which this Handbook is a part have been pub- tutions that the authors are affiliated with, or the lished in journals and conference volumes. Papers countries they represent. written by members of the project and dealing The project was supported by a grant from the with many of the subjects covered in this Hand- U.K. Department for International Development book can be found in the April 2000 issue of The (DFID) and by the World Bank Institute, the World World Economy; the May 2001 issue of The Review Bank Research Support Budget, the government of of International Economics; Olarreaga and Rocha the Netherlands, and the Société Générale de Sur- (2000); Hoekman and Martin (2001); Stern veillance (SGS), Switzerland. We are particularly (2001); Martin and Pangestu (forthcoming); grateful to Susan Prowse and Charlotte Seymour- Maskus and Wilson (2001); and Mattoo and Stern Smith for their encouragement and guidance. (forthcoming). The editors are indebted to Maria Kasilag, Rebecca All of the members of the World Bank's trade Martin, Ana Rivas, Rob Simms, and Lili Tabada for team have been active contributors to the project, in excellent administrative support throughout the particular Will Martin, who has played a key role in project. We thank the following people for their managing the project and provided leadership in active engagement and participation in the delivery working with partners in Asia. Carsten Fink, and of various parts of the project: Kym Anderson, Ade- Keith Maskus have been unstinting in their willing- laide University; Bijit Bora, at the time with the Unit- ness to contribute and comment on materials. Fran- ed Nations Conference on Trade and Development cis Ng played an essential role in providing data to xix D E V E L O P M E N T, T R A D E , A N D T H E W T O participants. The Trade and Production Database that helped improve the final product. We thank CD-ROM included with the Handbook is the result those who acted as readers and discussants of chap- of the painstaking work of Marcelo Olarreaga and ters and the handbook as a whole, including Claude Alessandro Nicita. Miroslava Zervoudakis and Barfield, Paul Collier, Richard Eglin, Alan Gelb, Joe Faezeh Foroutan were key contributors to the pro- Francois, Jacob Kol, Patrick Messerlin, Douglas Nel- ject's Website, , which son, David Palmeter, Garry Pursell, Jayanta Roy, has become the major dissemination vehicle for the Alan Winters, Luc de Wulf, Jamel Zarrouk, and output generated by team members. Ambassador B. K. Zutshi. The chapters included in the Handbook benefited Finally, we acknowledge the considerable time from comments and feedback obtained from par- and effort that has been devoted to the preparation ticipants in workshops, conferences, and seminars, of the second CD-ROM included with this Hand- who are too numerous to be mentioned by name book, "Applied Trade Policy for Developing Coun- here. We are very grateful to the people and institu- tries: Outline, Content, and Readings for a Short tions that hosted and helped arrange these meet- Course." Our thanks go to the principal authors, ings, in particular Richard Eglin, Sam Laird, and Jaime de Melo, Centre d'Etudes et de Recherches sur Peter Tulloch of the WTO secretariat, who organ- le Développement International (CERDI), and ized a major conference in September 1999 and a Marc Bacchetta, WTO, and to their organizations, as day-long review seminar of the first draft of the well as to Olivier Jammes (CERDI) for CD-ROM Handbook in May 2001. Members of the WTO del- development. egations in Geneva provided invaluable suggestions xx Contributors Rudolf Adlung World Trade Organization Kym Anderson Adelaide University, Australia, and CEPR Marc Bacchetta World Trade Organization Bijit Bora World Trade Organization Antonia Carzeniga World Trade Organization Rupa Chanda Indian Institute of Management Philippe Chauvet World Trade Organization Rafael Cornejo Inter-American Development Bank Valentina Delich FLACSO/Argentina and LATN Jaime de Melo University of Geneva, CERDI, and CEPR Luc De Wulf Independent Consultant Liam Ebrill International Monetary Fund Philip English World Bank Simon J. Evenett World Trade Institute, Berne, and CEPR J. Michael Finger American Enterprise Institute, Washington,D.C. Carsten Fink World Bank Joseph F. Francois Erasmus University, Rotterdam, and CEPR Carlo Gamberale World Trade Organization Luis Jorge Garay S. Inter-American Development Bank Reint Gropp International Monetary Fund James Hodge University of Cape Town Bernard Hoekman World Bank and CEPR Peter Holmes University of Sussex, U.K. Gary N. Horlick O'Melveny & Myers, Washington, D.C. Robert E. Hudec Tufts University, Boston. Stefano Inama UNCTAD Veena Jha UNCTAD Hanaa Kheir-El-Din Cairo University Masamichi Kono Financial Services Authority (Japan) Sam Laird UNCTAD and University of Nottingham, U.K. Miguel F. Lengyel FLACSO/Argentina and LATN David F. Luke Organization for African Unity Catherine L. Mann Institute for International Economics, Washington, D.C. Will Martin World Bank Keith E. Maskus University of Colorado Aaditya Mattoo World Bank Andreas Maurer World Trade Organization xxi D E V E L O P M E N T, T R A D E , A N D T H E W T O Constantine Michalopoulos Independent consultant Francis Ng World Bank Alessandro Nicita University of Geneva Marcelo Olarreaga World Bank and CEPR T. Ademola Oyejide University of Ibadan, AERC, and the Development Policy Center, Ibadan, Nigeria Arvind Panagariya University of Maryland Mari Pangestu Centre for Strategic and International Studies, Jakarta Frank J. Penna The Policy Sciences Center, Inc., New Haven, Conn. Francisco Javier Prieto Organization of American States Garry Pursell Independent consultant Vinod Rege International trade consultant Dani Rodrik Harvard University Kamal Saggi Southern Methodist University, Dallas, Tex. Pierre Sauvé OECD Maurice Schiff World Bank Philip Schuler University of Maryland Howard J. Shatz Harvard University Eleanor Shea O'Melveny & Myers, Washington, D.C. Beata K. Smarzynska World Bank and CEPR Brian Rankin Staples Trade Facilitation Services, Ottawa Sherry M. Stephenson Organization of American States Robert M. Stern University of Michigan Janet Stotsky International Monetary Fund Arvind Subramanian International Monetary Fund David G. Tarr World Bank Simon Tay Member of Parliament, Singapore Diana Tussie FLACSO/Argentina and LATN Lee Tuthill World Trade Organization Coenraad J. Visser University of South Africa Jayashree Watal World Trade Organization John S. Wilson World Bank L. Alan Winters University of Sussex, U.K. and CEPR xxii Abbreviations ACIS Advance Cargo Information System ACP African, Caribbean, and Pacific (Cotonou Convention, formerly Lomé) ACWL Advisory Centre on WTO Law AD antidumping AERC African Economic Research Consortium AGOA African Growth and Opportunity Act (U.S.) AMS aggregate measure of support APEC Asia-Pacific Economic Cooperation ASEAN Association of Southeast Asian Nations ASYCUDA Automated System for Customs Data and Management (UNCTAD) ATC Agreement on Textiles and Clothing (WTO) ATPA Andean Trade Preferences Act BDV Brussels Definition of Value BIT Bilateral Investment Treaty BTN Brussels Tariff Nomenclature CAP Common Agricultural Policy (EU) CBD Convention on Biological Diversity CCC Customs Cooperation Council (now the WCO) CCCN Customs Cooperation Council Nomenclature CEFACT Center for Facilitation of Procedures and Practices for Administration, Commerce, and Transportation (UN) CGE computable general equilibrium (model) c.i.f. cost, insurance, and freight CRM customs reform and modernization CRTA Committee on Regional Trade Agreements (WTO) CTE Committee on Trade and Environment (WTO) CTH change in tariff heading CVD countervailing duty DSB Dispute Settlement Body (WTO) DSP dispute settlement procedures (WTO) DSU Dispute Settlement Understanding (WTO) EBA Everything but Arms (EU initiative for LDCs) EC European Community EDI electronic data interchange EDIFACT Electronic Data Interchange for Administration, Commerce, and Transport (UN) EEC European Economic Community EFTA European Free Trade Association xxiii D E V E L O P M E N T, T R A D E , A N D T H E W T O EPZ export-processing zone ERP effective rate of protection EU European Union FAO Food and Agriculture Organization of the United Nations FDI foreign direct investment f.o.b. free on board FSC foreign sales corporation FTA free trade area FTAA Free Trade Area of the Americas GATS General Agreement on Trade in Services (WTO) GATT General Agreement on Tariffs and Trade (WTO) GDP gross domestic product GMO genetically modified organism GNP gross national product GPA Agreement on Government Procurement (WTO) GSP Generalized System of Preferences GTAP Global Trade Analysis Project HCC Heads of Customs Conference (NAFTA) HS Harmonized Commodity Description and Coding System ICC International Chamber of Commerce ICTSD International Centre for Trade and Sustainable Development IDB Integrated Data Base (WTO) IECC International Express Carriers Conference IF Integrated Framework for Technical-Related Assistance, Including Human and Institutional Capacity Building to Support Least-Developed Countries in Their Trade and Trade-Related Activities IFIA International Federation of Inspection Agencies ILO International Labour Office IMF International Monetary Fund IPRs intellectual property rights ISIC International Standard Industrial Classification ISO International Organization for Standardization ITC International Trade Centre (UNCTAD and WTO) ITC International Trade Commission (U.S.) ITCB International Textiles and Clothing Bureau ITO International Trade Organization LATN Latin American Trade Network LDC least-developed country (UN classification) MAI Multilateral Agreement on Investment MEA Multilateral Environmental Agreement MENA Middle East and North Africa MERCOSUR Common Market of the South MFA Multifibre Arrangement MFN most-favored-nation MRA mutual recognition agreement MTA multilateral trade agreement MTN multilateral trade negotiation NAFTA North American Free Trade Agreement NATO North Atlantic Treaty Organization NGO nongovernmental organization NRP nominal rate of protection xxiv Abbreviations NTB nontariff barrier NTM nontariff measure OAU Organization of African Unity OECD Organisation for Economic Co-operation and Development OMA orderly marketing arrangement PPM production and processing method PSI preshipment inspection PTA preferential trading agreements Quad Canada, European Union, Japan, and the United States QR quantitative restriction R&D research and development RCA revealed comparative advantage RIA regional integration agreement ROO rules of origin SADC Southern African Development Community SCM subsidies and countervailing measures S&D special and differential (treatment) SDR Special Drawing Right SGS Société Générale de Surveillance SITC Standard International Trade Classification SPS sanitary and phytosanitary STE state trading enterprise TABD Transatlantic Business Dialogue TBT technical barriers to trade TMB Textiles Monitoring Body (WTO) TPO trade promotion organization TPRB Trade Policies Review Body (WTO) TPRM Trade Policies Review Mechanism (WTO) TRIM trade-related investment measure TRIPS Trade-Related Aspects of Intellectual Property Rights (WTO agreement) TRQ tariff rate quota UNCITRAL United Nations Committee on International Trade Law UNCTAD United Nations Conference on Trade and Development UNDP United Nations Development Programme USTR U.S. Trade Representative VER voluntary export restraint WCO World Customs Organization WHO World Health Organization WIPO World Intellectual Property Organization WITS World Integrated Trade Solution (World Bank) WTO World Trade Organization xxv blank Introduction M any countries have been less than institutions, trade policies, narrowly defined, are successful in integrating into the still important in today's international economic world economy and benefiting from trade reform landscape. Barriers to exports of some products in programs. The reasons are multifaceted and com- which developing countries have a comparative prise a mix of domestic and international factors. advantage remain high--tariffs on some agricultur- Barriers to trade and investment remain high in al products are over 100 percent. Agricultural subsi- many nations, with policy regimes implying signifi- dies in OECD countries exceeded US$300 billion in cant anti-export bias. Numerous countries have 2000, contributing to global price instability and been affected by civil strife and war. And in spite of impeding the ability of developing countries to the trade preferences granted by member countries compete on export markets. of the OECD, industrial country tariff structures are Trade between developing countries began to still characterized by escalating tariffs, with high grow rapidly in the 1990s, increasing the signifi- tariff peaks for agricultural products and for labor- cance of their own trade barriers for export interests intensive products such as clothing. in these countries. Antidumping actions are no There is general agreement that many complementa- longer limited to OECD economies but have come ry policies and institutions are needed to support trade to be used intensively by a number of developing policy reforms in order to create an enabling environ- countries. Barriers to trade in services are many ment for supply-side responses that generate employ- times those that apply to trade in merchandise, ment and economic growth. As Dani Rodrik argues in especially where movement of the service provider Chapter 1 of this Handbook,if trade policy reform is to is necessary. In many cases these barriers and detri- be successful, it must be embedded in and supported mental policies can be removed only through inter- by an effective institutional setting,and it must be com- national negotiations. plemented by other reforms. A large and complex International trade agreements, in particular the "behind-the-border" agenda has to be addressed if WTO agreement, have become the focal point for trade reform is to have its intended effect. Much many discussions on trade and investment policy. depends on complementary policies that define the As a result, policymakers and citizens of developing business environment--on policies regarding invest- countries are confronted with demands that a num- ment in human capital (education),infrastructure,and ber of trade policy­related issues be addressed in the quality of public and private sector governance. the context of multilateral or regional negotiations. The Handbook focuses on a number of the elements of This offers opportunities to pursue what are regard- that agenda, as well as on more"traditional"trade poli- ed as desirable domestic reforms, but it also poses cy issues such as the design of the tariff regime. risks associated with agreements or rules that may not be supportive of development prospects. The traditional mechanism driving trade agree- Trends in the Multilateral Trading System ments has been the reciprocal exchange of commit- Although the challenges confronting developing ments to reduce trade barriers. This mechanism countries primarily concern domestic policies and results in greater welfare improvements than can be xxvii D E V E L O P M E N T, T R A D E , A N D T H E W T O obtained through unilateral reform, as it generates WTO auspices to address developing country con- liberalization both at home and abroad and makes cerns. Similar calls were put forward in the prepara- politically feasible domestic trade reforms that oth- tions for the 2001 ministerial meeting in Doha. erwise might be blocked by powerful vested inter- The Doha Development Agenda that emerged ests. International cooperation can also be a useful from the meeting clearly reflects the increased device for pursuing domestic reforms that are indi- prominence of development concerns in WTO rectly linked to trade. As tariff barriers have fallen deliberations--in turn, the result of increased par- and quantitative restrictions have disappeared, the ticipation by developing countries in the trading focus of trade agreements has increasingly shifted system. All that was done, however, was to define an toward regulatory regimes that can have an impact agenda. Achievement of a prodevelopment outcome on trade and investment. remains a major challenge. Resistance to liberaliza- Multilateral negotiations on nonborder policies, tion of "hard-core" sectors such as agriculture and administrative procedures, and domestic legal textiles that are of key interest to developing coun- regimes have proved much more complex than talks tries is very strong; conversely, many low-income on traditional market access. Because it is more dif- countries are unwilling to extend the reach of the ficult to trade"concessions,"the focus tends to be on WTO to cover new issues. Implicitly, if not explicit- the identification of specific rules that should be ly, much of the discussion and debate at Doha con- adopted. Given the disparities in economic power cerned defining the limits of the WTO. Developing and resources among countries, the outcome often countries played a central role in this debate, with reflects the status quo in high-income countries. many resisting the further expansion of the WTO These may be fully consistent with the development into the territory of domestic regulation. priorities of low-income countries, but there is no The Doha Ministerial Declaration launches nego- presumption that this will be the case. tiations on market access for manufactures, dispute Developing country misgivings regarding the settlement, WTO rules, disciplines on regional inte- rule-making dimensions of the WTO became gration, environment, and intellectual property increasingly prominent in the 1990s. These con- rights (geographical indications). These talks will cerns centered on the costs required for implement- complement ongoing negotiations on agriculture ing some WTO agreements, the lack of adequate and services, as mandated by the Uruguay Round financial assistance, and the failure of high-income agreements. Negotiations are to be concluded by countries to grant "special and differential" treat- 2005. At the next WTO ministerial meeting, in 2003, ment to developing countries. (Most of the provi- negotiations will be launched on four "Singapore sions in the WTO agreements calling for such issues"--competition, investment, trade facilita- treatment are "best endeavor" commitments that tion, and transparency in government procure- are not binding on high-income countries.) A more ment--if agreement on modalities can be obtained fundamental concern was that the rules of the game by explicit consensus at that time. were not always compatible with national efforts to Whether the end result will be prodevelopment reduce poverty and increase economic growth. will depend to an important degree on the extent to For the rules to make sense for all members, which developing and industrial country trade bar- stakeholders in developing countries must partici- riers are lowered, and on the rules that emerge. A pate in the domestic policy formation process, be key determinant of the outcome of the negotiations able to inform national representatives of their will be effective and proactive developing country views, and hold their representatives accountable participation. This, in turn, requires a good under- for outcomes. If WTO agreements were unambigu- standing of where national interests lie and a good ously seen by constituencies in developing countries understanding of the substantive issues, not just by as being conducive to (or consistent with) the government officials but also by the private sector attainment of development objectives, these agree- and civil society. There is clearly a need to strength- ments could play a much more beneficial and effec- en capacity to undertake analysis and to identify tive role. In the run-up to the 1999 WTO ministerial national reform priorities, market access con- meeting in Seattle, a number of prominent straints, and the potential merits and implications observers and policymakers called for the launch of of multilateral disciplines. This Handbook is a "Development Round" of negotiations under intended as a contribution to that effort--as a use- xxviii Introduction ful resource for analysts and stakeholders engaged views do not necessarily reflect those of the institu- in the design of trade-related policies. tions with which they are affiliated. Not everyone will necessarily agree with all the policy recommendations made by the authors. After Objectives of the Handbook all, as we noted above, on a number of issues there is A major challenge confronting developing countries no "one size fits all" answer, and this is especially is to use international negotiations and cooperation true of regulatory policies. What matters most is to as instruments for improving their terms of trade ask the right questions and to determine the status and their access to export markets and as mecha- quo in a given area. It is important to obtain as nisms for adopting and implementing domestic pol- much information as possible regarding alternative icy reforms that will raise living standards and policy options, to understand what type of analysis reduce poverty. The design of trade policy reform is is needed to provide policy guidance, and to have a a complex matter that extends far beyond tariffs and good understanding of the prevailing multilateral quotas applied at the border. It must be comple- rules of the game. mented by policies designed to ensure that enter- Although much of what is contained in the Hand- prises can compete on world markets. There is no book is motivated by the fact that the issues are on "one size fits all" package of policy reform, and no the agenda of international negotiations, the magic bullet. Approaches will and must differ across emphasis of many contributors is on economic and countries, reflecting different circumstances, endow- development dimensions. The institutions and poli- ments, legal systems, and cultures. cies that are important for development and eco- One goal of this Handbook is to provide informa- nomic growth extend far beyond the subject areas tion on the implications of--and options offered that the WTO deals with or can deal with. Although by--international trade agreements, especially the the WTO can be useful in helping countries address WTO, for developing countries that seek to use specific bottlenecks and constraints that impede trade as a vehicle for development. Contributors trade, most of the trade policy agenda is domestic. It were asked to write relatively short chapters on a is therefore vital that policymakers and civil society variety of trade policy­related topics that are have a good understanding of what their national important from a development perspective and that priorities are and what makes for good policy, are subject to or affected by multilateral rules, or informed by the experiences of other countries, in may become so. The chapters assess the economics order to determine what types of multilateral coop- of the issues, survey what cross-country experience eration can help countries benefit from trade inte- suggests are good practices, and consider the pros gration. and cons of the possibilities for using international Relatively little emphasis is given in the Hand- cooperation as an instrument for improving both book to an enumeration of WTO disciplines. There domestic policy and access to export markets. are many readily available resources that can pro- Although there is an emphasis on the WTO, many vide the interested reader with such information, of the issues addressed also arise in the context of starting with the WTO Website, . regional integration agreements. The CD-ROM provided with this Handbook, Notwithstanding its length, this Handbook can "Applied Trade Policy for Developing Countries," only partially address the many policy issues that contains all of the major agreements and many arise in the course of efforts to integrate into the other WTO documents. Therefore, only key aspects world economy. The focus is on trade policy, broad- of WTO rules are discussed. Relatively more atten- ly defined to cover both traditional instruments of tion is given to the General Agreement on Trade in commercial policy--tariffs, customs administra- Services (GATS) and the Trade-Related Aspects of tion, and so on--and "new" issues such as services, Intellectual Property Rights (TRIPS) agreement intellectual property, and the behind-the-border than to General Agreement on Tariffs and Trade regulatory agenda that has implications for market (GATT) disciplines, since a wealth of analysis and access conditions. The approach is one of multiple information exists on "traditional" trade policy voices; the contributors include many authors who instruments. The chapters in the Handbook dealing have no connection to the World Bank. In all cases, with merchandise trade issues focus primarily on contributors wrote in a personal capacity, and their those subjects that are of greatest interest to devel- xxix D E V E L O P M E N T, T R A D E , A N D T H E W T O oping countries--tariff peaks, preferences, rules of Each of the eight parts begins with a short intro- origin, customs clearance and trade facilitation, duction that is intended as a reader's guide to the local content and other industrial policy measures, issues and to further reading. Annotated references, and export promotion. drawn in part from Hoekman and Kostecki (2001), One topic that is of major importance to develop- are listed in each introduction for those who are ing countries--agriculture and agricultural trade interested in pursuing in-depth discussion and policies--is not addressed in any depth in this analysis. To facilitate consultation of the citations in Handbook because it is the subject of a companion individual chapters, we have compiled an integrated volume (Ingco and Nash forthcoming). The same is bibliography, found at the end of the Handbook. true for another major issue area: trade and poverty. The boxes included in the chapters illustrate spe- Winters, in Chapter 5 of this volume, summarizes cific points or describe specific cases. Many of these the main messages and conclusions that have boxes were prepared by the editors of the Hand- emerged from the literature on trade and poverty. book, drawing on papers prepared for this project Those seeking a more in-depth treatment should and on the literature. The chapter authors are not consult McCulloch, Winters, and Cirera (2001), responsible for the content of those boxes written which reviews the literature and good practices con- by the volume editors or by other contributors. cerning the design of trade policy reforms from a The appendix includes a glossary and a set of poverty alleviation perspective. tables that present data on trade barriers for a large sample of countries. The glossary provides a listing of major WTO articles and provisions for the con- Structure of the Handbook venience of readers who are not familiar with the This volume has eight parts. The chapters in Part I WTO, as well as succinct descriptions and defini- place trade policy reform in a development context tions of key trade-related institutions and policies. and discuss key dimensions of reform. Part II deals A CD-ROM containing data on tariffs, trade, and with the main aspects of the WTO. Parts III, IV, and production is packaged with the Handbook. The V cover the areas that are the subject of WTO rules: database is described briefly in the appendixes and trade in goods, trade in services, and the protection is more fully documented on the CD-ROM. We of intellectual property. Part VI contains discussions have also included in the appendixes and on the of a number of regulatory issues; many of these CD-ROM a short guide to the most commonly used have not been subject to multilateral rules but are indicators and indices that can be applied to the now being introduced into the WTO agenda due to data. More detailed datasets and analytical tools strong interest on the part of some high-income that can be used for negotiations are being devel- countries and nongovernmental organizations. Spe- oped in cooperation with UNCTAD; this set of tools cific process-related concerns of developing coun- will be released in mid-2002. tries are the subject of Part VII; these include A second CD-ROM contains the teaching modules participation in the WTO, capacity building, and developed by Jaime de Melo and Marc Bacchetta implementation of WTO agreements. Finally, the during their many years of conducting an intensive chapters in Part VIII summarize a number of rules two-week course for government officials, cospon- of thumb for good trade policy and review sored by the World Bank Institute and the WTO. The approaches to using the WTO (and regional agree- CD-ROM also includes an extensive set of readings ments) as instruments for promoting development. and, as noted above, official WTO documents. xxx I TRADE POLICY REFORM IN CONTEXT I t is useful to make a distinction between appropriate macroeconomic incentive environment reduction in border barriers to trade and and embedded in a comprehensive development and "behind-the-border" complementary poli- poverty reduction strategy. Complementary institution- cies that are critical in supporting trade policy reforms. al reform efforts and improvements in the legal and The first set of policies focuses on creating incentives regulatory environment that increase investor confi- for efficient growth by reducing the average level and dence are vital if trade liberalization is to serve as an the dispersion of border protection, eliminating non- engine of growth. Key elements of the associated tariff barriers (NTBs), and strengthening the public behind-the-border trade agenda include efficient regu- institutions needed to ensure that goods cross fron- latory regimes, institutions that support the participa- tiers with low transactions costs (i.e., efficient customs tion of national firms in international markets, and regimes that minimize red tape). The second set has measures to enhance the competitiveness of these firms to do with regulatory standards and policies to ensure by providing access to crucial services inputs. Dani that supply responses to liberalization are efficient, Rodrik, in Chapter 1, examines some of these issues. equitable, and enduring. Important issues here The trade agenda has become increasingly com- include liberalization of trade in services sectors to plex in the past decade, and where it starts and enhance competitiveness, policies to promote access ends is not clear. What is clear is that the standard to information and technology, and strengthening of approach found in most textbooks--which focuses institutions in order to benefit from participation in on policy instruments that are applied at the border regional and multilateral trade arrangements. and that affect the domestic prices of goods or The chapters in this part focus on aspects of the com- export prices (tariffs, quotas, export subsidies, and plementary agenda. A strategy for sustained trade taxes)--is too narrow. In practice, the trade agenda expansion and growth must be framed within an spans all policies that have the effect of discriminat- 1 T R A D E P O L I C Y R E F O R M I N C O N T E X T ing against foreign providers (suppliers) of goods, A key message that emerges from the chapters in services, and production factors (knowledge, labor, this part is the need for analysis that focuses not and capital), and it takes in the functioning of insti- only on trade policy narrowly defined but also on tutions that affect the investment climate in a coun- the complementary reforms and institutions that are try. The recognition that trade policy has a much required if trade reforms are to benefit society. Such wider ambit than border policies implies that gov- analysis should include a diagnosis of the current sit- ernments and civil society must have a broad focus uation, benchmarking in relation to good practice and must consider the interrelationships between and competitors, determination of the incentive different policy areas and the operation and effec- and redistributional implications of status quo poli- tiveness of existing institutions. Kym Anderson, in cies and possible changes, and identification of the Chapter 2, emphasizes the need for an economy- complementary actions that are needed to make wide perspective on trade policy reform. trade reform an effective component of a poverty- Key complementary factors that often determine reducing growth strategy. the success of trade policy reform are the real exchange rate and the ability of the government to Further Reading maintain revenue collection objectives. As Howard J. Schatz and David G. Tarr document in Chapter 3, Jeffrey Sachs and Andrew Warner, "Economic Reform although countries may maintain different types of and the Process of Global Integration," Brookings exchange rate regimes, allowing the real exchange Papers on Economic Activity, 1 (1995): 1­118, is a rate to appreciate significantly over time has often widely read and influential empirical study that finds led to the failure of trade reforms. Chapter 4, by an unambiguous positive relationship between open- Liam Ebril, Janet Stotsky, and Reint Gropp, examines ness and economic performance. Dani Rodrik, Has the fiscal implications of trade liberalization. Tariff Globalization Gone Too Far? (Washington, D.C.: Insti- revenue remains important for many low-income tute for International Economics, 1997), provides a countries. In pursuing further tariff reform, efforts skeptical view of the benefits of globalization for have to be made to develop alternative domestic growth and welfare in the absence of the institutions tax bases and to ensure that reliance on tariff rev- and policies needed to manage downside risks. An enues does not needlessly distort resource allocation accessible account of the effects of the inward-look- incentives. Cross-country experience suggests that ing, import-substituting development strategies pop- policy reforms can be designed so as to maintain or ular in the 1960s and 1970s, as well as the shift increase revenue collection. toward more outward-looking policies in the 1980s, Although the available research indicates that trade is given by Jagdish Bhagwati in Protectionism (Cam- liberalization reduces poverty overall, segments of the bridge, Mass.: MIT Press, 1988). Edward Buffie, Trade poor may be hurt by it, and in Chapter 5, L. Alan Win- Policy in Developing Countries (Cambridge, U.K.: Cam- ters looks at the interactions between trade reform bridge University Press, 2001), analyzes trade policy and poverty alleviation. Reform programs supporting in an integrated framework that allows for economic liberalization must be complemented by efforts to dynamics and incorporates the structural features of strengthen social safety nets. Since some of the poor developing countries. Neil McCulloch, L. Alan Win- are likely to be so destitute that any decrease in ters, and Xavier Cirera, Trade Liberalization and Pover- incomes will impose extreme hardship, it is important ty: A Handbook (London: Centre for Economic Policy to identify which of them may be adversely affected Research, 2001), provides a comprehensive treat- by reforms and to determine the most appropriate set ment of the links between trade and poverty in the of policies to complement trade reform. context of the WTO. 2 1 D A N I R O D R I K Trade Policy made and implemented, estab- lishes new constraints and oppor- tunities for economic policy more Reform as broadly, creates a new set of stake- holders while disenfranchising the Institutional previous ones, and gives rise to a new philosophy (alongside a new Reform rhetoric) on what development policy is all about. Hence, trade reform ends up being much more than a change in relative prices: it results in institutional reform of a major kind. E In the language of economics, conomists are trained to think institutional reform changes not only policy param- about trade policy reform in terms eters but also behavioral relationships. Correspond- of changes in the levels of tariffs and quantitative ingly, the resource-allocation and dynamic restrictions (QRs) and the shifts in relative prices consequences of trade reform become harder to dis- brought about by these alterations. They use eco- cern using the type of analysis that is the applied nomic models, supplemented by quantitative esti- economists' stock in trade. Household behavior and mates of elasticities, to analyze the implications of investment decisions get altered in ways that are dif- changes in tariffs and QRs for production, con- ficult to track in the absence of knowledge about the sumption, and trade. By tweaking their models suf- "deep parameters" of the economy. When the ficiently, they can predict the likely impacts on reform is well designed and consistent with the employment, poverty and distribution, macroeco- institutional needs of the economy, it can spur nomic balances, and the government budget. If they unexpected levels of entrepreneurial dynamism and are ambitious (reckless?), they will also pass judg- economic growth. When it is not, it can result in a ment on dynamic efficiency, technological progress, stagnation that will appear surprising. and long-run economic growth. Viewing trade reform as institutional reform Policymakers often have a different perspective on helps clarify the criteria by which trade reform trade reform. For them, the actual changes in tariff should be evaluated. My main argument in this schedules are typically only a small part of the process. chapter is that the relevant criterion is neither open- What is at stake is a deeper transformation of the pat- ness to trade nor consistency with existing WTO terns of behavior within the public sector, and of the rules.1 The yardstick that matters is the degree to government's relationship with the private sector and which trade reform contributes to the construction the rest of the world. The reform goes beyond particu- of a high-quality institutional environment at home. lar levels of tariffs and QRs: it sets new rules and My working hypothesis, supported by empirical expectations regarding how these policy choices are evidence to which I will refer below, is that a high- 3 T R A D E P O L I C Y R E F O R M I N C O N T E X T quality institutional environment has greater eco- oriented reforms in Latin America and the growing nomic payoffs than a liberal trade regime or adher- realization that these reforms have paid too little ence to WTO rules. attention to mechanisms of social insurance and to In practice, there may be some important safety nets. The third and most recent is the Asian spillovers among these objectives. To cite an impor- financial crisis, which has shown that allowing tant illustration, a free trade regime is likely to financial liberalization to run ahead of financial reduce the corruption and rent-seeking associated regulation is an invitation to disaster. A number of with trade interventions. Similarly, tariff bindings recent empirical studies have highlighted the under the WTO may generate greater predictability importance of high-quality institutions in shaping in incentives and solidify property rights--two economic performance (see, especially, Kaufmann, important attributes of a high-quality institutional Kraay, and Zoido-Lobatón 1999; Acemoglu, John- framework. But while free trade and WTO rules can son, and Robinson 2000). contribute to the emergence of high-quality institu- Following Lin and Nugent (1995: 2306­07), it is tions, these are not one and the same. Institutional useful to think of institutions broadly as "a set of development takes time and often requires humanly devised behavioral rules that govern and unorthodox and divergent choices. Some of the shape the interactions of human beings, in part by most spectacular cases of development in the post- helping them to form expectations of what other war period have been the product of gradualist, people will do." All well-functioning market two-track modes of institutional reform (Rodrik economies are "embedded" in a set of nonmarket 2000b). The type of investments in institution- institutions, without which markets cannot per- building required for full adherence to WTO agree- form adequately. I will highlight below five types of ments on, say, customs valuation or intellectual market-supporting institutions in particular: prop- property rights (IPRs) may not be the first order of erty rights, regulatory institutions, institutions for business for low-income countries with more macroeconomic stabilization, institutions for social urgent needs (Finger and Schuler 2000). Since insurance, and institutions of conflict management. human resources, administrative capacity, and I emphasize as well the variety of institutional political capital are scarce, especially in developing setups that is compatible with superior economic countries, policymakers need to have a good sense performance. of the priorities. An implication of this line of reasoning is that we Property Rights should think of the trade regime and WTO rules as being at the service of developing countries' institu- As North and Thomas (1973) and North and Wein- tional needs, not vice versa. Governments that gast (1989), among many others, have argued, the understand this are the ones that are likely to make establishment of secure and stable property rights the most of trade reform. has been a key element in the rise of the West and the onset of modern economic growth. It stands to reason that an entrepreneur would not have the Institutional Prerequisites for Development incentive to accumulate and innovate unless s/he Price reforms--in external trade, in product and had adequate control over the return to the assets labor markets, in finance, and in taxation--were the that are thereby produced or improved. Note that rallying cry of the reformers of the 1980s, along the key word is "control," rather than "ownership." with macroeconomic stability and privatization. By Formal property rights do not count for much if the 1990s, it had become clear that incentives would they do not confer control rights. By the same not work, or would generate perverse results, in the token, sufficiently strong control rights may do the absence of adequate institutions. Three sets of dis- trick even in the absence of formal property rights. parate developments have conspired to put institu- In Russia today, shareholders have property rights tions squarely on the agenda of reformers. One of but often lack effective control over enterprises, these was the dismal failure in Russia of price whereas in China's township and village enterprises reform and privatization in the absence of a sup- control rights have spurred entrepreneurial activity portive legal, regulatory, and political apparatus. A even in the absence of clearly defined property second is the lingering dissatisfaction with market- rights. 4 Trade Policy Reform as Institutional Reform As these instances illustrate, establishing "proper- States has both the world's freest markets and the ty rights" is rarely a matter of just passing a piece of world's toughest antitrust enforcement. The lesson legislation. Legislation in itself is neither necessary that market freedom requires regulatory vigilance nor sufficient for the provision of secure control has been driven home recently by experience in East rights. In practice, control rights are upheld by a Asia. In the Republic of Korea and in Thailand, as in combination of legislation, private enforcement, so many other developing countries, financial liber- and custom and tradition. They may be distributed alization and capital account opening led to finan- more narrowly or more diffusely than property cial crisis precisely because of inadequate prudential rights. Moreover, property rights are rarely absolute, regulation and supervision. even when set formally in the law. Each society In developing countries, where market failures are decides for itself the scope of allowable property pervasive, regulatory institutions may need to rights and the acceptable restrictions on their exer- extend beyond the standard list covering antitrust, cise. Intellectual property rights are protected assid- financial supervision, securities regulation, and the uously in the United States and most advanced like. Recent models of coordination failure and cap- societies but not in many developing countries. By ital market imperfections make it clear that strategic contrast, zoning and environmental legislation government interventions may often be required to restricts the ability of households and enterprises in escape low-level traps and elicit desirable private rich countries to do as they please with their "prop- investment responses.2 The experience of Korea and erty" to a much greater extent than is the case in Taiwan (China) in the 1960s and 1970s can be inter- developing countries. All societies recognize that preted in that light. The extensive subsidization and private property rights can be curbed if doing so government-led coordination of private investment serves a greater public purpose. It is the definition in these two economies played a crucial role in set- of what constitutes a "greater public purpose" that ting the stage for self-sustaining growth. It is clear varies. that many other countries have tried and failed to replicate these institutional arrangements. And even Korea may have taken a good thing too far by main- Regulatory Institutions taining the cozy institutional linkages between the Markets fail when participants engage in fraudulent government and chaebols well into the 1990s, at or anticompetitive behavior. They fail when trans- which point these ties may have become dysfunc- actions costs prevent the internalizing of technolog- tional. Once again, the lesson is that desirable insti- ical and other nonpecuniary externalities. And they tutional arrangements vary, and that they vary not fail when incomplete information results in moral only across countries but also within countries over hazard and adverse selection. Economists recognize time. these failures and have developed the analytical tools required to think systematically about their Institutions for Macroeconomic Stabilization consequences and the possible remedies. Theories of the second best, imperfect competition, agency, Markets are not necessarily self-stabilizing. Keynes and mechanism design, to name but a few, offer an and his followers worried about shortfalls in aggre- almost embarrassing choice of regulatory instru- gate demand and the resulting unemployment. ments for countering market failures. Theories of More recent views of macroeconomic instability political economy and public choice offer cautions stress the inherent instability of financial markets against unqualified reliance on these instruments. and its transmission to the real economy. All In practice, every successful market economy is advanced economies have come to acquire fiscal overseen by a panoply of regulatory institutions that and monetary institutions that perform stabilizing regulate conduct in goods, services, labor, asset, and functions, having learned the hard way about the financial markets. A few acronyms from the United consequences of not having them. Probably most States will suffice to give a sense of the range of important among these institutions is a lender of institutions involved: FTC, FDIC, FCC, FAA, last resort--typically, the central bank--that guards OSHA, SEC, EPA, and so on. In fact, the freer are the against self-fulfilling banking crises. markets, the greater is the burden on regulatory There is a strong current within macroeconomics institutions. It is not a coincidence that the United thought that disputes the possibility or effectiveness 5 T R A D E P O L I C Y R E F O R M I N C O N T E X T of stabilizing the macroeconomy through monetary represented well by the Japanese case, is one in and fiscal policies. There is also a sense in policy cir- which social insurance is provided through a com- cles, particularly in Latin America, that fiscal and bination of enterprise practices (such as lifetime monetary institutions, as currently configured, have employment and enterprise-provided social bene- added to macroeconomic instability, rather than fits), sheltered and regulated sectors (mom-and- reduced it, by following procyclical rather than anti- pop stores), and an incremental approach to cyclical policies. These developments have spurred liberalization and external opening. the trend toward central bank independence and Social insurance legitimizes a market economy have helped open a new debate on designing more because it renders it compatible with social stability robust fiscal institutions. Some countries (Argenti- and social cohesion. But the existing welfare states na being the most significant example) have given in Western Europe and the United States engender a up on a domestic lender of last resort altogether by number of economic and social costs--mounting replacing their central bank with a currency board. fiscal outlays, an "entitlement" culture, long-term The debate over currency boards and dollarization unemployment--that have become increasingly illustrates the obvious, but occasionally neglected, apparent. Partly because of this experience, devel- fact that the institutions needed by a country are oping countries, such as the countries in Latin not independent of that country's history. America that adopted the market-oriented model following the debt crisis of the 1980s, have not paid sufficient attention to creating institutions of social Institutions for Social Insurance insurance. The upshot has been economic insecuri- One of the liberating effects of a dynamic market ty and a backlash against the reforms. How these economy is that it frees individuals from their tradi- countries will maintain social cohesion in the face tional entanglements--the kin group, the church, of large inequalities and volatile outcomes, both of the village hierarchy. The flip side is that it uproots which are being aggravated by the growing reliance them from traditional support systems and risk- on market forces, is an important question that has sharing institutions. Gift exchanges, the fiesta, and no obvious answer. kinship ties--to cite just a few of the social arrange- ments for equalizing the distribution of resources in Institutions of Conflict Management traditional societies--lose many of their social insurance functions. And as markets spread, the tra- Societies differ in their cleavages. Some are made up ditional ways of managing the risks that have to be of an ethnically and linguistically homogenous insured against become much less effective. A mod- population marked by a relatively egalitarian distri- ern market economy is one where idiosyncratic bution of resources. Others are characterized by (individual-specific) risk to incomes and employ- deep cleavages along ethnic or income lines. These ment is pervasive. divisions often hamper social cooperation and The huge expansion of publicly provided social engender social conflict. Economists have used insurance programs during the 20th century is one models of social conflict to shed light on questions of the most remarkable features of the evolution of such as: Why do governments delay stabilizations advanced market economies. In the United States it when delay imposes costs on all groups? Why do was the trauma of the Great Depression that paved countries rich in natural resources often do worse the way for major institutional innovations in this than countries that are resource-poor? Why do area: social security, unemployment compensation, external shocks often lead to protracted economic public works, public ownership, deposit insurance, crises that are out of proportion to the direct costs and legislation favoring unions. In Europe the roots of the shocks themselves? of the welfare state reached in some cases to the tail Healthy societies have a range of institutions that end of the 19th century. But the striking expansion make such colossal coordination failures less likely. of social insurance programs, particularly in the The rule of law, a high-quality judiciary, representa- smaller economies most open to foreign trade, was tive political institutions, free elections, independent a post­World War II phenomenon. Social insurance trade unions, social partnerships, institutionalized need not always take the form of transfer programs representation of minority groups, and social insur- paid out of fiscal resources. The East Asian model, ance are examples of such institutions. What makes 6 Trade Policy Reform as Institutional Reform these arrangements function as institutions of con- tries that face the prospect of adopting the EU's flict management is that they entail a double "com- Common Agricultural Policy or its antidumping mitment technology": they warn the potential regime. It all depends on the circumstances and on "winners" from social conflict that their gains will be how national governments are able to use such cir- limited and assure the "losers" that they will not be cumstances. expropriated. They tend to increase the incentives One way that governments can use institutional for social groups to cooperate by reducing the payoff arbitrage to good effect is to enhance the credibility to socially uncooperative strategies. of domestic institutions. For example, the new disci- plines imposed on developing country governments by the WTO--in the areas of tariff bindings, quanti- Trade Policy and Institutional Reform tative restrictions, services, subsidies, trade-related What is the link between trade policy reform and investment measures (TRIMs), and intellectual these institutions? Trade reform often entails the property--can be viewed as helping these govern- importation of institutions from abroad. Sometimes ments overcome traditional weaknesses in their style this is the outcome of deliberate policy actions to of governance. These disciplines impose a certain "harmonize" a country's economic and social insti- degree of predictability, transparency, rule-bound tutions with those of its trading partners. Member- behavior, and nondiscrimination in areas of policy ship in the WTO, for example, requires the adoption often subject to discretion and rent-seeking. In the of a certain set of institutional norms: nondiscrimi- same vein, perhaps the greatest contribution of the nation in trade and industrial policies, transparency North American Free Trade Agreement (NAFTA) to in the publication of trade rules, WTO-consistent the Mexican economy was the element of irre- patent and copyright protection, and so on. Similar- versibility and "cementing" that the agreement has ly, membership in the European Union (EU) contributed to Mexico's economic reforms. In requires the adoption of wide-ranging legal and Europe the accession of Greece, Portugal, and Spain bureaucratic requirements set down in Brussels. to the EU has made return to military dictatorship in At other times, institutional arbitrage is the result those countries virtually unthinkable. of the working out of market forces. Mobility of Imported institutions, however, can also turn out employers around the world, for example, makes it to be ill suited or counterproductive. Many of the harder to tax corporations and tilts national regimes labor standards that some labor groups in the North toward the taxation of nontraded goods and factors, would like developing countries to adopt--such as such as labor. Financial integration raises the premi- higher minimum wages or restrictions on some um for macroeconomic stability and makes central kinds of child labor--may fit in this category. The bank independence look more desirable. Finally, new patent restrictions called for by the Trade-Relat- openness can change national institutions by alter- ed Aspects of Intellectual Property Rights (TRIPS) ing the preferences that underlie them. Civil liber- agreement of the WTO are at best a mixed blessing ties and political freedoms are among the most for countries such as India that have so far benefited important imported concepts in the developing from cheap pharmaceuticals. A similar argument world; the demands for democracy to which these can be made about pressures for tightening environ- ideas give rise are a direct product of openness in mental standards in developing countries. this broad sense. Successful institutional reforms typically combine Arbitrage in markets for goods and capital, in the imported blueprints with local flavor. A good exam- absence of second-best complications, is associated ple of this in the area of trade comes from Mauri- with normatively desirable outcomes; it increases tius, where superior economic performance has efficiency. One cannot make the same presumption been built on a peculiar mix of orthodox and het- where arbitrage in institutions is concerned. There erodox strategies. This economy's success derives in are no theorems stating that institutional conver- large part from an export-processing zone (EPZ), gence, harmonization, or "deep integration" which operates under free trade principles. The EPZ through trade is inherently desirable. While many of has enabled a boom in exports of garments to Euro- the examples cited above involve outcomes that are pean markets and an accompanying investment desirable (greater democracy, for instance), this is boom at home. Yet the island's economy has com- not true of all possible outcomes. Think of the coun- bined the EPZ with a domestic sector that was high- 7 T R A D E P O L I C Y R E F O R M I N C O N T E X T ly protected until the mid-1980s. The origins of this tary and phytosanitary measures (SPS), and intel- essentially dual-track strategy (not unlike that fol- lectual property rights (IPRs)--a sum equal to a lowed in China) lay in the social and political make- year's development budget for many of the least- up of the island and in the decision by policymakers developed countries. Would this be money well not to disrupt a fragile ethnic situation through an spent? Finger argues that for the vast majority of across-the-board liberalization that would have dis- developing countries, the answer is no. Although advantaged established import-substituting groups. these countries would benefit from the strengthen- The EPZ scheme, in fact, provided a neat way ing of their institutions in the relevant areas, the around the political difficulties. The creation of the reality is that "WTO obligations reflect little aware- EPZ generated new opportunities in trade and ness of development problems.""Other alternatives, employment without removing protection from the e.g., basic education for women and girls, would import-substituting groups or from the male work- have much more attractive rate-of-return numbers" ers who dominated the established industries. The (Finger 1999). It is a safe bet that any new trade segmentation of labor markets early on between round will shorten the leash on developing coun- male and female workers, with women predomi- tries further, even if pressure in the controversial nantly employed in the EPZ, was crucial, as it pre- areas of environment and labor can be fended off. vented the expansion of the EPZ from driving wages Integration into the world economy has other, up in the rest of the economy and hurting import- more subtle institutional requirements, as well. substituting industries. New profit opportunities Openness implies heightened exposure to external were created at the margin while leaving old oppor- risk and, consequently, greater demand for social tunities undisturbed. insurance. Greater provision of social insurance One can cite other instances of heterodox trade seems to be a key factor behind the empirical regu- reforms that proved successful because they suited larity that governments tend to be bigger in existing political and institutional realities. Korea's economies where trade makes up a higher share of outward orientation during the 1960s, for example, GDP (Rodrik 1998). More broadly, openness was achieved not through import liberalization (of increases the premium on institutions of conflict which there was little), but through export subsi- management (Rodrik 1999). dization (of which there was a lot). This type of It is often overlooked that the most successful reform is now prohibited under existing WTO rules "globalizers" of an earlier era--the East Asian on subsidies. Similarly, China's two-track reform "tigers"--had to abide by few international con- strategy in agriculture, industry, and trade, which straints and had to pay few of the costs of integra- maintained nonmarket institutional forms while tion during their formative growth experience in aligning incentives correctly at the margin, has been the 1960s and 1970s. Global trade rules essentially wildly successful. These are cases in which imagina- gave them a free ride, and capital mobility was hard- tive experimentation with institutional reform has ly an issue. This is why these countries can hardly be had, in all likelihood, greater payoffs than the considered poster children for today's globalization. wholesale transplantation of institutions from Korea, Taiwan (China), and the other East Asian advanced industrial countries would have had.3 economies had the freedom to do their own thing, and they used it abundantly. As noted above, they combined their reliance on trade with unorthodox Integration into the World Economy as a policies--export subsidies, domestic content Model of Institutional Reform requirements, import-export linkages, patent and WTO membership entails institutional reforms that copyright infringements, restrictions on capital are not only demanding, but also of a particular flows (including on foreign direct investment), kind. One can question, as Michael Finger has elo- directed credit, and so on--that are either preclud- quently done, the fit between these reforms and the ed by today's rules or greatly frowned on. The envi- needs of developing countries, particularly of the ronment for today's globalizers is quite different. least developed among them. Finger has calculated None of the institutional reforms needed for that it would cost a typical developing country insertion in the world economy is bad in and of US$150 million to implement requirements under itself, and in fact, many of them can be indepen- the WTO agreements on customs valuation, sani- dently desirable, as I argued above. Some can also 8 Trade Policy Reform as Institutional Reform have unintended benefits. For example, a govern- phenomena (overvalued currencies or macro insta- ment that is forced to protect the rights of foreign bility) or geographic determinants (such as location investors perhaps becomes more inclined to protect in the tropical zone) to trade policies proper. Once the basic human rights of its own citizens, too. This simple corrections are made for such problems, one was a potent argument in U.S. debates, prior to rarely finds a statistically significant relationship China's accession to the WTO, about China's per- between the level of tariff and nontariff barriers and manent normalized trade relations (PNTR) status. economic growth across countries. But one has to recognize that a strategy of institu- There are, in fact, reasons to be skeptical about tional reform based on global integration is a strate- the existence of a general, unambiguous relation- gy of trickle-down institutional reform. The ship between trade openness and growth. The rela- reforms may or may not trickle down, and even tionship is likely to be a contingent one, dependent when they do, they will rarely constitute the most on a host of country and external characteristics. effective way of targeting the desired ends, whether The fact that practically all of today's advanced those ends are legal reform, improved observance of countries embarked on their growth behind tariff human rights, or reduced corruption. Institutional barriers and reduced protection only subsequently change is costly and requires the expenditure of surely offers a clue of sorts. Moreover, the modern scarce human resources, administrative capabilities, theory of endogenous growth yields an ambiguous and political capital. The priorities implied by glob- answer to the question of whether trade liberaliza- al insertion will not always coincide with the priori- tion promotes growth. The answer varies depending ties of a more fully developmental agenda. on whether the forces of comparative advantage push the economy's resources in the direction of activities that generate long-run growth (via exter- Can We Rely on a Growth Payoff from nalities in research and development, expansion of Openness? product variety, upgrading of product quality, and Global integration carries opportunity costs because so on) or divert them from such activities. Finally, as of the institutional consequences that such a strategy I have stressed throughout, the institutional setting entails. These costs have to be traded off against the in which trade policy operates is more important expected benefits. All economists know that gains for economic performance than the levels at which from trade exist, but the standard gains from trade specific trade barriers are set. tend to be small. The tendency in policy discussions No country has developed successfully by turning has been to go considerably beyond the standard case its back on international trade and long-term capi- for trade and to claim that open trade policies pro- tal flows. Very few countries have grown over long duce significant boosts in economic growth rates. periods of time without experiencing an increase in This claim is apparently supported by a large cross- the share of foreign trade in their national product. national empirical literature. Recently, Francisco In practice, the most compelling mechanism that Rodríguez and I reviewed the extensive literature on links trade with growth in developing countries is the relationship between trade policy and growth that imported capital goods are likely to be signifi- (Rodríguez and Rodrik 2001) and reached the con- cantly cheaper than those manufactured at home. clusion that there is a significant gap between the Policies that restrict imports of capital equipment, message that the consumers of this literature have raise the price of capital goods at home, and thereby derived and the "facts" that the literature has actually reduce real investment levels have to be viewed as demonstrated. The gap emerges from a number of undesirable prima facie. Exports, in turn, are factors. In many cases the indicators of "openness" important, since that is what one purchases import- used by researchers are problematic as measures of ed capital equipment with. trade barriers or are highly correlated with other But it is equally true that no country has devel- sources of poor economic performance. In other oped simply by opening itself up to foreign trade cases the empirical strategies used to ascertain the and investment. The trick in the successful cases has link between trade policy and growth have serious been to combine the opportunities offered by world shortcomings, the removal of which results in signif- markets with a domestic investment and institu- icantly weaker findings.4 One common problem has tion-building strategy to stimulate the animal spir- been the misattribution of either macroeconomic its of domestic entrepreneurs. Almost all of the 9 T R A D E P O L I C Y R E F O R M I N C O N T E X T outstanding cases--East Asia, China, India since the along these lines. I have argued here that the first early 1980s--involve partial and gradual opening question policymakers contemplating trade reform up to imports and foreign investment. should ask is not whether the reform will result in The appropriate conclusion to draw from the evi- higher volumes of trade, render their trade regime dence is not that trade protection should, as a rule, more liberal, or increase market access abroad but be preferred to trade liberalization. There is no evi- whether it will improve the quality of institutions at dence from the past 50 years that trade protection is home. The results of trade negotiations--whether systematically associated with higher growth. The bilateral, regional, or multilateral--should be point is simply that the benefits of trade openness judged by the same yardstick. should not be oversold. When other worthwhile policy objectives are competing for scarce adminis- Notes trative resources and political capital, deep trade liberalization often does not deserve the high prior- This chapter draws heavily on several earlier papers, in particular ity it typically receives in development strategies. Rodrik 1999, 2000a, and 2000b. This is a lesson that is of particular importance to 1 It should go without saying that openness to trade and adher- countries, such as those in Africa, that are in the ence to WTO rules are not the same thing. A country can fol- early stages of reform. low free trade policies without being a member of the WTO, and many WTO rules are at variance with free trade (as in the cases of antidumping, safeguards, and regional agreements). Conclusion 2 See Hoff and Stiglitz (2000) for a useful survey and discussion. A high-quality policy environment is one that sends 3 See Kapur and Webb (2000) and Pistor (2000) for useful dis- clear signals to producers and investors, precludes cussions of the limitations of importing legal and institutional rent-seeking, does not waste economic resources, is forms from abroad. consistent with the administrative capabilities of the 4 Our detailed analysis covers the five papers that are probably government, and maintains social peace. Trade pol- the best known in the field: Dollar (1992); Sachs and Warner icy reform contributes to economic development (1995); Ben-David (1993); Edwards (1998); and Frankel and insofar as it helps build high-quality institutions Romer (1999). 10 2 K Y M A N D E R S O N Economywide taking an economywide perspec- tive when considering the effects of actual policies at home or Dimensions of abroad or of potential policy reforms. Given the significance of Trade Policy and agriculture in low-income coun- tries, the chapter focuses primari- Reform ly on the possible direct and indirect effects of policies on this sector, emphasizing the need to consider the impact of input as well as output price distortions on producer incentives. E very country has an interest in Direct Effects of Policies: A Single-Sector trade policy reform. This is true Perspective even for the most open of economies because, although that government may not be distorting Historically, the governments of poor agrarian incentives, government policies of many other economies have taxed farmers in one way or anoth- countries are distorting the prices received by the er (Krueger, Schiff, and Valdés 1988). Sometimes it open economy's exporters in international markets. has been an in-kind tax, such as a proportion of Moreover, it is relative prices that matter: the incen- grain output. In other settings, where a cash crop tives facing producers or consumers of a particular was being exported, producers often have been product can be distorted not only by policies direct- required to sell to a statutory marketing authority ly affecting the price of that product but also, and that paid them only a fraction of the export price. sometimes even more strongly, by policies affecting Either way, farmers receive less than the free-market the prices of products that are substitutes or com- price for their produce. Except in the unlikely event plements in production or consumption. Govern- that all of those taxes come back to farmers in the ment intervention in currency markets also can form of government goods and services they other- have nontrivial distortionary effects on incentives. wise would have purchased with that taxed income, Farmers, for example, may receive the international the incentive to produce and market farm products price for their produce and yet be harmed by having is reduced. to convert from foreign to domestic currency at an Governments of such agrarian economies typical- artificially low exchange rate. ly return little of the proceeds of those taxes to farm This chapter explores not just the direct but also the families, especially at early stages of the country's various indirect ways in which trade and trade-related development. Rather, the taxes tend to be used to policies affect the welfare of people in developing develop urban infrastructure, pay officials relatively countries. Its purpose is to identify the importance of high wages, subsidize food consumption, and so on. 11 T R A D E P O L I C Y R E F O R M I N C O N T E X T Until recently, it was widely believed that taxing pri- import-protection policy encourages the allocation mary producers for such purposes would not reduce of too many resources to agriculture's import-com- output significantly because farm families were poor peting industries, and it also harms consumers of or had no alternative uses for their time, land, and those importables via higher food prices. other resources. Empirical studies during the past What implications does the Krueger, Schiff, and half-century, however, have shown that farmers in Valdés study have for reforming agricultural policy even the poorest settings are quite price-responsive in the average developing country? Reducing export (Askari and Cummings 1977). When the proceeds restrictions would cause the domestic price of from growing a marketable product are reduced, exported farm products to rise by up to one-eighth, farm households divert at least some of their helping producers of those exportables but hurting resources to producing other products or to leisure domestic buyers of the products (who may be down- pursuits. Only the very poorest subsistence farmers stream processors). That reform might also encour- might be enticed by such taxes to work harder, but age producers of import-competing farm products even that response may be welfare reducing in that to switch their production to exportables that now they then have less recreational time and are likely to carry higher prices. If the country's food import live less healthy and shorter lives. restrictions also were reduced, producers of those It matters that farm household resources are importables would see their output price decline diverted from producing the taxed good because and would consider switching to other farm prod- such diversion means that society's resources are ucts. This would reinforce the encouragement of not being used where they are most productive. A exportable production in agriculture, insofar as the farmer discouraged from specializing in growing a resources used in the two different farm subsectors cash crop, for example, has less to spend on other are substitutable. Both types of reform improve the products and therefore is less able to encourage oth- efficiency of resource use in the sector by encourag- ers to specialize in doing what they do best also. ing greater exploitation of the country's agricultural Likewise, it matters if farmers have to pay more for comparative advantage. That is, reforms that boost inputs purchased from nonfarm sectors (for exam- the relative profitability of the industries previously ple, because of import taxes on these goods): they discouraged by the government's trade-restrictive then buy less of those inputs in relation to other policies tend to be welfare enhancing. inputs than is optimal. This direct effect is, however, of much less economic importance than the indi- Indirect Effects of Policies: An Intersectoral rect effect of such industrial or service sector pro- Perspective tection policies, as is made clear in the next section. Of course, not all agricultural producers in devel- The above lesson applies not just within the agricul- oping countries face artificially depressed prices for tural sector but also to interactions between it and their products. Indeed import-competing produc- other sectors. That is, farmers also can be discour- ers of some key food items enjoy protection from aged, albeit indirectly, by nonagricultural policy import competition that raises the domestic price of interventions. One source of such discouragement their produce above free-market levels. An empiri- comes from import protection to producers of non- cal study of 18 developing countries between the farm products. In an economy producing just two mid-1970s and the mid-1980s contrasted the treat- sets of goods, importables and exportables, a tax on ment of major farm export products with that of imports is equivalent to a tax of equal size on key imported foods (Krueger, Schiff, and Valdés exports whenever the two sets of goods use com- 1988). The authors found that domestic prices of mon resources such as labor and capital. Both taxes the imported foods averaged about 20 percent raise the prices of importables relative to exporta- above prices at the countries' borders, whereas bles, and by the same amount, and it is that price domestic prices for key agricultural export items ratio which determines the allocation of resources were, on average, 11 percent below international between the two sectors (Lerner 1936). levels. Both types of distortion are harmful to More generally, when domestic prices of some national economic welfare: whereas depressed industrial or services sector products are raised arti- export prices result in too few resources being ficially by restrictions on their importation or by devoted to the production of those exportables, an other price-support measures, resources are drawn 12 Economywide Dimensions of Trade Policy and Reform to those import-competing sectors at the expense of toral resource use remain and could worsen when other industries in the primary sectors, including the average level of agricultural assistance is raised. export industries (Clements and Sjaastad 1984). Second, the economy is invariably made up of more Historically, industrial tariffs have been a major than just those two sectors, so similar levels of assis- source of indirect discrimination against agricul- tance would have to be provided to fishing, mining, ture, but a wide range of other distortionary mea- and other sectors to ensure an overall improvement sures is found in service industries as well. in the efficiency of national resource use. Third, sup- The importance of this cause of inefficient pose the farm assistance were to be provided via, say, resource allocation cannot be stressed enough, as it input subsidies for fertilizer and water, as in fact has crucial implications for reform. Two examples often happens, even in poor countries. It turns out will illustrate the point. Taking, again, the average that agricultural support via input policies would be country in the Krueger, Schiff, and Valdés study, less efficient and possibly even counterproductive suppose that the agricultural sector is a net exporter because it would encourage the use of only a subset (which means the country is a net importer of non- of inputs rather than all farm inputs (Warr 1978). farm products) and that the food-importing sub- Most damning of all, manufacturers would perceive sector is almost as large as the agricultural their situation as deteriorating if support for pri- export­focused subsector. Within agriculture, the mary production were increased and, if no change in restrictions that reduce the domestic price of farm the political-economy forces had been at work, they exportables by 11 percent and raise the price of food would presumably demand a return to the status importables by 20 percent would boost the overall quo ante, perhaps through another hike in industri- average price of farm products, but by less than 10 al tariffs. Clearly, tariff compensation to farmers is a percent. Taking a single-sector perspective as in the far riskier reform strategy for improving the use of previous section might lead one to believe that national resources than the first-best strategy of eliminating those agricultural policies and thereby reducing industrial tariffs. reducing farm prices on average would be welfare improving. The Additional Indirect Effect of Distorting As it turns out, however, such a conclusion would Exchange Rates follow only if there were no distortions in the rest of the economy. If the manufacturers in this economy The Krueger, Schiff, and Valdés (1988) study also were to enjoy an average nominal rate of protection examined the extent to which unsustainable current from import competition of, say, 25 percent (for account deficits, overvalued official exchange rates, example, as a result of a uniform 25 percent tariff), and the like artificially inflate the value of a nation's then, prior to reform and notwithstanding the posi- currency from the viewpoint of farmers. Such poli- tive direct assistance to farmers, there would already cies encourage the production (and discourage the be too many resources in industrial relative to agri- domestic consumption) of nontradables relative to cultural pursuits. In that case reducing support for tradables and thereby represent another source of farming would be likely to exacerbate the inefficient inefficiency in national resource use and another resource allocation rather than improve it. To disincentive to farm. ensure a welfare-improving policy reform in this Empirically, for the 18 countries studied by case, it would be necessary to first lower the degree Krueger, Schiff, and Valdés, these macroeconomic of assistance to manufacturers and then, when the policies proved less of a disincentive to agricultural level of assistance to industry equaled that to farm- producers than did industrial protectionism. ers, phase down both simultaneously.1 Nonetheless, they added to farmers' difficulties. If it is too difficult politically to lower tariff protec- Together, the indirect negative impact of industrial tion to manufacturers, might a similar national wel- and macroeconomic policies on farmers' incentives fare improvement be achievable by raising the level was two-and-a-half times as large as the direct neg- of assistance to agriculture? In theory, maybe, but in ative effects of agricultural export policies in the practice such a tariff-compensation strategy would decade 1974­84, equivalent to depressing the price be unwise, on a number of grounds. First, if rates of of farm exportables by 38 percent, compared with assistance to different industries within each of the just 11 percent by direct measures. This indirect dis- two sectors are not equal, inefficiencies in intrasec- incentive also applied to import-competing farm- 13 T R A D E P O L I C Y R E F O R M I N C O N T E X T ers. In this sample of developing countries, the latter used intensively in the industry in which it has a enjoyed direct nominal protection of 20 percent in technological advantage. that decade, so even the most favored farmers in For nationalistic and cultural reasons, permanent those countries were being disadvantaged by the immigration of labor has not been made easy in dominance of the adverse indirect effects of non- recent decades, but numerous countries have toler- agricultural policies on agricultural incentives. ated temporary movements of labor, bringing mutu- What would be the economywide implications of al gains to the countries involved. Much more reducing import tariffs in the above case? Reducing important in the past two decades, however, has the food import restrictions alone would probably been the growth in movement of capital across boost production of exported farm goods, which national borders. Foreign direct investment can would improve resource allocation within the farm bring with it not just financial capital but also mana- sector. But it would also free mobile resources that gerial and marketing skills, technological knowl- could then move to nonfarm activities, which, on edge, and intellectual property--forms of capital average, are more protected than farming. Hence, that foreign firms might not be willing to see export- whether the overall efficiency of national resource ed if they were unable to retain control over them. use would rise or fall is an empirical question if only Developing countries seeking to exploit fully their a subset of import restrictions and exchange rate comparative advantages therefore need to relax their distortions is to be removed. Only if the most pro- restrictions on foreign investment inflows. By the tected industries were to be liberalized first would same logic they also need to allow foreign invest- resources necessarily move to less protected indus- ment outflows so that domestic owners of capital tries and sectors and thereby guarantee an improve- also can earn the highest rewards possible. ment in the efficiency of the use of these resources in producing tradables. Even then, there is the possibil- The Dynamic Consequences of ity that those mobile resources would move into the Trade Reform production of more nontradables if the currency remained overvalued. This is the reason for the value Freeing up trade in goods, services, currencies, and of comprehensive reform that simultaneously frees capital not only improves the efficiency of national trade in goods, services, and currencies. resource use and consumer welfare at a point in time but also contributes to economic growth. The mech- anisms by which openness contributes to growth are What about Markets for Factors of gradually becoming better understood, thanks to the Production? pioneering work of such theorists as Grossman and An economywide perspective on trade reform would Helpman (1991) and Rivera-Batiz and Romer be incomplete unless it also extended to restrictions (1991). In a helpful survey of the subsequent litera- on factor flows. Theorists in the 1950s pointed to the ture, Taylor (1999) identifies several channels possibility that trade in goods could be a complete through which openness to trade can affect an econ- substitute for trade in productive factors in terms of omy's growth rate. They include the scale of the mar- both the volume of product trade and the welfare ket when knowledge is embodied in the products gains from trade (Mundell 1957). That theoretical traded, the effect of knowledge spillovers, and the possibility holds only under fairly restrictive condi- degree to which redundant creation of knowledge is tions, however. More recently, attention has been avoided through openness. More important from a drawn to the possibility that trade in some produc- policymaker's viewpoint, the available empirical evi- tive factors complements rather than substitutes for dence strongly supports the view that open trade in products (Markusen 1983). That can hap- economies grow faster (Edwards 1993; USITC 1997). pen when other productive factors are sector-specif- ic and goods trade is thus insufficient to equalize What if Trade Reform Harms the factor prices across countries. In that case trade in Environment? internationally mobile factors can generate further welfare gains from trade. It can also happen when Ideally, in adopting an economywide perspective, all there are differences in technologies across coun- significant influences of trade reform on human tries; then each country should import the factor welfare should be considered. That could include a 14 Economywide Dimensions of Trade Policy and Reform whole range of so-called noneconomic policy the Krueger, Schiff, and Valdés (1988) study shows, objectives, as well as standard economic effects such the indirect effect of nonagricultural and macroeco- as those on the natural environment, poverty, nomic policies on farmers' welfare can be several unemployment, food security, and distribution of times as large as the direct influence on incentives of income and wealth across regions and households. agricultural polices affecting export-oriented farm- Space is not available to discuss each of these here, ers. This is also true within a sector, and even more but excellent treatments are available in such books so to the extent that productive factors are more as Corden (1997). The main conclusion to be drawn readily substitutable within than between sectors. from that literature is that whatever the domestic In lobbying for trade reform, care is needed to policy objectives one has in mind, trade policy ensure that trade liberalization is not accompanied instruments are virtually never first-best ways of or followed by "re-instrumentation" of support. achieving those objectives. There are numerous ways to support producers This conclusion does not mean that trade reform other than through trade policy, and many of them can be undertaken without regard for society's are even more inefficient than trade measures. It other objectives. Welfare improvement via trade lib- would be counterproductive to lobby for the eralization cannot be guaranteed if optimal domes- removal of a trade restriction if it led to such an tic policies are not in place. There is no better inferior replacement. illustration of this than with respect to the natural A major aspect of exporters' lobbying activities environment. Reducing restrictions on exports of often involves encouraging the removal of impedi- logs, for example, in the absence of any other forest ments to market access abroad. Here again, an econ- resource policies is likely to lead to excessive defor- omywide perspective is needed (as is vigilance in estation. Another example is the reduction in Mon- preventing re-instrumentation). Consider, for golia's export tax on cashmere, which encouraged example, the interests of developing countries with the excessive grazing of common pastures. In these a strong comparative advantage in agriculture. They and in many other such cases overexploitation was would be likely to benefit directly from reduction in the result of property rights being poorly defined or agricultural protectionism in advanced industrial poorly policed. Clearly, better resource and environ- countries, but they could also benefit, albeit indi- mental policies are required before optimal social rectly, from a reduction in manufacturing protec- welfare can be achieved. tion in those same countries. The most obvious Note, however, that those resource and environ- example is a reduction in the very high barriers to mental policies are warranted, regardless of the imports of textiles, clothing, and footwear. Greater degree of openness of the economy. All that trade global production and trade in those products reform requires in addition is that the levels of envi- would result from reduced protection, with the out- ronmental policy intervention be adjusted when put expansion concentrated in newly industrializ- trade is liberalized to ensure that any additional ing countries. A direct consequence would be an environmental damage which accompanies opening expanded demand for cotton, wool, and leather up is matched in value terms with the marginal inputs--but that is only part of the impact on gains from trade expansion. Of course, trade reform agrarian developing countries. Probably more need not cause additional environmental damage; important is that such reform would speed the at least equally possible is the prospect that the industrialization of the more densely populated changes in production and consumption that developing countries, which would attract resources accompany trade liberalization will actually reduce away from their farm sectors. An indirect conse- pollution or resource depletion (Anderson 1997). quence, therefore, would be increased demand for food imports by those newly industrializing coun- tries. This suggests there is scope for agrarian and Implications for Reform-Minded Producers newly industrializing developing countries to act and Trade Policymakers collectively in pushing hard for greater market A clear implication of this economywide perspec- access for farm and textile products in advanced tive for producers seeking to influence government economies. In return, developing countries would policy is that their focus should not be confined to be expected to provide more access to their markets measures directly affecting their own industries. As for the goods and services exported by advanced 15 T R A D E P O L I C Y R E F O R M I N C O N T E X T economies--another dimension of the intersectoral connectedness of the global economy.2 Notes 1 In practice, a greater degree of refinement is possible, taking into account not only input price distortions (to get a measure of effective assistance to value added rather than just the nom- inal boost to the price of output) but also the degree of inter- sectoral substitutability or complementarity in production and consumption. See Corden (1971); Vousden (1990, ch. 9). 2 For recent empirical studies on the economywide effects of global trade reform and their implications for farm production and trade, see, for example, Hertel and others (forthcoming); Anderson, Hoekman, and Strutt (2001). 16 3 H O W A R D J . S H AT Z D A V I D G . TA R R Exchange Rate Although as a group, develop- ing countries progressively liber- alized their trade regimes during Overvaluation the 1980s and 1990s, some gov- ernments continue to take and Trade actions to defend their exchange rates that are counter to their Protection long-run trade liberalization efforts. One classic pattern is to attempt to defend an overvalued exchange rate through protec- tionist trade policies.3 Experi- ence shows that protection in A defense of an overvalued lthough both fixed and flexible exchange rate will significantly retard the country's exchange rate systems (and their medium-to-long-run growth prospects. In fact, an variants) have their advantages and disadvantages, overvalued exchange rate is often the root cause of more than half the countries in the world maintain protection, and the country will be unable to return fixed or managed exchange rates.1 While we do not to the more liberal trade policies that allow growth discuss the relative merits of these exchange rate without exchange rate adjustment. systems in this chapter, we note that as a practical Moreover, a devaluation of the nominal matter, exchange rate management in many coun- exchange rate appears to be a necessary condition tries in the world has resulted in overvaluation of for achieving a large depreciation of the real the real exchange rate, in some cases leading to gross exchange rate, as virtually all real devaluations distortions.2 (For further discussion of the links (above 25­35 percent) have been accompanied by between trade and macroeconomic management, nominal devaluations (Ghei and Hinkle 1999). see the CD-ROM, "Applied Trade Policy," that Sustained efforts to use downward adjustment of accompanies this Handbook.) wages and prices as a means of restoring a compet- Since governments are frequently confronted itive real exchange rate have frequently led to with the problems of external shocks and external severe recessions or depressions. trade deficits in the context of a fixed exchange rate Worldwide experience has shown that defending regime, a concise survey of worldwide experience the exchange rate has no medium-run benefits, with the effects of overvalued exchange rates in since falling reserves will eventually force devalua- terms understandable to policymakers should be tion. It is better that the devaluation be accom- useful. This chapter presents theory, cross-country plished without further debilitating losses in econometric evidence, and important case studies reserves and lost productivity due to import con- of the effects of overvalued exchange rates. trols. Experience with devaluations shows that after 17 T R A D E P O L I C Y R E F O R M I N C O N T E X T the devaluation, the exchange rate will reach a new result, less foreign exchange is available for need- equilibrium and that the equilibrium is strongly ed imports. influenced by the policies of the central bank and · Foreign exchange may be rationed and allocated the government. inefficiently by the government. · Efforts to defend an overvalued exchange rate through very tight monetary policy can plunge The Problems of an Overvalued the country into severe recession. Exchange Rate Countries that attempt to maintain overvalued The Need to Restore Internal Balance exchange rates significantly impede their growth in the medium to long term. Theory, cross-country When a country experiences a deficit in its trade bal- statistical studies, and case histories all reinforce the ance, it is not in "external" balance. It follows from a basic findings that exchange rate overvaluation can national income accounting identity that a trade reduce economic efficiency, misallocate resources, deficit means the country is spending more than its increase capital flight, and, most perniciously, lead income. That is, the trade deficit allows the country to exchange and trade controls. to consume or spend beyond its income (or beyond the value of what it is producing). When a country's expenditure does not equal its income, it is not in The Theory "internal" balance. These external and internal Theory suggests that there are many channels imbalances can severely impede country economic through which an overvalued exchange rate hurts performance, and it is these imbalances that coun- the economy and growth: tries suffering from external shocks often face. Although a nominal devaluation is designed to · It discriminates against exports. Since a signifi- correct the problem of external balance, it will also cant portion of the costs of production is paid in be important to ensure internal balance; otherwise, domestic currency, the overvalued exchange rate the trade deficit may not be corrected by the nomi- results in a reduction of exporters' incentives and nal devaluation. For many developing countries the ability to compete in foreign markets. This chokes trade deficit reflects the government's fiscal deficit, foreign exchange receipts and damages a coun- which is often financed by monetary expansion. try's ability to purchase the imports needed for The monetary expansion in turn leads to inflation. economic activity. In this environment the impact on the real exchange · Import-competing industries are faced with rate of a nominal devaluation is likely to be eroded increased pressure from foreign companies, by inflation, since high inflation tends to appreciate resulting in calls for protection against imports the real exchange rate, making elimination of the from industrial and agricultural lobbies. The trade deficit problematical. political pressures for protection eventually prove In general, monetary or fiscal policies will have to to be overwhelming, and governments yield to be combined with exchange rate policies to achieve lobbying and impose higher tariffs on imports. both internal and external balance simultaneously. This closes the economy to international compe- This is a special case of a more general principle of tition and reduces access to needed imported economics: multiple policy targets typically require inputs and technology. As a result, growth falls. multiple policy instruments. In this chapter, howev- Devaluation serves the dual purpose of uniformly er, we focus on the experience of countries that have protecting import-competing industries and limited the use of exchange rate adjustment as an increasing incentives for exporters. economic policy instrument. · Productivity advances are less rapid because the export sectors and the import-competing sectors, Problems with "Automatic" Adjustment where productivity advances are often fastest, are Mechanisms disadvantaged by an overvalued exchange rate (Cottani, Cavallo, and Khan 1990). Unless the central bank takes offsetting action, a · Overvaluation induces capital flight among trade deficit will result in a decline in the domestic domestic citizens anticipating a devaluation. As a money supply. Thus, one response to an overvalued 18 Exchange Rate Overvaluation and Trade Protection exchange rate is to hold the nominal exchange rate with similar results, Ghura and Grennes (1993) ana- fixed and assume that domestic prices and wages lyzed the relationship between the real exchange will fall and so help bring tradable goods prices back rate and macroeconomic performance in 33 Sub- to internationally competitive levels. This is the Saharan African countries between 1972 and 1987. "specie flow mechanism" described by David Hume They found that misalignment, or overvaluation, in the 18th century. The problem with this strategy was associated with lower levels of growth of real is that in most modern economies, prices and wages GDP per capita, lower levels of exports and imports, tend to be sufficiently inflexible downward that sus- lower levels of investment, and lower levels of sav- tained and substantial periods of unemployment ings, even when they corrected for other causes. must be endured if the strategy is to have a chance of succeeding. Most countries are unwilling to Case Studies of the Effects of Overvaluation endure these high costs. (See Sachs and Larraín 1999 for a further discussion.) For example, as is The economic histories of developing countries that described below, Chile endured a deep recession in followed a classic import-substituting industrializa- 1982­83 before it devalued in 1984, and the fran- tion strategy after World War II provide good illus- cophone African countries in the CFA zone experi- trations of the negative effects of an overvalued enced disastrous consequences from overvaluation; exchange rate combined with trade controls. Latin in some, the economic contractions were compara- America, more than any other region, followed this ble to the Great Depression in the United States. strategy, but it was not alone. We select illustrative The CFA zone experience also casts doubt on the episodes from Argentina, Chile, Uruguay, Turkey, claim that countries should avoid devaluation in and the CFA zone of Africa. order to retain international investors. The zone certainly had stable prices and exchange rates, but Argentina, Chile, and Uruguay its failure to solve the problems brought on by the overvalued real exchange rate substantially Argentina, Chile, and Uruguay all followed import- decreased its attractiveness to foreign investors. substituting industrialization policies that led to a Capital flight increased in anticipation of an even- bias against exports, extremely uneven rates of trade tual devaluation (Clément and others 1996). protection across sectors, and controlled financial systems. They also experienced recurrent balance of payments crises and slow growth (Corbo, de Melo, Cross-Country Economic Performance and Tybout 1986). By the early 1970s, all three had Cottani, Cavallo, and Khan (1990) investigated the accelerating inflation, bottlenecks in production, effects of real exchange rate misalignment and vari- slow export growth, and balance of payments diffi- ability on the economic performance of 24 develop- culties (Corbo and de Melo 1987). In response, they ing countries between 1960 and 1983. They found went through two phases of stabilization and that exchange rate misalignment was strongly relat- reform, one in the mid-1970s and the other during ed to low growth of per capita GDP. Misalignment 1979­82. The second phase is most relevant for was also related to low productivity (capital did not evaluating the effects of an overvalued exchange go to the companies or sectors that could make the rate and import controls on economic perform- best use of it), slow export growth, and slow agricul- ance. tural growth. In the second phase all three countries used a A study of growth in 12 countries between 1965 nominal exchange rate anchor to halt inflation. The and 1985 (Edwards 1989) reinforced these exchange rate appreciated, and when it became findings.4 The greater the misalignment, the lower apparent that the nominal rate could not be sus- the growth during the period. Furthermore, tained, capital flight resulted. In Uruguay and exchange controls and trade impediments, proxied Argentina, where there were no capital controls, by the black-market exchange rate premium, were major capital outflows occurred. In Chile, where negatively related to growth. there were capital controls, people engaged in capi- There is strong evidence that overvaluation of real tal flight by buying imported consumer durables. exchange rates was greatly implicated in Africa's This capital flight occurred in all three countries poor economic performance. Among other studies well before the onset of the debt crisis in 1982. 19 T R A D E P O L I C Y R E F O R M I N C O N T E X T Other problems resulted. Profitability fell in the and a staged lowering of uniform tariffs, from 35 tradable goods sectors. In Argentina, which percent in 1984 to 11 percent by 1991. An important remained quite restrictive to imports throughout, feature of the new nominal exchange rate system the gross margins of exporting businesses were hurt was a crawling band, which policymakers intended much more than those of import-competing busi- to use to maintain the international competitiveness nesses. In Uruguay the rate of growth of nontradi- of Chilean exports (Dornbusch and Edwards 1994). tional exports fell sharply between 1979 and 1981. In fact, although they used the nominal rate as the In Chile the leading growth sectors during the peri- policy variable, they focused on the real exchange od were construction, internal trade, and financial rate, adjusting the nominal rate for the differential services--all nontradables--even though reforms between domestic and foreign inflation. Taking an during the 1975­79 period had reduced the bias index of 100 as the value of the real rate in 1977, the against exports significantly by June 1979. real exchange rate appreciated to 84.5 in 1981, fell to 118.2 in 1984, and then, following the introduction of the new policy, depreciated to 145.2 in 1985. It Chile: The Aftermath continued depreciating, to 180.1 in 1990 (Corbo Chile is now well known for its economic success. and Fischer 1994). In 1998 the Chilean legislature Since 1984, it has had an average annual rate of approved a further lowering of the uniform tariff to growth of real GDP of more than 7 percent. Its poli- 6 percent, in stages, and in late 1999 Chile aban- cies following the 1982­83 crises are instructive. doned the exchange rate band system for a float. Chile experienced high rates of growth in the late The improved incentives to exporters from the 1970s, following a deep contraction in 1974­75. The reduction in the import tariff and the devaluation growth surge was the result of a number of deregula- led to an expansion of nontraditional exports (by 10 tion and reform measures, including institution of a percent a year from 1985 to 1995) and to efficient uniform 10 percent tariff on all goods except auto- import substitution. Macroeconomic stabilization, mobiles. Nevertheless, inflation persisted, hurting tax reform, and cuts in government spending com- the reforms, and in 1979 Chile fought back by set- bined to promote savings and investment. And pri- ting a fixed exchange rate as a nominal anchor. Com- vatization of state-owned firms, rehabilitation of bined with other policies, this at first led to large the financial sector through recapitalization, and external borrowings, most of which were at variable strengthened bank regulation spurred private busi- interest rates. In the early 1980s the external financ- ness activity. ing dried up as confidence in the sustainability of the exchange rate ebbed. Making matters worse, Chile Turkey experienced a deterioration in the terms of trade. Then foreign interest rates rose, further hurting the Three episodes from the post­World War II history of Chilean financial and business sectors. In 1982­83 Turkey, recounted in Krueger (1995), provide another Chile experienced its worst depression since the illustration of the problems created by an overvalued 1930s, as real GDP fell 15 percent. exchange rate combined with import restrictions. During and immediately after the recession, Chile Like the Latin American countries, Turkey followed experimented with a number of policies, including an import-substituting industrialization growth strat- an increase in tariff rates to switch domestic spend- egy. Starting in 1953, export growth ceased for a num- ing to domestic products. In June 1982 the govern- ber of reasons, and inflation accelerated. The ment abandoned the fixed exchange rate, combination of inflation and a fixed nominal eliminated compulsory wage indexation, and initi- exchange rate meant a strengthened real exchange ated a series of nominal devaluations. For a short rate and a bias against exports. Foreign exchange time, Chile allowed the exchange rate to float became scarce, and the country started import licens- (Corbo and Fischer 1994). Then, however, it fol- ing in 1954. By 1957, export earnings were falling, and lowed an erratic policy, implementing five different imports were severely restricted, damaging domestic exchange rate regimes (Labán and Larraín 1995). economic activity. In 1958, Turkey could not finance In 1985 the government embarked on the strategy imports, and it appeared that the country would not it maintains to this day: an export-oriented struc- even be able to obtain gasoline for trucks to move that tural adjustment. This included steady devaluations year's harvest to ports. 20 Exchange Rate Overvaluation and Trade Protection In response, Turkey adopted an IMF stabilization countries, average growth was 2.8 percent. In addi- plan that featured devaluation, import liberaliza- tion, CFA countries achieved an annual average tion, and fiscal and monetary restraint. Real GDP, export growth rate of 7 percent. which had been declining, started growing immedi- In the mid-1980s the economic performance of the ately in response to the availability of imports. Infla- CFA zone countries began to deteriorate, for two rea- tion fell, and export earnings began to rise again. In sons: the appreciation of the French franc, and a the 1960s Turkey was among the most rapidly grow- series of primary-commodity price shocks (Azam ing developing countries. and Devarajan 1997). Devarajan (1997), in a study of In the late 1960s, Turkey's exchange rate again 12 CFA countries, found an average overvaluation of became overvalued as a result of moderate inflation 31 percent in 1993 on the eve of the devaluation, with throughout the decade (5 to 10 percent annually) Cameroon's real exchange rate the most overvalued and a fixed nominal exchange rate. The high (78 percent) and Chad's real rate the only underval- demand for imports, together with the bias against ued one.5 Eight of the 12 had overvaluations of 20 exports, caused foreign exchange to become scarce. percent or more. Making matters worse, other The resulting problems in obtaining imports led to African countries were devaluing during the 1980s, a slowdown in both production and real invest- contributing to the overvaluation of the real rates of ment. The country responded in 1970 with a nomi- the CFA zone countries compared with those of their nal devaluation, and the result was extremely rapid export competitors. Elbadawi and Majd (1996) export growth. Turkey then experienced rapid eco- showed statistically that CFA membership and, by nomic growth through 1975. implication, the high level of the real exchange rate The third episode occurred in the late 1970s. were partly to blame for the poor economic perform- Large fiscal deficits, failure to adjust the internal ance of the CFA countries in the late 1980s. price of oil following the 1973 oil shock, and an Because of the overvaluations and mounting overvalued exchange rate, made worse by extremely structural problems, such as rigidly high wages, eco- high inflation, spurred this new crisis. Once again, nomic performance started to deteriorate. The zone the country ended up with severely constrained saw no economic growth between 1986 and 1994, a imports, falling real output, and falling income. period when other Sub-Saharan African countries were growing at 2.5 percent a year (Clément 1994). In fact, some of the countries suffered an output The CFA Zone Countries contraction comparable to that of the Great The currency of the countries of the CFA zone of Depression in the United States (Table 3.1). Africa was fixed precisely to the French franc and is A number of other ill effects stemmed from the now fixed to the euro. Until the second half of the period of overvaluation in the CFA zone. Several 1980s, these countries experienced stable and posi- countries suffered large increases in poverty tive economic performance (Elbadawi and Majd (Devarajan and Hinkle 1994). For example, in Côte 1996). For example, their average annual real GDP d'Ivoire the incidence of poverty doubled between growth rate between 1973 and 1981 was 5.7 percent, 1985 and 1992, from 30 to 60 percent. Devarajan whereas for 18 non-CFA Sub-Saharan African and Hinkle also note that banking systems in a Table 3.1 Comparing "Great Depressions": Cameroon, Côte d'Ivoire, and the United States (percentage decline in per capita GDP) Measure of output decline Cameroon Côte d'Ivoire United States Purchasing power paritya 31.4 29.1 -- Purchasing power parity with terms of trade adjustmenta 38.5 34.5 -- Market pricesb 41.5 18.8 30.9 -- Not available. a. Authors' calculations for 1986­92 from the Penn World Table Mark 5.6, described in Summers and Heston (1991) and available on the Website . Data for after 1992 are unavailable. b. Authors' calculations from the peak to the trough of the depression period (1986­94 for Cameroon and Côte d'Ivoire; 1929­33 for the United States). Data are from World Bank (1999) and U.S. Bureau of the Census (1975). 21 T R A D E P O L I C Y R E F O R M I N C O N T E X T number of countries became insolvent or illiquid as countries often apply high or prohibitive trade pro- a result of private sector inability to repay debts, tection on selected products or vis-à-vis selected government and public enterprise arrears, and capi- countries. Even given a limited objective of reducing tal flight. Export earnings collapsed in response to the demand for foreign exchange, an increase in the adverse terms of trade shocks and the overvalu- imports will occur through informal channels, ation of the real exchange rate. The contractionary depending on how porous the borders are. With macroeconomic policies adopted by most CFA diverse protection, while some sectors will be pro- countries reduced import levels, and inflation tected, the burden of the costs of adjustment to the remained low, but budgetary and external deficits overvalued exchange rate will be borne by the rose. The fixed nominal rate and various policy- unprotected sectors, by those sectors that are more induced rigidities in domestic prices, particularly in susceptible to informal or illegal imports, and by the wages and nontradable goods prices, meant that export sectors. Countries typically eventually deval- adjustment had to come through reduced employ- ue, but it is better that the devaluation be accom- ment, output, and growth.6 plished without debilitating losses in reserves and Constrained by their fixed exchange rates, at least lost productivity due to import controls. two of the CFA zone countries tried to carry out As the experience cited here shows, governments "mock devaluations," with subsidies to exports and must avoid policies that contribute to an overvalued increases in import tariff rates. In Côte d'Ivoire the exchange rate. Although we do not advocate any scheme collapsed after a short trial because of particular type of exchange rate regime in this chap- administrative difficulties, inability to give the ter, we emphasize that whatever regime is employed, export subsidy plan a sufficient budget, and lack of policies should be aimed at maintaining a competi- support by the government. In Senegal administra- tive real exchange rate. tion of the plan proved difficult, and the scheme encouraged overinvoicing by exporters and smug- Notes gling and underinvoicing by importers. The plan also proved costly to the budget, as tariffs were The authors thank Arup Banerji, Julian Berengaut, Dominique already high and the increases could not generate Desruelle, Lawrence Hinkle, Fred King, Kiyoshi Kodera, Albert Mar- tinez, Will Martin, Francis Ng, Paul Ross, Maurice Schiff, and semi- much more revenue. nar participants at the World Bank for helpful comments on earlier Finally, on January 12, 1994, the countries held a drafts of this paper. "maxi-devaluation," changing the rate to the French franc from 50:1 to 100:1.7 The CFA devaluation had 1 As of the beginning of 1999, the IMF (1999: app. I) reported arrangements for 185 countries. The exchange rate regimes excellent intermediate-term effects on growth. For can be categorized as pegged (84 countries), floating (75 the 12 CFA countries in Devarajan's sample, World countries), and limited flexibility (26 countries). Of the 84 Bank data showed that real GDP growth between countries with pegged exchange rates, 37 have no separate 1990 and 1993 averaged almost minus 0.3 percent legal tender, 8 use a currency board arrangement, 24 peg to annually, weighted by GDP (World Bank 1999). another currency, and 15 peg to a composite of currencies. Of those using a floating rate, 27 maintain a managed float and From 1994 to 1997, however, growth in the sample 48 an independent float. countries averaged 5.1 percent annually, according to the same data source.8 Cameroon, the largest 2 See Global Currency Report (1999). Of 160 countries listed, 38 had black market premiums of more than 10 percent at the country in the CFA zone, grew at an annual rate of end of 1998. Of the 38, 19 had premiums of more than 25 minus 3.4 percent in the first period but by 4.5 per- percent, 13 had premiums of more than 50 percent, and 10 cent in the second period (World Bank 1999). (Afghanistan, Algeria, Angola, Iraq, the Democratic Republic of Devarajan (1997) found that a year after the devalu- Korea, Liberia, Libya, Myanmar, São Tomé and Principe, and ation, the average undervaluation for the group was Somalia) had premiums of more than 100 percent. The black- 2 percent, but with significant variance. market exchange rate is likely to be overly depreciated in rela- tion to an equilibrium long-run real exchange rate, since an actual real depreciation would increase the supply of and Conclusion reduce the demand for foreign exchange. See Ghei and Kamin (1999) for a detailed explanation and econometric evidence. Worldwide experience has shown that defending 3 Ghei and Pritchett (1999) call this the "import compression syn- the exchange rate has no medium-run benefits. In a drome." Since devaluations (which reduce imports) are often classic pattern, once reserves are drawn down, accompanied by reductions of trade barriers (which increase 22 Exchange Rate Overvaluation and Trade Protection imports), econometric evidence on the import-reducing impact of devaluation has been weak. Ghei and Pritchett argue that devaluations significantly reduce imports if there is proper adjustment for the simultaneous reduction of trade protection. 4 The 12 countries studied were Brazil, Colombia, El Salvador, Greece, India, Israel, Malaysia, the Philippines, South Africa, Sri Lanka, Thailand, and Yugoslavia. 5 The countries in the study were Benin, Burkina Faso, Cameroon, Central African Republic, Chad, Congo, Côte d'Ivoire, Gabon, Mali, Niger, Senegal, and Togo. 6 For example, both Senegal and Côte d'Ivoire had rigid labor laws that kept wages high throughout the predevaluation peri- od (Foroutan,1997). Clément (1994) noted that throughout the CFA zone, rising wage costs contributed to substantial drops in public enterprise profitability, expanding the public sector financing requirement. Extensive controls over both producer prices and retail prices, particularly nontradable goods prices, added to the price rigidities in many countries. 7 The Western and Central African monetary unions (comprising Benin, Burkina Faso, Cameroon, Central African Republic, Chad, Congo, Côte d'Ivoire, Equatorial Guinea, Gabon, Mali, Niger, Senegal, and Togo) changed their rates from 50 CFA francs: 1 French franc to 100 CFA francs: 1 French franc. At the same time, Comoros changed its rate from 50 Comoros francs: 1 French franc to 75 Comoros francs: 1 French franc. 8 The unweighted averages are 0.1 percent for 1990­93 and 4.7 percent for 1994­95. 23 4 L I A M E B R I L L J A N E T S T O T S K Y R E I N T G R O P P Fiscal Dimensions the reforms undertaken and the initial circumstances of the par- ticular country. of Trade Table 4.1 shows a taxonomy of trade reform measures and their Liberalization expected effect on the govern- ment's fiscal position. As noted above, the impact can be positive, negative,or neutral,depending on the nature of the restrictions and the characteristics of the particu- lar country. For the most part, however, the reforms will general- D ly enhance revenue collections or espite substantial trade liberaliza- will have an ambiguous effect.The sequencing of trade tion over the past decade, many liberalization in programs supported by the IMF and developing countries continue to have restrictive the World Bank normally gives the highest priority to trade regimes characterized by high tariffs and per- the removal of nontariff barriers, which tend to be the vasive nontariff barriers (NTBs). Given the now most distortionary, followed by measures to rational- well-established nexus between open trade regimes ize the tariff structure. Below, we take a brief look at and improved export and growth performance, fur- some of these measures and their implications for the ther trade liberalization to promote sustainable country's fiscal position. growth and integration into the global trade system Nontariff barriers encompass a whole range of remains essential. Because trade liberalization has practices, including quotas, bans, export and import implications for fiscal revenues, and because many licensing, and state trading monopolies. In addition low-income countries continue to rely to a signifi- to the economic efficiency arguments, the initial focus cant extent on trade taxes as a source of revenue, on removing NTBs has the advantage of also increas- attention must center on the fiscal dimension of ing fiscal revenues. Quotas and bans provide no rev- trade reform in designing a strategy for trade liberal- enue to the budget and offer ample opportunities for ization. rent-seeking behavior or for smuggling. Accordingly, the conversion of quotas into equivalent tariffs (nor- mally accompanied by a scheduled timetable for fur- The Revenue Impact of Trade Liberalization ther reduction in tariff rates) or the removal of bans Trade liberalization has often been delayed by con- will, other things being the same, have an immediate cerns that it will have a negative impact on fiscal positive effect on fiscal revenues as rents are trans- revenues and contribute to macroeconomic insta- ferred to the government in the form of trade tax rev- bility. In fact, the impact of trade liberalization on enues. For these reasons, removal of NTBs should be revenues is generally ambiguous and depends on addressed early in the reform process. 24 Fiscal Dimensions of Trade Liberalization Table 4.1 Revenue Impact of Trade Liberalization Trade reform Expected revenue impact Replace nontariff barriers with tariffs Positive Eliminate tariff exemptions Positive Eliminate trade-related subsidies Positive Reduce tariff dispersion Ambiguous/positive Eliminate state trading monopolies Ambiguous/positive Reduce high average tariffs Ambiguous Lower maximum tariff Ambiguous Reduce moderate or low average tariffs Negative Eliminate export taxes Ambiguous/negative Sources: IMF and World Bank staff estimates. Eliminating tariff exemptions (excluding export structure, or one with few tariff bands, will mini- duty drawback schemes) and trade-related subsi- mize tax evasion and ease the task of customs dies should have a direct positive effect on the gov- administrators by reducing opportunities for mis- ernment's fiscal position. Moreover, not only does classification and valuation mistakes. the existence of tariff exemptions, especially discre- If a country has already implemented substantial tionary exemptions, provide an incentive for trade reforms, at some point further reductions in importers to seek additional exemptions, but their rates (other things being the same) will result in lower proliferation also increases the incentive for classify- revenue collection, at least in the short run. Given the ing taxable products as exempt, which has a nega- longer-term growth benefits of trade reform, howev- tive impact on revenues. Thus, in addition to their er, the appropriate response would be to offset any direct positive fiscal effect, eliminating discre- potential revenue loss by using other, less distorting, tionary exemptions and other complexities can broader-based taxes (for example, a value-added tax), contribute to improved governance. applied equally to both domestically and foreign-pro- The fiscal impact of reducing tariffs depends on duced goods. The distortion to the economy from their initial levels and coverage and on the extent to taxing both imports and domestic substitutes at which they are reduced. In principle, given an equivalent tax rates is generally less than that of taxing unchanged level of imports, lowering tariffs will imports alone, and taxing both yields larger revenues. reduce trade taxes. Since, however, the lower rates For developing countries in which trade taxes are are also likely to increase the demand for imports, an important source of revenue, a further reduction the net impact on revenues will depend on the price in the average tariff could be perceived to have a elasticity of import demand. The higher the elastic- negative effect on revenues and to inhibit the pace ity is, the more likely it is that a reduction in tariffs of further reforms. In these cases mobilizing alter- will have a net positive impact on fiscal revenues. native sources of revenue and diversifying tax In countries with prohibitively high tariffs, there sources away from trade taxes is critical but is likely is a strong incentive for tax evasion, either through to be a long-term process requiring an early start misclassification or by smuggling and avoiding pay- toward a broader-based tax. Since such a process ing the tax altogether. Therefore, lowering such tar- takes time to prepare and implement, technical iffs is likely to generate higher revenues because it assistance from the IMF should be sought at an reduces the cost of compliance and increases the early stage of the liberalization process and should volume of recorded trade subject to taxation as also be used to support the trade reform measures smuggling activities subside. More generally, a by improving customs administration. But even in reduction in tariff dispersion will tend to bolster countries that are highly dependent on trade taxes, revenues by reducing incentives for tax evasion. there is no reason to delay implementing trade The reform of the trade regime in the direction of reform measures that have a positive or neutral a more uniform tariff structure could increase fiscal impact on revenues. In fact, heavy reliance on such revenues as a result of increased transparency and taxes strengthens the case for proceeding more rap- simplification of tax administration. A uniform idly with the revenue-increasing elements of trade 25 T R A D E P O L I C Y R E F O R M I N C O N T E X T reform, especially the tariffication of NTBs and the dence suggests that liberalization of quantitative curtailment of exemptions. restrictions tended to bolster revenues and that tar- Significant progress has been made in reducing iff reforms did not result in revenue losses. the role of export taxes, in part because their elimi- · Fiscal considerations were the main factors cited nation is now generally accepted as a way of as limiting the extent of targeted trade reforms. enhancing growth prospects and strengthening a Greater trade liberalization could have been tar- country's external position. The effect on revenues geted and achieved if more attention had been from lowering export taxes will depend on how given to supportive fiscal policies and to revenue- much and how rapidly the reduction expands total neutral trade measures. trade and reduces illegal activities such as smug- · The effects of trade reform on revenues also gling. Since export taxes are often claimed to be depend significantly on the accompanying substitutes for some form of income tax on hard-to- macroeconomic policies and, in particular, on an tax sectors such as agriculture, their reduction or appropriate exchange rate policy. elimination is likely to be accepted by the country authorities if it is implemented as one element of an Lessons for the Design of Trade Policy overall tax reform package for broadening the tax base. Elimination of export taxes will also have a The discussion in this chapter suggests that there is positive impact on the producers of the commodi- scope for so tailoring the pattern of trade liberaliza- ties that are affected. In the case of agricultural tion as to avoid adverse consequences for revenues. products, these producers may be among the poorer Accordingly, adjustment programs supported by the segments of society (Box 4.1). IMF and the World Bank should, at the start of the Finally, in many programs supported by the IMF reform process, focus on broad-based trade liberal- or the World Bank, substantial trade liberalization ization measures, with a front-loading of those ele- has been accompanied by a devaluation of the ments that are likely to have a positive impact on exchange rate in order to, among other things, pro- revenues. The discussion also underscores the vide incentives for exporters so that they can take importance of sound macroeconomic policies and, advantage of the more liberal trade regime. In gen- in particular, the need for an appropriate exchange eral, the effect of devaluation on trade taxes is rate and for efforts to broaden the domestic tax base. ambiguous (Tanzi 1989) and will depend on the Nevertheless, the problems posed by trade liberal- price elasticity of import demand; if import ization in cases where it is perceived to have an ini- demand is inelastic, the devaluation will result in a tial negative effect on revenue (especially in higher value of imports in local currency terms and countries that rely significantly on trade taxes) will increase revenues at any given level of tariffs. should not be minimized. Even if alternative rev- enue and expenditure measures are readily avail- able, there are likely to be political and economic Case Studies and Other Empirical Evidence challenges. In addition, the consideration of fiscal Several studies (for example, Ebrill, Stotsky, and alternatives will take place in the context of pro- Gropp 1999; Sharer and others 1998) have exam- gram design, which usually involves fiscal pressures ined the actual impact on fiscal revenues following in many areas of both revenue and expenditures, the implementation of trade reforms. These studies including outlays in support of structural reform strengthen the observation noted above that the other than trade policy. sequencing of trade reforms can be done in a way Since sustained trade liberalization could eventu- that minimizes its adverse effect on revenues. Some ally lead to a reduction in the share of trade tax rev- of the main conclusions from these studies follow. enues in total receipts, maintaining revenue performance will require compensating domestic · For countries that initially started with highly tax reforms. Given the long gestation period of tax restrictive trade regimes, trade reforms were imple- policy and administrative reforms, it is critical that mented with a view toward protecting budgetary the reform of domestic taxes be considered at the revenues, and for the most part countries were able very outset of the trade reform exercise and that to achieve significant liberalization without com- technical assistance be sought at an early stage of promising their fiscal objectives. Empirical evi- the liberalization process. 26 Fiscal Dimensions of Trade Liberalization B O X 4 . 1 E X P O R T TA X E S Developing countries often impose export taxes on GI's interventions in the domestic market had a simi- primary commodity exports. Export taxation is one lar depressing effect on producer prices, which fluc- policy instrument that is not subject to WTO disci- tuated around US$5 per kilogram during the 1980s. plines, reflecting WTO members' focus on import Who benefited from the bourbon vanilla cartel policies. This makes it particularly important to and CAVAGI's domestic policies? Indonesian pro- determine the economic effects of such policies. In ducers were clearly the winners. The losers were some cases taxes are imposed in lieu of royalties for Madagascar's producers--mainly smallholders, the extraction of minerals; in others they are used numbering about 60,000, with an average pro- to provide protection to industries that process pri- duction of 130 kilograms and an average income mary commodities. In the latter case they can have of US$650 per plantation. adverse impacts on the poor that need to be care- A recent study provides estimates of the produc- fully monitored and analyzed. The export taxes er prices that would have prevailed in Madagascar mean that primary producers and farmers receive a had the marketing board been abolished. These price below that prevailing in world markets for are close to US$26, well above the US$5 price their commodities. Elimination of the tax will raise fixed by CAVAGI. Taking into account the increase their incomes but may bankrupt established pro- in production that such a change in prices would cessing facilities that are viable only if they pay have generated, laissez-faire policies would have lower-than-world prices for their inputs. Such increased the vanilla producer surplus eightfold. plants may employ poor urban labor, giving rise to Perhaps surprisingly, given the market power that a policy conundrum. In such cases a careful analysis Madagascar had in international markets, free of the appropriate trade regime for poverty allevia- trade (no intervention) would have also increased tion and the provision of safety nets is needed. Madagascar's welfare, by 0.5 percent of GDP--the Sometimes export taxes are used in an attempt to outcome of a major gain equivalent to 2.2 percent exercise market power, and in such cases the policy of GDP for producers, partially offset by a 1.7 per- can have a very adverse effect on the poor. An exam- cent of GDP loss to the marketing board. ple is Madagascar's marketing board for vanilla. An alternative to free trade would have been for In 1960 Madagascar, the world's lowest-cost CAVAGI to eliminate its interventions in the domestic producer of high-quality bourbon vanilla, account- market but to continue to exploit its market power in ed for 60 percent of world exports of natural vanil- international markets through an export tax. Esti- la. From its dominant position, Madagascar mates suggest that the optimal export tax would organized a bourbon vanilla cartel, with Comoros have been close to US$25 per kilogram instead of and Reunion, which set high export prices. Mada- the US$61 implicit tax that CAVAGI was imposing on gascar restricted supply by regulating its domestic producers. This would still have resulted in a dou- market through a marketing board (CAVAGI) that bling of the vanilla producers' surplus and, when fixed low producer prices and required licenses for combined with the tax revenues, would have gener- growing, preparing, and exporting vanilla. ated a welfare gain close to 1 percent of GDP. If this strategy were to be assessed by the effect it A likely explanation as to why these alternative poli- had on export prices of vanilla from Madagascar, it cies were not pursued is that the marketing board's was a clear success. The export price of vanilla revenue would have declined under both scenarios. increased from US$10 per kilogram in the late 1960s This suggests that Madagascar's marketing board to more than US$65 in the early 1990s. However, pricing policies had objectives other than welfare Madagascar's share of world markets declined to 30 maximization and that the heavy implicit taxation of percent as Indonesia, which was outside the cartel, small producers generated an important income took advantage of high world prices to develop its redistribution from the rural poor to the urban elite. export capacity. The entry of Indonesia into world markets left the total value of Madagascar's exports Source: Prepared by the volume editors, based on de Melo, constant throughout the 1970s and 1980s. CAVA- Olarreaga, and Takacs (2000). 27 5 L . A L A N W I N T E R S Trade Policies ed with an eye to its direct poverty consequences. It should, rather, be set on a sound basis for Poverty overall, with recognition that some modification may be Alleviation inevitable for political and other reasons. The primary way to deal with poverty is through general antipoverty policies. Trade Reform and Poverty International trade scholars have long understood that although H owever it is defined, poverty is not for small countries, trade interventions are generally a direct result of international inefficient and wasteful, their inefficiency is usually trade. Rather, poverty reflects low earn- dominated quantitatively by their redistributive ing power, few assets, poor access to communal effects. That is, the net losses from intervention will resources, poor health and education, powerless- generally represent large positive effects for some ness, and vulnerability. It does not matter what people and households and large negative effects for causes these features so long as they exist, nor what others. Correspondingly, although removing inter- relieves them if they can be relieved. Trade policy ventions will generally be income enhancing overall, matters only to the extent that (a) it affects the it is likely to generate both winners and losers.1 For direct determinants of poverty and (b) relative to example, liberalizing an import sector typically the whole range of other possible policies, it offers redistributes real income from producers to con- an efficient policy lever for poverty alleviation sumers as prices fall, and between different factors of (more poverty bang for a buck of forgone opportu- production in such a way that some gain while oth- nities). ers lose more than average. Trade liberalization may have adverse conse- The important positive issues are empirical: does quences for some--including some poor people-- trade liberalization generally create poverty, and that should be avoided or ameliorated to the under what circumstances might it do so in specific greatest extent possible. My fundamental belief, cases? In Winters (2000a), I develop a detailed tax- however, is that trade liberalization aids growth, onomy linking trade shocks to household and indi- which, in turn, aids poverty alleviation. I also vidual poverty and extract 11 key questions that believe that a widespread reform will contain help to answer the latter question (see Box 5.1). enough positive elements so that, in general, only a Finding an instance in which a trade reform causes few people will end up as net losers. Trade policy poverty may not constitute a refutation of an intel- should therefore generally not be closely manipulat- lectually interesting hypothesis, but it does pose a 28 Trade Policies for Poverty Alleviation B O X 5 . 1 K E Y Q U E S T I O N S F O R D E T E R M I N I N G T H E P O V E R T Y I M PA C T O F A T R A D E R E F O R M Will the effects of changed border prices be passed dicting either the price effects or the factor inten- through to the rest of the economy? Trade policy sities of affected sectors can be complex, as was and shocks operate primarily through prices. If seen with the Latin American reforms of the price changes are not transmitted--for example, 1980s and 1990s. In addition, if factor supplies because governments continue to fix the internal show some elasticity, part of a trade shock will prices of goods that they have ostensibly liberal- show up as changes in employment rather than ized internationally--the most direct effects on in factor prices. At the limit, a factor with a per- poverty, whether positive or negative, will be nul- fectly elastic supply will experience only employ- lified. ment effects. This is most pertinent for labor Is reform likely to give poor consumers access to markets. If the prevailing wage is determined by new goods? Perhaps the most direct effect of subsistence levels, switching people from one trade reform on poverty is through the prices of activity to another has no perceptible effect on goods and services in which poor households poverty. If, however, the trade-affected sector have large net positions. The biggest price shocks pays higher wages (because, say, it has an institu- occur when either the initial or final price is finite tionally enforced minimum wage), increases in and the other is infinite (that is, when there is no activity will tend to reduce poverty, and declines market). A shock that completely undermines an will tend to increase it. The formal-informal divide important market--say, for a cash crop or a form is important in this respect. In all this, it is impor- of labor--is likely to have major implications for tant to remember the difference between the poverty. Similarly, making new opportunities, functional and the personal distribution of goods, or services available to the poor can great- income. Falling wages for unskilled labor gener- ly enhance welfare. ate poverty only to the extent that the poor Is reform likely to affect different household mem- depend disproportionately on such wages. bers differently? Within a household, claims on Will the reform actually affect government rev- particular goods and endowments of particular enue strongly? One's immediate reaction is that assets (labor) are typically unevenly distributed. It cutting tariffs will reduce government revenue. is possible that poverty impacts will be concen- Although at the limit this is clearly true (zero tar- trated on particular members--usually females iffs yield zero revenue), many trade reforms actu- and children, who may lose personally even ally have small or even positive revenue effects, when the household gains in aggregate. especially if they convert nontariff barriers into Will spillovers be concentrated on areas and activ- tariffs, remove exemptions, and get tariff rates ities that are relevant to the poor? The sectors of an down to levels that significantly reduce smug- economy are interlinked, and, if substitutability is gling. Even where revenue falls, it is not inevitable high, a shock will be readily transmitted from one that expenditure on the poor will decline. That, sector to another. Frequently the diffusion will be ultimately, is a policy decision. so broad that it has little effect on any particular Will reform lead to discontinuous switches in activ- locality or sector, but sometimes--for example, ities? If so, will the new activities be riskier than the where trade in services is very local--the trans- old ones? If a trade liberalization allows people to mission is narrow but deep. Then it is necessary combine "national" and "international" activities, to ask whether the second-round effects have it is most likely to reduce risk: foreign markets are serious poverty implications. Agricultural stimuli likely to be less variable than domestic ones, and can confer strong propoor benefits on local even if they are not, risk spreading is likely to economies via benign spillovers. decrease overall risk. If, however, trade reform What factors are used intensively in the most leads to more or less complete changes in activi- affected sectors? Changes in the prices of goods ties, there is a possibility that risk will increase, if affect wages according to factor intensities. Pre- the new activity is riskier than the old one. (continued) 29 T R A D E P O L I C Y R E F O R M I N C O N T E X T B O X 5 . 1 ( C O N T I N U E D ) Does the reform depend on or affect the ability of Thus, if a reform implies very large shocks for par- poor people to take risks? The very poor cannot ticular localities, mitigation through phasing, or, bear risk easily. Because the consequences of better, through compensatory and complemen- even small negative shocks are so serious for the tary policies, could be called for. There is a trade- poor, they may be unwilling to take opportunities off, however, because, typically, larger shocks will that increase their average income if the chance reflect bigger shortfalls between current and of losses also increases. This might leave them potential performance and hence larger long-run with only the negative elements of a reform pack- gains from reform. age. Similarly, if a reform makes it more difficult Will transitional unemployment be concentrated for the poor to continue their traditional risk-cop- on the poor? The nonpoor typically have assets ing strategies, it may increase their vulnerability that carry them through periods of adjustment. to poverty even if it raises mean incomes. The situation might be unfortunate for them, but If the reform is broad and systemic, will any it is not poverty strictly defined. The poor have growth it stimulates be particularly unequalizing? few assets, so even relatively short periods of Economic growth is the key to sustained poverty transition could induce a descent deep into reduction. Only if it is very unequalizing will it poverty. If the transition impinges on the poor, increase absolute poverty. there is a strong case for using some of the long- Will the reform imply major shocks for particular run benefits of reform to ease their adjustment localities? Large shocks can create qualitatively strains. different responses from smaller ones; for exam- ple, markets can seize up or disappear altogether. Source: Winters (2000a). real policy question: should we automatically con- theoretical. For this and other reasons, losers will demn a trade reform because it means that one usually be better able to articulate their interests poor person loses or one person is pushed into than gainers, and so the volume of opinion is not a poverty? I believe very strongly that we should not. sufficient indicator of the relative strengths of the Rather, the identification of hardship arising from a pluses and minuses of a policy change. This is par- generally desirable policy reform should stimulate ticularly true given that the poor are generally much the search for complementary policies to minimize less able to advertise and defend their interests than the adverse consequences and reduce the hurt that are wealthier groups. they cause. Rejecting any reform that adversely In what follows, I explore three responses to the affects any poor person is a recipe for long-run stag- possibility that trade reform can create poverty: nation and for an ultimate increase in poverty. Even manipulating trade policy itself, compensating the the requirement that no household fall temporarily losers or the poor, and pursuing complementary into poverty is likely to be extremely restrictive in policies to try to ensure that as few people as possi- poor countries. The more utilitarian view that the ble are net losers. number of households (or persons) in poverty should not increase is more appropriate, although Can Trade Policy Be Managed to Alleviate even then, consideration of the depth of poverty is Poverty? required. All judgments ultimately must be quantitative, One natural response to the possibility that trade not just qualitative. In practical circumstances, it is liberalization could exacerbate poverty in certain easier to identify losers from trade policy than sections of a society is to "manage" liberalization in potential gainers. Losers are identifiable, concrete, a way that eliminates or at least reduces the prob- and personified (see Krueger 1990), whereas the lems. At the conceptual level, this is just common gains are diffuse and appear merely prospective and sense: poverty alleviation is arguably our highest 30 Trade Policies for Poverty Alleviation priority, whereas trade policy is just a means to an growth as a cause of growth) or include much more end. It makes sense to marshal all the tools we have than just trade policy--and that their econometrics toward achieving our principal goals, and indeed, it are weak. But Rodriguez and Rodrik do not argue would be perverse to do anything different. that trade liberalization is harmful to growth, nor But on a practical level, the question is how to use do they deal with other evidence for a beneficial trade policy to achieve poverty objectives. First, relationship, such as the manifest failure of closed there is the possibility that we do actually have goals economies and the findings of a number of case other than poverty alleviation--for example, as studies (see, for example, Srinivasan and Bhagwati regards average incomes, security, foreign policy, or 1999). Thus while Rodriguez and Rodrik should environmental sustainability--and these would certainly inspire greater modesty in policy advice need to be factored in. Second, even leaving aside and renewed research efforts, they have not (yet) additional objectives, we need to decide which reversed the presumption that openness is likely to measure of poverty we are aiming at: there are boost long-run growth. choices even among income- or consumption- The difficulty of establishing an empirical link based measures, let alone among the various con- between liberal trade and growth arises at least part- cepts and dimensions that characterize modern ly from two difficulties, both of which should views of poverty. Third, there may be questions inform our policy attitude. The first is the difficulty about trading poverty in one region against that in of measuring trade stances once one comes inside another, and there will certainly be, fourth, tradeoffs the boundary of near autarchy: tariffs need to be between poverty today and poverty tomorrow. aggregated, quantitative restrictions assessed and Fifth, what else figures in the policy packages among then aggregated, and the degrees of credibility, vul- which we are deciding? Are other policy instru- nerability to lobbying, and enforcement measured ments frozen at current levels, so that the question (see Winters 2000b). This suggests that while one is only one of how trade reform impinges directly should staunchly recommend openness, one needs on the real incomes of the poor? Or can we presume to be cautious about declaring particular regimes that other policies will be optimized, so that, say, open or not. Which was the more open in 1997, boosting incomes in the top decile at the expense of Brazil, or Chile? Both had average most-favored- higher prices for the poor is acceptable because it nation (MFN) tariffs of around 11­12 percent, but will permit a redistribution via the tax-benefit sys- in Chile there appeared to be little discretion and tem that more than offsets the initial growth in the little sensitivity to industrial lobbying, whereas in income gap? These questions illustrate that saying Brazil political pressures could be observed almost "manage trade policy" is not helpful until one spec- every day. ifies how to manage it. The second difficulty is that, although liberal trade policies are likely to be beneficial under any circumstances (because they enlarge the set of Don't Do It opportunities), a quasi-permanent effect on growth One response to the fear that a trade liberalization almost certainly requires combination with other will cause poverty is, "don't do it." But even if the good policies as well. The latter point is made direct effect of a reform might be to worsen poverty repeatedly by the IMF and the World Bank in their overall, this is not generally a satisfactory response. policy advice.2 Krueger (1990) has argued that Although it has proved difficult to isolate the effects openness is likely to be correlated with better policy of trade liberalization on economic growth empiri- in a number of dimensions, and supporting evi- cally, the predominant view is that it has an impor- dence for this assertion might be detected in Ades tant role. The well-publicized cross-country studies and Di Tella (1997, 1999), on corruption, and in that supported this view in the 1990s (for example, Romer (1993), on inflation. Thus, openness brings Dollar 1992; Sachs and Warner 1995; Edwards advantages not only on its own but also as part of a 1998) have recently received rough treatment from constellation of policies designed to ensure efficien- Rodriguez and Rodrik (2001). The latter argue, with cy and competition in markets, and transparency some justification, that these studies' measures of and predictability in policymaking. openness are flawed--in particular, because they The second part of the openness-poverty link either are endogenous (at least as much due to concerns the connection from growth to poverty. 31 T R A D E P O L I C Y R E F O R M I N C O N T E X T Growth needs to be strongly biased against the poor a series of exceptions is not strong. One needs very before it is likely to worsen poverty absolutely. (The compelling evidence of the efficacy of such inter- effect on inequality is a different story but, in my ventions, and such evidence is, on the whole, miss- view, a distracting one.) There are examples of such ing. Simply appealing to the experience of East Asia a negative relationship (see, for example, White and is not persuasive; it is not beyond dispute that these Anderson 2000), but they are equally balanced by countries' trade interventions were important or cases in which growth disproportionately favors the beneficial (Lee 1996 suggests the opposite for poor. Thus, Dollar and Kraay's (2001) finding that, Korea), and it is far from certain that other coun- on average, growth is good for the poor does seem tries have the policymaking institutions to be able to be robust (as it was in the earlier work of Gallup, to replicate East Asian policy stances effectively. Radelet, and Warner 1998), and so does their con- In addition to efficiency considerations, we must clusion that growth driven by trade liberalization is recognize that trade liberalization is a political act no different in that respect. A challenge to the latter and that governments must generate sufficient view can be found in the early work of Lundberg political support to sustain the reform. Even the and Squire (2000), but on further investigation they most rigorous reforms need tempering for political too concluded that trade liberalization benefited the reasons; see, for example, Edwards and Lederman poor, albeit by less than it helped better-off house- (1998) on Chile, where certain agricultural goods holds. were granted special protection in the form of price To conclude, I would argue that although there bands. Recognizing the need for such compromise remain a number of pressing research questions in is not the same as recommending it, however, and it the area, a liberal trade regime almost certainly is important to remember that the poor are even assists poverty alleviation in the long run. Thus, lib- weaker in political markets than in economic ones. eralization should have a place in the armory of a Rarely will protecting the poor and reaping signifi- poverty-conscious government. This does not cant political support for a reform coincide. Gov- imply a call for the immediate dismantling of all ernments are well advised to do everything possible trade restrictions, and it certainly does not imply to avoid using the instruments of trade policy for that opening the border is all that is needed, but it political purposes. One of the most powerful tools does, I believe, mandate a serious and credible com- for avoiding political pressures is uniformity-- mitment to openness in the foreseeable future. explicitly treating all commodities equally. Such a motivation was clearly articulated in Chile as it entered its big reform in the mid-1970s (Edwards Don't Do It All and Lederman 1998). A second response is,"don't do it all: while everyone There may be a stronger case for exceptions to lib- is in favor of liberalization in general, certain sectors eralization for the sake of direct poverty alleviation: or products should be exempt." In fact, all countries the outcome is objectively measurable and can have such exceptions (agriculture in Europe and arguably be isolated politically from general inter- clothing in the United States, for example), but that vention. Thus, if particular products can be clearly does not necessarily make them good economics. identified with the poor as either consumption or Considering overall economic performance, there production goods, it may be justified to postpone undoubtedly are cases in which an isolated inter- their liberalization significantly. There are, however, vention in trade would be beneficial to immediate some important caveats to such a recommendation. economic welfare--where externalities, informa- First, the calculation needs to be rigorous in defin- tion failures, or just random shocks can be usefully ing "the poor" whose interests are being protected. overcome by a well-judged intervention. But given (Ravallion and van de Walle 1991 show how in the difficulty of identifying these cases, of prevent- Indonesia the poor and the very poor had conflict- ing their capture by interest groups, and of avoiding ing interests in rice liberalization.) Second, the giving a systemic signal that lobbying for interven- products do need to be tightly linked to the poor in tion pays, it is not clear that it will be beneficial order that the distributional gains of protecting overall to pursue them. Thus, although one does not them are not offset by efficiency losses elsewhere in need to progress all the way to free trade to reap the the economy. The goods concerned need to be of benefits of liberalism, the general case for planning great significance to the poor--almost always, a 32 Trade Policies for Poverty Alleviation foodstuff on the consumption side, and frequently with long adjustment periods and postponing liber- an agricultural good on the production side--and alization because "the time is not ripe." The key is of little interest to other sections of society. On the credibility that reform will actually occur. Adjust- latter dimension, for example, it was reported in the ment costs may be lower if adjustment can be 1970s that in addition to any effects on the poor that spread somewhat through time, but they are proba- Egypt's bread subsidies may have had, they also bly increased if adjustment is resisted in the hope made it worthwhile for pig farmers to feed their that the threat of liberalization will go away. Several stock on fresh bread. trade reforms have been accelerated once they have Third, close monitoring is required to ensure that been launched; examples include implementation the desired effects do actually emerge. An important of free trade in the European Economic Communi- goal to keep in mind is that the poor continue to ty, of the Kennedy Round tariff cuts, and of the tar- have access to the effects of the policy and do not get iff cuts planned in the Association of Southeast hustled out of the way by more articulate and pow- Asian Nations (ASEAN) Free Trade Agreement. erful middle-class interests. In general, the poor fare Usually the acceleration happens at the behest of the very badly in discretionary allocation mechanisms private sector, presumably because, once it is precisely because they are marginalized politically accepted that reform will occur, business is keen to and socially, as well as economically. Thus, for adjust rapidly. example, export restrictions to keep down the price Thus, undertaking a major trade liberalization in of a local staple will probably not benefit the poor if phases is probably desirable, just as the Uruguay low prices mean that nonprice rationing is required. Round, for example, permitted long adjustment Fourth, a long-term plan is needed to help reduce periods. The phasing should, however, not merely the dependence of the poor on the policy interven- entail postponing the largest adjustments longest; tion. Otherwise, the intervention just amounts to it should pay attention to the different adjustment stopping the clock, which offers little prospect of needs of different sectors and to the interactions long-run development. between different parts of the package. For exam- Overall, "don't do it all" is not a suitable policy ple, if the inputs and outputs of a particular sector recommendation. The politics, especially protecting are liberalized at very different rates, the sector the interests of the poor, will be easier if the govern- could face either negative or excessively positive ment can explicitly reject special pleading on the incentives for production during the transition.3 grounds that everyone is receiving equal treatment. Whatever the transition period, credible commit- De facto, there may be--there may have to be-- ment to the final goal is important, for without it, some slippage in such an attempt, but it seems to neither current nor potential future activities will me undesirable to go into the process expecting or look desirable, and there will be a diversion of recommending slippage. The only exception I effort into lobbying. would make would be for temporary exemptions for goods or services that can be clearly and closely Compensatory Policies for Developing linked to the poor. A high burden of proof should Countries be placed on candidates for such exemptions to prove their efficacy. If trade liberalization causes poverty among certain sections of society, the next question is whether soci- ety can offset the effect directly. Despite the theoreti- Don't Do It Now cal attractions of lump-sum budgetary transfers for "Don't do it now" is a more useful response than economists, governments are not generally attracted the others in some circumstances. For example, to them because of their cost, their transparency trade reform in the midst of recession seems likely (and the transparency of their abuse), and the to give rise to more, and more durable, transitional appearance that they do little to cure "the problems" unemployment than reform in a boom. Again, that individuals face. Rather, assistance is usually where investment is necessary to allow the produc- offered, if at all, in terms such as retraining, reloca- tion of export-quality goods, it may be desirable to tion assistance, and temporary income support. In allow time for that to occur. There is, however, a fact, while these approaches probably do have a con- world of difference between committing to policies tribution to make, even they face severe difficulties. 33 T R A D E P O L I C Y R E F O R M I N C O N T E X T Official retraining has mixed success under any cir- thresholds are set low enough; and, since relieving cumstances, and, what is worse, it is difficult to sepa- poverty is more or less universally recognized as a rate those cases where trade is to blame from those responsibility of the state, there is little argument where it is not. Unless one is willing to underwrite about the legitimacy of such interventions. almost any adjustment, identification of cases is a Targeting is a major problem for safety nets, not major difficulty. Making a general commitment, only technically but also because the middle classes however, is not attractive because of the potentially are often better able to access them than the poor. huge cost and because doing so shifts private risk to Sustainability is another difficulty; a major trade the public sector, with all the attendant moral hazard shock could put severe financial pressure on a problems. It is not the role of the state, nor is it feasi- scheme just at a time when it is most needed. Raval- ble, to absorb every negative shock that might afflict lion (1999) offers some useful thoughts on setting individuals. Yet it is difficult to make a moral case as up safety nets. Workfare is a good start, provided to why trade shocks warrant adjustment assistance that the wage is low enough, that there is little or no while other shocks do not.4 administrative discretion in its application, and that A further complication arises in giving compen- the tasks set are seen to be of communal interest. In sation in a way that encourages rather than discour- fact, Ravallion suggests that local communities ages adjustment. European agricultural policy is select the projects to be undertaken under workfare essentially designed to protect farmers from the and that better-off communities should be asked to consequences of declining comparative advantage, cofinance the projects. Workfare has to be supple- yet it has the effect of rewarding current, not for- mented, however, by schemes to provide food to mer, farmers. Compensation may be decoupled people such as the elderly and infirm who cannot from current output but not from farming as an work and to children (through, for example, food- activity. for-education schemes). These supplementary In cases where liberalization leads to the loss of schemes may be tripped on and off according to jobs, government can insist on, and perhaps help need, but they should have a permanent infrastruc- finance, redundancy payments. These payments can ture and sensitive and quick triggers. Expenditure help some people avoid poverty, if they use their on safety nets is almost by definition countercycli- money productively, but they are not guaranteed to cal, and so a firm commitment by government is do so. (See Winters 2000a on the "new poor" in required to ensure that the money does not dry up Zimbabwe.)5 Moreover, redundancy payments typi- in times of greatest need. cally reward past service, not current need, and so Examples of useful safety nets can be found in they are not particularly well targeted for poverty Bangladesh. According to the Consumer Unity & alleviation purposes. Trust Society, General compensatory policies, including safety nets, are designed to alleviate poverty from any It is generally recognized that programs such as source directly. They replace the problem of identi- Food for Education (FFE), Vulnerable Group fying the shock with the task of identifying the poor. Development (VGD), Test-Relief, and Food for Ideally, countries should already have such pro- Work positively induce alleviation of poverty. grams in place. Indeed, a major part of the effect of For example, during the unprecedented floods these programs arises from their mere existence of 1998, about 4.5 million VGD cards were dis- rather than their use: they facilitate adjustment by tributed in Bangladesh, which provided crucial assuring the poor that there is a minimum (albeit a help at a critical time. The FFE program has barely acceptable one) below which they will not be helped increase school attendance of poor chil- allowed to fall. Such schemes, if trade-adjusting dren by 21%. (CUTS 1999: 110) countries do already have them, offer the advan- tages over tailor-made schemes of automaticity, The safety nets in Zambia and Zimbabwe, by con- immediacy, and a degree of "road-testing," and they trast, are currently regarded as too poorly run and also avoid the problems associated with targeted underfunded to be able to offer serious assistance to trade adjustment assistance. If they are sensibly losers from trade liberalization. constructed, they need not entail huge expenditure; Safety nets are not the only answer to the threat of there is little chance of moral hazard problems if the increasing poverty from trade liberalization, but 34 Trade Policies for Poverty Alleviation they are an important part of the response. They relatively underdeveloped in most of the smallhold- can generally be targeted better than other policies, er areas, increasing numbers of resettled and com- and they are not very distortionary of market forces. munal households are now becoming involved as If countries do not have safety nets already, they producers of the main crops. This has primarily should consider setting them up as part of the con- been the result of "outgrower" schemes and of text for a trade liberalization that may create short- sourcing or subcontracting by large-scale commer- term poverty. The safety nets should not, however, cial farms. The Horticultural Promotion Council be trade shock­specific. (HPC) estimates that around 3,000 small-scale farmers are now growing for export on a contract basis, accounting for approximately 10 percent of Complementary Policies for Zimbabwe's exports.6 In January 1999 the HPC Better-Functioning Markets established the Small-Scale Linkage Programme, A critical issue concerning the poverty impacts of designed to provide communal and resettled farm- trade liberalization, especially for surprises connect- ers with the knowledge and skills to produce high- ed with it, is the functioning of markets. Trade liber- value, off-season export crops. alization must be accompanied by monitoring to determine whether any markets are failing. Policies Credit Markets designed to ensure that markets continue to func- tion or to develop, where required, will have high Development economics affords many examples of payoffs for both aggregate income and poverty alle- how missing credit markets have prevented devel- viation. Some important circumstances are dis- opment, and the same phenomenon is visible in cussed next. responses to trade liberalization. Thus, for example, achieving minimum consignment size might entail hiring draft power or seasonal labor, but this is not Infrastructure Support possible without credit. Similarly, establishing Potential opportunities for poor producers to bene- informal businesses in areas such as trading may fit from a more open trading regime have been lost require more capital than the poor can raise. These because critical infrastructure was either absent or cases in which the poor are not able to respond to had deteriorated. In both Zimbabwe and Zambia incentives as strongly as the less poor replicate the remote farmers found their opportunities con- results of López, Nash, and Stanton (1995) in their strained by inability to reach major market centers. panel study of Mexican agriculture. In the same way, many of the benefits from relaxed retailing regulations and from availability of new or Labor Mobility cheaper goods have been confined to urban and periurban areas. The secret of spreading the benefits of increasing labor demand widely is labor mobility. If markets are segmented for cultural or geographic reasons, Market Institutions breaking down these barriers through information The poor frequently seem unable to attain the eco- and facilitating physical mobility will have an equal- nomic mass required for the establishment of mar- izing effect. kets that, once in place, may be viable. Policy should aim at the creation of the market as an institution, Establishing Businesses not at the ongoing subsidization of market activity. One aspect of facilitating the participation of the If the regulations for establishing new businesses are poor in markets may be to find means to allow them cumbersome, if the businesses' ability to obtain to combine very small consignments of inputs or inputs (especially utilities) is weak, or if regulations outputs into reasonably sized bundles and so reduce on expansion and on labor recruitment and separa- transactions cost sufficiently to make dealing with tion are restrictive, this could curtail the willingness poor producers worthwhile. Horticulture in Zim- of entrepreneurs to start or expand operations. A babwe offers an illustration of a successful policy of success story of business deregulation is the growth this kind (Winters 2000a). Although horticulture is of maize hammer milling in Zimbabwe. Following 35 T R A D E P O L I C Y R E F O R M I N C O N T E X T domestic deregulation, 3,500 new hammer mills This is not, however, a license to postpone the opened, mainly in rural areas, and the share of ham- design, announcement, and locking in of the reform mer millers in total maize milling has increased to itself. Any of these delays--for example, announc- almost 80 percent.7 These mills are mechanically ing that liberalization is necessary but that its form simple and robust (being based on swinging or will be worked out once certain other reforms have rotating hammers in a grinding chamber) and can been implemented--would seem likely to result in be used by unskilled labor. They provide quality the worst of all worlds. It would create uncertainty maize meal products to nearby customers in poor and incentives to lobby government and, indeed, communities, saving them significant transport would look to many commentators like a de facto costs. In 1995 hammer mills were estimated to reluctance to liberalize trade. In particular, in the employ 7,512 permanent workers (751 in urban absence of a clear and monitorable plan for specific areas); when casual workers and rural activities are pieces of infrastructure, a general wish to wait until included, the sector employs some 13,000 workers. the roads or ports are "ready" is just a recipe for About 18 percent of the employees in urban ham- indefinite postponement. A credible plan for liber- mer mills are female, as are 8 percent of the employ- alizing the borders--albeit one with significant ees in rural areas.8 transition periods--will be an important stimulus to reforming these other areas in ways that will typ- ically have other benefits as well. Prerequisites or Concomitants? It is also well to record that there are disadvan- In many cases actual policy debate appears to hinge tages as well as advantages to phased adjustment. on whether complementary policies of the sort just Populations can certainly suffer from reform fatigue described should be prerequisites for a trade liberal- and would actually be more comfortable with a def- ization."Everyone accepts that trade liberalization is inite, even if ambitious, reform plan than with one desirable in the long run," the argument goes, "but that drifts into the indefinite future. Phased adjust- various supporting policies must be in place before ment implies a longer time spent out of equilibri- it is attempted." Here, even more than in the matters um, and in most discussions it is not proved that the discussed above, we have no formal analysis to fall integral of shallow adjustment costs over a long back on. There is a literature on sequencing reform period is smaller than that of deep costs over a within the trade sector and between trade and capi- shorter period.9 Moreover, delay postpones the ben- tal accounts, but there are no convincing empirical efits of full reform. Finally there are likely to be generalizations about sequencing in the sense dis- aggregate gains from trade reform even in the cussed here. Moreover, the question is only partly absence of complementary policies. A trade reform economic; part of it is political and concerns increases opportunities for desirable exchange, and whether a reform postponed is a reform preempted. these will exist even with poor infrastructure, and I argued above that there may well be a case for even though there would have been more opportu- phasing in a reform over a long period provided nities had the infrastructure been better. And this that the final destination is clear (and not likely to applies to the poor as much as to other people. It is be contested) and that the transition is well possible that in the absence of complementary poli- designed and does not amount just to postponing cies, the poor will suffer (say, because of rising all effective change until the last moment. Given prices) whereas with such policies they will gain that a well-conceived and well-executed reform because they will receive income gains to offset the generates a potentially infinite-lived stream of bene- price rises. But there is no general theorem to this fits, whether this occurs over three years or, say, nine effect; the case remains to be made. years is not that important. The same logic applies to delays required to put complementary policies in Conclusion place (or, indeed, to compensatory mechanisms if that is the route chosen). Thus, for example, there Trade reform almost invariably brings with it two may be a case for delaying the implementation of a changes that help in the battle against poverty: it liberalization while legislation on business forma- induces efficiency in the use and allocation of tion or labor market operation is put in place and resources (the economist's beloved static gains), and plans for protecting market institutions are laid. it fosters long-run growth. It also entails temporary 36 Trade Policies for Poverty Alleviation adjustment costs that reduce incomes immediately, Notes although these costs are almost always outweighed Preparation of this chapter was supported by Globkom, the Parlia- by the long-term benefits. Finally it has a host of mentary Commission on Swedish Policy for Global Development, direct and indirect effects on poverty that could go to which I am most grateful. I also thank Costas Michalopoulos for either way, depending on consumption and produc- comments on the outline and Rosie Bellinger for logistic help. tion patterns and on the nature of reform. 1 I say "generally" because second-best considerations or market The general presumption is that reform will help failures could reverse the result. alleviate poverty, but the direct and indirect effects 2 Mosley (2000) argues that the attempts of the IMF and the just mentioned make it likely that some will lose World Bank to prove this proposition have not been very suc- from liberalization--especially one that is narrowly cessful. (His attempts to prove the contrary are similarly focused sectorally--and it is certainly possible that flawed, however.) some of the poor will suffer. Still, others will gain, 3 Technically, the effective rate of protection (ERP) could and these will quite possibly include others among become very distorted. In Zambia, Oxfam and the Institute for the poor. Thus, tradeoffs are necessary. There is no Development Studies (IDS) found decreases in the ERP for alternative to case-by-case analysis if policymakers maize, as fertilizer prices (which were said to account for 76 wish to predict and preempt adverse effects, even percent of the cost of production) increased by more than out- put prices. This accounts for the loss of output there. though prediction is very difficult. One must be alive to the possibility that "predict and preempt" 4 See Decker and Corson (1995) on the U.S. Trade Adjustment policies will be captured by powerful interest Assistance Program, which doubles the length of unemploy- groups. Given these groups' strong interests in trade ment insurance coverage, from 26 to 52 weeks, for workers certified as displaced by trade liberalization. After serious abuse policy and the apparent ease with which trade poli- in its early years, when it was merely a transfer (over 70 per- cy can be captured (because its domestic costs are cent of claimants went back to work for the employer from usually hidden and the issues can be so easily pre- whom they were said to have been displaced), a training ele- sented in terms of standing up to foreigners), there ment was added. This had the effect of screening out are grave dangers in setting out to manipulate trade claimants who did not want or need training, but it apparently did nothing to increase the earning power of recipients. policy directly to avoid adverse poverty impacts. Only in the most obvious cases are the dangers like- 5 The "new poor" are retired public sector officials who have not ly to be worth incurring. managed to invest their redundancy payments sufficiently pro- ductively to maintain themselves above poverty levels. My general prescription, then, is for a rigorously liberal trade policy (even though it is recognized 6 These small-scale "outgrowers" tend to supply the four main that some slippage may occur for political reasons). packinghouses in Zimbabwe, which are the large-scale pro- General compensatory policies should then be used ducers that seek to add volume and diversify risk. to cure immediate hardship, and complementary 7 The 1995­96 Zimbabwe National Hammer Miller Status Study, policies should be pursued to enlarge long-term funded by the U.S. Agency for International Development. gains. Assessing likely impacts in the design of poli- 8 Ibid. cy reforms is of great importance. The set of ques- 9 This is not to deny the possibility--as, for example, if a major tions posed in Box 5.1 can help policymakers in the shock creates hysteresis in labor markets--but it needs to be design and implementation of reforms. proved. 37 II THE WORLD TRADE ORGANIZATION T he WTO was created in 1995 as one of discussions of the "engine" of the WTO--the princi- the outcomes of the Uruguay Round of ple of reciprocity (Chapter 7, by J. Michael Finger multilateral trade talks. The Uruguay and L. Alan Winters); the accession process (Chapter Round, which concluded in 1994 after eight years of 8, by Constantine Michalopoulos); and the dispute complex and sometimes contentious negotiations, settlement mechanism (Chapters 9, by Valentina was a landmark in the history of the trading system. Delich, and 10, by Robert E. Hudec). The last is the Agriculture and textiles and clothing became subject aspect of the WTO that attracts most attention. The to stronger multilateral disciplines, and the trading WTO is unique among international organizations system was extended to include intellectual property in that it has a well-functioning, binding dispute set- and trade in services. The WTO establishes the rules tlement mechanism. This is of great importance to of the trade policy game for its members, which developing countries, which generally will not be increasingly include developing countries. (Member- able to induce compliance with negotiated rules in ship at the time of writing stood at 144, but more bilateral disputes with large industrial economies. In than 50 developing countries have yet to join the practice, because countries value the trading sys- WTO.) A good understanding of how the WTO tem, the large and powerful tend to abide by the works and what it does is a necessary condition for rulings of dispute settlement panels, providing an maximizing the benefits of membership. incentive for developing countries to ensure that The chapters in this part discuss some of the they are able to use the system. major features of the WTO that are relevant to The ability to use the WTO system is a function of developing countries. A brief summary of the basic many factors. Among the necessary conditions are rules and the institutional mechanisms of the WTO that countries participate in the negotiations on the (Chapter 6, by Bernard Hoekman) is followed by rules of the game and that they use the WTO in a 39 T H E W O R L D T R A D E O R G A N I Z AT I O N proactive manner. Much of this Handbook is aimed Kluwer Law International, 1999). An early study of at helping countries do so. the GATT system that remains well worth reading is The chapters in Part II are not intended to provide Gerard Curzon, Multilateral Trade Diplomacy (Lon- comprehensive coverage of the WTO. Those seek- don: Michael Joseph, 1965). Robert Hudec's seminal ing an in-depth treatment of the WTO, its negotiat- Developing Countries in the GATT Legal System (Lon- ing history, and its dispute settlement case law are don: Trade Policy Research Centre, 1987) is an indis- referred to the sources listed below. pensable source for those seeking to understand the approach taken toward development issues in the WTO. An informative history of the Uruguay Round Further Reading and Sources of negotiations is presented in John Croome, Reshaping Information the Trading System (Deventer: Kluwer, 1999). The WTO Website, , provides direct Michael Trebilcock and Robert Howse, in The Regula- access to most of the documents submitted to the tion of International Trade (London: Routledge, institution, as well as to reports and case law. The 1998), provide a comprehensive treatment of WTO International Center for Trade and Sustainable Devel- rules, as well as a comparison between WTO disci- opment has links to all the major nongovernmental plines and those that apply in the European Union organizations on its Website, , and and the North American Free Trade Agreement it publishes an informative newsletter, Bridges, that (NAFTA). A recent analysis and description of the monitors WTO issues from a development perspec- economics and politics of the world trading system tive. WTO dispute settlement procedures are dis- can be found in Bernard Hoekman and Michel cussed in detail by David Palmeter and Petros C. Kostecki, The Political Economy of the World Trading Mavroidis in Dispute Settlement in the World Trade System: The WTO and Beyond, 2d ed. (New York: Organization: Practice and Procedure (The Hague: Oxford University Press, 2001). 40 6 B E R N A R D H O E K M A N The WTO: Since 1947, the GATT has been the major focal point for industrial country governments Functions and seeking to lower trade barriers. Although the GATT was initially Basic Principles largely limited to a tariff agree- ment, over time, as average tariff levels fell, it increasingly came to concentrate on nontariff trade policies and domestic policies having an impact on trade. (See the Glossary to this volume for a list of trade-related policies used T by countries.) Its success was he WTO, established in 1995, adminis- reflected in a steady expansion in the number of ters the trade agreements negotiated contracting parties. By the end of the Uruguay by its members, in particular the General Agree- Round (1994), 128 countries had joined the GATT. ment on Tariffs and Trade (GATT), the General Since the entry into force of the WTO, membership Agreement on Trade in Services (GATS), and the has grown to 144, as of the end of 2001. Trade-Related Aspects of Intellectual Property The WTO differs in a number of important Rights (TRIPS) agreement. (These and other major respects from the GATT. The GATT was a rather WTO agreements are contained in the CD-ROM flexible institution; bargaining and deal-making lay "Applied Trade Policy," which is included with this at its core, with significant opportunities for coun- Handbook.) The WTO builds on the organizational tries to "opt out" of specific disciplines. In contrast, structure that had developed under GATT auspices WTO rules apply to all members, who are subject to as of the early 1990s. binding dispute settlement procedures. This is The origins of the GATT were in the abortive attractive to groups seeking to introduce multilater- negotiations to create an International Trade Orga- al disciplines on a variety of subjects, ranging from nization (ITO) following World War II. Negotiations the environment and labor standards to competi- on the charter of such an organization were con- tion and investment policies to animal rights. But it cluded successfully in Havana in 1948, but the talks is a source of concern to groups that perceive the did not lead to the establishment of the ITO because (proposed) multilateral rules to be inappropriate or the U.S. Congress was expected to refuse to ratify the worry that the adoption of specific rules may affect agreement. Meanwhile, the GATT was negotiated in detrimentally the ability of governments to regulate 1947 by 23 countries--12 industrial and 11 develop- domestic activities and deal with market failures. ing--before the ITO negotiations were concluded.1 The main function of the WTO is as a forum for As the ITO never came into being, the GATT was the international cooperation on trade-related poli- only concrete result of the negotiations. cies--the creation of codes of conduct for member 41 T H E W O R L D T R A D E O R G A N I Z AT I O N governments. These codes emerge from the small number of contracting parties to the GATT exchange of trade policy commitments in periodic (only 23 countries), the benchmark for MFN is the negotiations. The WTO can be seen as a market in best treatment offered to any country, including the sense that countries come together to exchange countries that are not members of the GATT. market access commitments on a reciprocal basis. It National treatment requires that foreign goods, is, in fact, a barter market. In contrast to the markets once they have satisfied whatever border measures one finds in city squares, countries do not have are applied, be treated no less favorably, in terms of access to a medium of exchange: they do not have internal (indirect) taxation than like or directly money with which to buy, and against which to sell, competitive domestically produced goods (Art. III, trade policies. Instead they have to exchange apples GATT). That is, goods of foreign origin circulating for oranges: for example, tariff reductions on iron in the country must be subject to taxes, charges, and for foreign market access commitments regarding regulations that are "no less favorable" than those cloth. This makes the trade policy market less effi- that apply to similar goods of domestic origin. cient than one in which money can be used, and it is The MFN rule applies unconditionally. Although one of the reasons that WTO negotiations can be a exceptions are made for the formation of free trade tortuous process. One result of the market exchange areas or customs unions and for preferential treat- is the development of codes of conduct. The WTO ment of developing countries, MFN is a basic pillar contains a set of specific legal obligations regulating of the WTO. One reason for this is economic: if pol- trade policies of member states, and these are icy does not discriminate between foreign suppliers, embodied in the GATT, the GATS, and the TRIPS importers and consumers will have an incentive to agreement. use the lowest-cost foreign supplier. MFN also pro- vides smaller countries with a guarantee that larger countries will not exploit their market power by Basic Principles raising tariffs against them in periods when times The WTO establishes a framework for trade poli- are bad and domestic industries are clamoring for cies; it does not define or specify outcomes. That is, protection or, alternatively, give specific countries it is concerned with setting the rules of the trade preferential treatment for foreign policy reasons. policy game, not with the results of the game. Five MFN helps enforce multilateral rules by raising principles are of particular importance in under- the costs to a country of defecting from the trade standing both the pre-1994 GATT and the WTO: regime to which it committed itself in an earlier nondiscrimination, reciprocity, enforceable com- multilateral trade negotiation. If the country desires mitments, transparency, and safety valves. to raise trade barriers, it must apply the changed regime to all WTO members. This increases the political cost of backsliding on trade policy because Nondiscrimination importers will object. Finally, MFN reduces negoti- Nondiscrimination has two major components: the ating costs: once a negotiation has been concluded most-favored-nation (MFN) rule, and the national with a country, the results extend to all. Other coun- treatment principle. Both are embedded in the main tries do not need to negotiate to obtain similar WTO rules on goods, services, and intellectual treatment; instead, negotiations can be limited to property, but their precise scope and nature differ principal suppliers. across these three areas. This is especially true of the National treatment ensures that liberalization national treatment principle, which is a specific, not commitments are not offset through the imposition a general commitment when it comes to services. of domestic taxes and similar measures. The The MFN rule requires that a product made in requirement that foreign products be treated no less one member country be treated no less favorably favorably than competing domestically produced than a "like" (very similar) good that originates in products gives foreign suppliers greater certainty any other country. Thus, if the best treatment grant- regarding the regulatory environment in which they ed a trading partner supplying a specific product is must operate. The national treatment principle has a 5 percent tariff, this rate must be applied immedi- often been invoked in dispute settlement cases ately and unconditionally to imports of this good brought to the GATT. It is a very wide-ranging rule: originating in all WTO members. In view of the the obligation applies whether or not a specific tar- 42 The WTO: Functions and Basic Principles iff commitment was made, and it covers taxes and establish "ceiling bindings": the member concerned other policies, which must be applied in a nondis- cannot raise tariffs above bound levels without criminatory fashion to like domestic and foreign negotiating compensation with the principal sup- products. It is also irrelevant whether a policy hurts pliers of the products concerned. The MFN rule an exporter. What matters is the existence of dis- then ensures that such compensation--usually, crimination, not its effects. reductions in other tariffs--extends to all WTO members, raising the cost of reneging. Once tariff commitments are bound, it is impor- Reciprocity tant that there be no resort to other, nontariff, Reciprocity is a fundamental element of the negoti- measures that have the effect of nullifying or ating process. It reflects both a desire to limit the impairing the value of the tariff concession. A num- scope for free-riding that may arise because of the ber of GATT articles attempt to ensure that this MFN rule and a desire to obtain "payment" for does not occur. They include Article VII (customs trade liberalization in the form of better access to valuation), Article XI, which prohibits quantitative foreign markets. As discussed by Finger and Winters restrictions on imports and exports, and the Agree- in Chapter 7 of this volume, a rationale for reciproc- ment on Subsidies and Countervailing Measures, ity can be found in the political-economy literature. which outlaws export subsidies for manufactures The costs of liberalization generally are concentrat- and allows for the countervailing of production ed in specific industries, which often will be well subsidies on imports that materially injure domes- organized and opposed to reductions in protection. tic competitors (see Chapter 17, by Pangestu, in this Benefits, although in the aggregate usually greater volume). than costs, accrue to a much larger set of agents, If a country perceives that actions taken by anoth- who thus do not have a great individual incentive to er government have the effect of nullifying or organize themselves politically. In such a setting, impairing negotiated market access commitments being able to point to reciprocal, sector-specific or the disciplines of the WTO, it may bring this situ- export gains may help to sell the liberalization polit- ation to the attention of the government involved ically. Obtaining a reduction in foreign import bar- and ask that the policy be changed. If satisfaction is riers as a quid pro quo for a reduction in domestic not obtained, the complaining country may invoke trade restrictions gives specific export-oriented WTO dispute settlement procedures, which involve domestic interests that will gain from liberalization the establishment of panels of impartial experts an incentive to support it in domestic political mar- charged with determining whether a contested kets. A related point is that for a nation to negotiate, measure violates the WTO. Because the WTO is an it is necessary that the gain from doing so be greater intergovernmental agreement, private parties do than the gain available from unilateral liberaliza- not have legal standing before the WTO's dispute tion. Reciprocal concessions ensure that such gains settlement body; only governments have the right to will materialize. bring cases. The existence of dispute settlement pro- cedures precludes the use of unilateral retaliation. For small countries, in particular, recourse to a mul- Binding and Enforceable Commitments tilateral body is vital, as unilateral actions would be Liberalization commitments and agreements to ineffective and thus would not be credible. More abide by certain rules of the game have little value if generally, small countries have a great stake in a they cannot be enforced. The nondiscrimination rule-based international system, which reduces the principle, embodied in Articles I (on MFN) and III likelihood of being confronted with bilateral pres- (on national treatment) of the GATT, is important sure from large trading powers to change policies in ensuring that market access commitments are that are not to their liking. implemented and maintained. Other GATT articles play a supporting role, including Article II (on Transparency schedules of concessions). The tariff commitments made by WTO members in a multilateral trade Enforcement of commitments requires access to negotiation and on accession are enumerated in information on the trade regimes that are main- schedules (lists) of concessions. These schedules tained by members. The agreements administered 43 T H E W O R L D T R A D E O R G A N I Z AT I O N by the WTO therefore incorporate mechanisms mitments on trade policies that are subject to bind- designed to facilitate communication between ing dispute settlement, can also have this effect. WTO members on issues. Numerous specialized committees, working parties, working groups, and Safety Valves councils meet regularly in Geneva. These interac- tions allow for the exchange of information and A final principle embodied in the WTO is that, in views and permit potential conflicts to be defused specific circumstances, governments should be able efficiently. to restrict trade. There are three types of provisions Transparency is a basic pillar of the WTO, and it in this connection: (a) articles allowing for the use of is a legal obligation, embedded in Article X of the trade measures to attain noneconomic objectives; (b) GATT and Article III of the GATS. WTO members articles aimed at ensuring "fair competition"; and (c) are required to publish their trade regulations, to provisions permitting intervention in trade for eco- establish and maintain institutions allowing for the nomic reasons. Category (a) includes provisions review of administrative decisions affecting trade, allowing for policies to protect public health or to respond to requests for information by other national security and to protect industries that are members, and to notify changes in trade policies to seriously injured by competition from imports. The the WTO. These internal transparency require- underlying idea in the latter case is that governments ments are supplemented by multilateral surveil- should have the right to step in when competition lance of trade policies by WTO members, becomes so vigorous as to injure domestic competi- facilitated by periodic country-specific reports tors. Although it is not explicitly mentioned in the (trade policy reviews) that are prepared by the sec- relevant WTO agreement, the underlying rationale retariat and discussed by the WTO General Coun- for intervention is that such competition causes cil. (The Trade Policy Review Mechanism is political and social problems associated with the described in Box 6.1.) The external surveillance need for the industry to adjust to changed circum- also fosters transparency, both for citizens of the stances. Measures in category (b) include the right to countries concerned and for trading partners. It impose countervailing duties on imports that have reduces the scope for countries to circumvent their been subsidized and antidumping duties on imports obligations, thereby reducing uncertainty regard- that have been dumped (sold at a price below that ing the prevailing policy stance. charged in the home market). Finally, under category Transparency has a number of important bene- (c) there are provisions allowing actions to be taken fits. It reduces the pressure on the dispute settle- in case of serious balance of payments difficulties or ment system, as measures can be discussed in the if a government desires to support an infant industry. appropriate WTO body. Frequently, such discus- sions can address perceptions by a member that a From GATT to WTO specific policy violates the WTO; many potential disputes are defused in informal meetings in Gene- Over the more than four decades of its existence, the va. Transparency is also vital for ensuring "owner- GATT system expanded to include many more ship" of the WTO as an institution--if citizens do countries. It evolved into a de facto world trade not know what the organization does, its legitimacy organization, but one that was increasingly frag- will be eroded. The trade policy reviews are a mented as "side agreements" or codes were negoti- unique source of information that can be used by ated among subsets of countries. Its fairly complex civil society to assess the implications of the overall and carefully crafted basic legal text was extended or trade policies that are pursued by their govern- modified by numerous supplementary provisions, ments. From an economic perspective, transparency special arrangements, interpretations, waivers, can also help reduce uncertainty related to trade reports by dispute settlement panels, and council policy. Such uncertainty is associated with lower decisions. Some of the major milestones are sum- investment and growth rates and with a shift in marized in Table 6.1. resources toward nontradables (Francois 1997). The GATT's early years were dominated by acces- Mechanisms to improve transparency can help sion negotiations and by a review session in the lower perceptions of risk by reducing uncertainty. mid-1950s that led to modifications to the treaty. WTO membership itself, with the associated com- Starting in the mid-1960s, recurring rounds of mul- 44 The WTO: Functions and Basic Principles B O X 6 . 1 : T R A N S PA R E N C Y: N O T I F I C AT I O N A N D S U R V E I L L A N C E Transparency at both the multilateral (WTO) level enhances communication, thereby strengthening and the national level is essential to ensure owner- the multilateral trading system. Country-specific ship of commitments, reduce uncertainty, and reviews are conducted on a rotational basis, and enforce agreements. Efforts to increase the trans- the frequency of review is a function of a mem- parency of members' trade policies take up a good ber's share in world trade. The four largest play- portion of WTO resources. The WTO requires that ers--the European Union, the United States, all trade laws and regulations be published. Article Japan, and Canada--are subject to review by the X of the GATT, Article III of the GATS, and Article 63 WTO General Council every two years. In princi- of the TRIPS agreement all require that relevant ple, the next 16 largest traders are subject to laws, regulations, judicial decisions, and administra- reviews every four years, and the remaining tive rulings be made public. More than 200 notifi- members are reviewed every six years. A longer cation requirements are embodied in the various periodicity may be established for least-devel- WTO agreements and mandated by ministerial and oped countries. The trade policy review (TPR) for council decisions. The WTO also has important sur- a country is based on a report prepared by the veillance activities, since it has a mandate to period- government concerned and on a report by the ically review the trade policy and foreign trade WTO Trade Policies Review Division. TPRs are regimes of members. The WTO's Trade Policy supplemented by an annual report by the Direc- Review Mechanism (TPRM), established during the tor-General of the WTO that provides an Uruguay Round, builds on a 1979 Understanding overview of developments in the international on Notification, Consultation, Dispute Settlement, trading environment. and Surveillance under which contracting parties By subjecting the trade policies of the largest agreed to conduct a regular and systematic review industrial country markets to regular public peer of developments in the trading system. The objec- review, the TPRM shifts the balance of power in tive of the TPRM is to examine the impact of mem- the WTO ever so slightly in favor of the develop- bers' trade policies and practices on the trading ing countries (Francois 2001). Equally important, system and to contribute to improved adherence the TPRM provides domestic interest groups with to WTO rules through greater transparency. The the information necessary to determine the costs legal compatibility of any particular measure with and benefits of national trade policies. The WTO disciplines is not examined, this being left for reports are not analytical in the sense of deter- members to ascertain. mining the economic effects of various national The TPRM was originally motivated in part by policies--the size of the implied transfers and the concerns stemming from the fact that the only beneficiaries and losers under the prevailing poli- available review of global trade policies at the cies. This task is left to national stakeholders time was produced by the United States (Keesing (think tanks and policy institutes). 1998). The TPRM is an important element of the WTO because it fosters transparency and Sources: Hoekman and Kostecki (2001); Francois (2001). tilateral trade negotiations gradually expanded the There are many similarities between the GATT and scope of the GATT to take in a larger number of the WTO, but the basic principles remain the same. nontariff policies. Until the Uruguay Round, how- The WTO continues to operate by consensus and to ever, no progress was made on agriculture or on tex- be member driven. There were, however, a number of tiles and clothing. The deal that finally allowed these major changes. Most obviously, the coverage of the sectors to be subjected to multilateral disciplines WTO is much wider. A change of great importance is included the establishment of rules for trade in that in contrast to the GATT, the WTO agreement is a services and enforcement of intellectual property "single undertaking"--all its provisions apply to all rights (IPRs), as well as the creation of the WTO. members. Under the GATT there was great flexibility 45 T H E W O R L D T R A D E O R G A N I Z AT I O N for countries to "opt out" of new disciplines, and in intended to strengthen the political guidance of the practice many developing countries did not sign spe- WTO and enhance the prominence and credibility cific agreements on issues such as customs valuation of its rules in domestic political arenas. Article II of or subsidies. This is no longer the case, implying that the Marrakech Agreement that established the WTO the WTO is much more important for developing charges the organization with providing a common countries than the GATT was. Also important were institutional framework for the conduct of trade changes in the area of dispute settlement, which relations among its members in matters to which became much more "automatic" with the adoption agreements and associated legal obligations apply. of a "negative consensus" rule. (All members must Four annexes to the WTO define the substantive oppose the findings in a dispute settlement to block rights and obligations of members. Annex 1 has adoption of reports.) Finally, the secretariat acquired three parts: Annex 1A, Multilateral Agreements on much greater transparency and surveillance func- Trade in Goods, which contains the GATT 1994 (the tions through the creation of the Trade Policy Review GATT 1947 as amended by a large number of Mechanism. understandings and supplementary agreements negotiated in the Uruguay Round); Annex 1B, which contains the GATS; and Annex 1C, the TRIPS Scope, Functions, and Structure of agreement. Annex 2 contains the Understanding on the WTO Rules and Procedures Governing the Settlement of The WTO is headed by a ministerial conference of all Disputes (DSU)--the WTO's common dispute set- members that meets at least once every two years. By tlement mechanism. Annex 3 contains the Trade contrast, under the GATT a decade could pass Policy Review Mechanism (TPRM), an instrument between ministerial meetings. The more frequent for surveillance of members' trade policies. Finally, participation by trade ministers under the WTO was Annex 4, Plurilateral Trade Agreements, consists of Table 6.1 From GATT to WTO: Major Events Date Event 1947 The GATT is drawn up to record the results of tariff negotiations among 23 countries. The agreement enters into force on January 1, 1948. 1948 The GATT provisionally enters into force. Delegations from 56 countries meet in Havana, Cuba, to consider the final draft of the International Trade Organization (ITO) agreement; in March 1948, 53 countries sign the Havana Charter establishing an ITO. 1950 China withdraws from the GATT. The U.S. administration abandons efforts to seek con- gressional ratification of the ITO. 1955 A review session modifies numerous provisions of the GATT. The United States is granted a waiver from GATT disciplines for certain agricultural policies. Japan accedes to the GATT. 1965 Part IV (on trade and development) is added to the GATT, establishing new guidelines for trade policies of and toward developing countries. A Committee on Trade and Develop- ment is created to monitor implementation. 1974 The Agreement Regarding International Trade in Textiles, better known as the Multifibre Arrangement (MFA), enters into force. The MFA restricts export growth in clothing and textiles to 6 percent per year. It is renegotiated in 1977 and 1982 and extended in 1986, 1991, and 1992. 1986 The Uruguay Round is launched in Punta del Este, Uruguay. 1994 In Marrakech, on April 15, ministers sign the final act establishing the WTO and embody- ing the results of the Uruguay Round. 1995 The WTO enters into force on January 1. 1999 Ministerial meeting in Seattle fails to launch a new round. 2001 A new round of trade talks (the Doha Development Agenda) is agreed on in Doha, Qatar. Source: Hoekman and Kostecki (2001). 46 The WTO: Functions and Basic Principles Tokyo Round codes that were not multilateralized ment and between trade and competition policy, in the Uruguay Round and that therefore bind only and with the issue of transparency in government their signatories. Together, Annexes 1 through 3 procurement. Specific committees address matters embody the multilateral trade agreements. Article II relating to the GATS or the TRIPS agreement. All of the WTO specifies that all the agreements con- WTO members may participate in all councils, tained in these three annexes are an integral part of committees, and other bodies, with the exceptions the WTO agreement and are binding on all mem- of the Appellate Body, dispute settlement panels, the bers. All of these instruments are discussed further Textiles Monitoring Body, and committees dealing in this chapter or in other chapters of this volume, with plurilateral agreements. The WTO is charged with facilitating the imple- About 40 councils, committees, subcommittees, mentation and operation of the multilateral trade bodies, and standing groups or working parties agreements, providing a forum for negotiations, functioned under WTO auspices in 2000, more than administering the dispute settlement mechanism, twice the number under the GATT. Such bodies are exercising multilateral surveillance of trade policies, open to all WTO members, but generally only the and cooperating with the World Bank and the IMF more important trading nations (less than half of to achieve greater coherence in global economic the membership) regularly send representatives to policymaking (Art. III WTO). Between meetings of most meetings. The degree of participation reflects the ministerial conference, which is responsible for a mix of national interests and resource constraints. carrying out the functions of the WTO, the organi- The least-developed countries, in particular, tend zation is managed by the General Council, at the not to be represented at these meetings; often, they level of diplomats. The General Council meets do not have delegations based in Geneva. All of about 12 times a year. On average, about 70 percent these fora, plus working parties on accession (aver- of all WTO members take part in its meetings, at aging close to 30 in the late 1990s), dispute settle- which members are usually represented by delega- ment panels, meetings of regional groups, meetings tions based in Geneva. The General Council turns of heads of delegations, and numerous ad hoc and itself, as needed, into a body that adjudicates trade informal groups add up to 1,200 events a year at or disputes (the Dispute Settlement Body, or DSB) or near WTO headquarters in Geneva. Most WTO that reviews members' trade policies (the Trade Pol- business is conducted in English, but many official icy Review Body, or TPRB). WTO meetings require French and Spanish inter- Three subsidiary councils, on goods, on services, pretation. and on intellectual property rights, operate under The main actors in the day-to-day activities are the general guidance of the General Council. Sepa- officials affiliated with the delegations of members. rate committees deal with the interests of develop- The WTO--like the 1947 GATT--is therefore ing countries (Committee on Trade and something of a network organization (Blackhurst Development); surveillance of trade restriction 1998). The WTO secretariat is the hub of a very actions taken for balance of payment purposes; sur- large and dispersed network comprising official veillance of regional trade agreements; trade-envi- representatives of members based in Geneva, civil ronment linkages; and WTO finances and servants based in capitals, and national business and administration. Additional committees or working nongovernmental groups that seek to have their parties deal with matters covered by the GATT, the governments push for their interests at the multilat- GATS, or the TRIPS agreement. There are commit- eral level. The operation of the WTO depends on tees, functioning under the auspices of the Council the collective input of thousands of civil servants on Trade in Goods, on subsidies, antidumping and and government officials who deal with trade issues countervailing measures, technical barriers to trade in each member country. (product standards), import licensing, customs val- Initiatives to launch multilateral trade negotia- uation, market access, agriculture, sanitary and phy- tions and to settle disputes--the two highest-profile tosanitary measures, trade-related investment activities of the WTO--are the sole responsibility of measures, rules of origin, and safeguards. In addi- WTO members themselves, not the secretariat. The tion, working groups have been established to deal member-driven nature of the organization puts a with notifications, with state-trading enterprises, considerable strain on the national delegations of with the relationships between trade and invest- members. Many countries have no more than one 47 T H E W O R L D T R A D E O R G A N I Z AT I O N or two persons dealing with WTO matters; a large and mechanisms have therefore been developed minority has no delegations in Geneva at all. over the years to reduce the number of members that are active participants in WTO deliberations. The first and most important device is to involve Decisionmaking only "principals," at least initially. To some extent Most decisionmaking in the WTO follows GATT this is a natural process; a country that has no agri- practices and is based on consultation and consen- cultural sector is unlikely to be interested in discus- sus. The consensus practice is of value to smaller sions centering on the reduction of agricultural countries, as it enhances their negotiating leverage trade barriers. In general the "Quad" economies-- in the informal consultations and bargaining that Canada, the European Union, Japan, and the United precede decisionmaking, especially if they are able States--are part of any group that forms to discuss to form coalitions. Although recourse to voting may any topic. They are supplemented by countries that be had if a consensus cannot be reached, in practice have a principal supplying interest in a product and voting occurs only very rarely. If a vote is needed, it by the major (potential) importers whose policies is based on the principle of "one member, one are the subject of interest. Finally, a number of vote." Unanimity is required for amendments relat- countries that have established a reputation as ing to general principles such as MFN or national spokespersons tend to be involved in most major treatment. Interpretation of the provisions of the meetings. Historically, such countries have included WTO agreements and decisions on waivers of a Egypt, India, and Yugoslavia. member's obligations require approval by a three- During the Tokyo and Uruguay Rounds, con- quarters majority vote. A two-thirds majority vote is tentious issues as to which deals had to be struck sufficient for amendments relating to issues other were often thrashed out in the "green room," a con- than the general principles mentioned above. ference room adjacent to the Director-General's Where not otherwise specified, and where consen- offices. Green-room meetings were part of a consul- sus cannot be reached, a simple majority vote is, in tative process through which the major countries principle, sufficient. In practice, voting does not and a representative set of developing countries--a occur. Indeed, in 1995 WTO members decided not total of 20 or so delegations--tried to hammer out to apply provisions allowing for a vote in the case of the outlines of acceptable proposals or negotiating accessions and requests for waivers but to continue agendas. Such meetings generally involved the to proceed on the basis of consensus (WT/L/93). active participation and input of the Director-Gen- Legislative amendments are also likely to be quite eral. The convention now is to call such meetings rare, as, in practice, changes to the various agree- green-room gatherings, no matter where they are ments occur as part of broader multilateral rounds. held. The green-room process became a contentious issue during the Seattle ministerial meeting; many developing countries that were excluded from criti- Management of the Secretariat and cal green-room meetings, where attempts were Daily Operations being made to negotiate compromise texts of a draft Unlike the World Bank and the IMF, the WTO does agenda for a new multilateral trade negotiation, felt not have an executive body or a board comprising a that they were not being kept informed of develop- subset of members some of whom represent a num- ments and were not being granted the opportunity ber of countries. Such executive boards facilitate to defend their views. Proposals have been made decisionmaking by concentrating discussions with- periodically to formalize the green-room process by in a smaller but representative group of members. creating an executive committee to manage the The closest the GATT ever came to such a forum WTO agenda, based on shares in world trade was the Consultative Group of Eighteen (CG18), (Schott and Buurman 1994). To date, no progress in established in 1975. It ceased meeting in 1985 and this direction has proved possible in the WTO. never substituted for the GATT Council of Repre- sentatives (Blackhurst 1998). Conclusion As of January 1, 2002, the WTO had a member- ship of 144. Achieving consensus among such a The Uruguay Round and the establishment of the large number of members is not a simple matter, WTO changed the character of the trading system. 48 The WTO: Functions and Basic Principles The GATT was very much a market access­oriented customs valuation, a good case can be made that institution: its function was to harness the dynamics implementation should be linked to national capac- of reciprocity for the global good. Negotiators could ity and international assistance (Hoekman 2002). be left to follow mercantilist logic, and the end A lesson from post­Uruguay Round experience result would be beneficial to all contracting parties. and thinking is that trade policy should be made This dynamic worked less well for developing coun- more central to the development process and devel- tries, where the burden of liberalization rested opment strategies. This needs to be done at both the much more heavily on the shoulders of govern- national and international levels. At the national ments. Even if they wanted to, their scope to use the level it is necessary in order to ensure that govern- GATT was often limited because exporters had ments have a basis on which to resist efforts to fewer incentives and were less powerful than in negotiate agreements in an area. Governments must industrial countries. The reciprocal, negotiation- be able to identify what types of rules will promote driven dynamic also worked much less well for development and what types would lead to an inap- issues that were "lumpy" and where the terms of the propriate use of scarce resources. At the interna- debate revolved around what rules to adopt, not tional level such a change is necessary in order to around how much of a marginal change was appro- enhance the communication between trade and priate. Once discussions center on rules, especially development assistance bodies in member coun- on disciplines for domestic policy and regulations, tries. One reason for the implementation assistance it is more difficult to define intraissue compromises problems that were encountered in the late 1990s that make economic sense. Cross-issue linkage was that the best-endeavors commitments on assis- becomes necessary. Disengagement was not an tance that were made by industrial country trade option during the Uruguay Round (because of the negotiators were not "owned" by counterpart agen- "single undertaking"), so the task was to come up cies in their governments that controlled develop- with a balanced package that ensured gains for all ment assistance money. Progress on both fronts players. One can argue whether the package that would do much to ensure that future negotiations emerged from the round was a balanced one; views do not give rise to problems of the type that were on this point differ widely. created in the Uruguay Round. Whatever the conclusion, it is clear that the approach taken toward ensuring and supporting Notes implementation of WTO agreements by developing This chapter draws on Hoekman and Kostecki (2001). countries was not an effective one. Limiting recog- nition of this problem to the setting of uniform 1 The founding parties to the GATT (giving the names used at transition periods was clearly inadequate. The case the time) were Australia, Belgium, Brazil, Burma, Canada, Cey- lon, Chile, China, Cuba, Czechoslovakia, France, India, for uniform application of agreements that involve Lebanon, Luxembourg, the Netherlands, New Zealand, Nor- reducing trade barriers--tariffs and nontariff barri- way, Pakistan, Southern Rhodesia, Syria, South Africa, the Unit- ers--is very strong. But in other areas requiring ed Kingdom, and the United States. Subsequently, China, minimum levels of institutional capacity, such as Lebanon, and Syria withdrew. 49 7 J . M I C H A E L F I N G E R L . A L A N W I N T E R S Reciprocity in eralization by developing coun- tries, and an "apples versus oranges" problem that arises the WTO because the WTO spans both border trade restrictions (tar- iffs, quotas, and the like) and within-border regulatory struc- Reciprocity: Mutual or correspondent concession of advantages or tures such as standards and privileges, as forming a basis for the commercial relations intellectual property. Failure to between two countries. recognize the apples versus --The Oxford English Dictionary1 oranges problem, we argue, has led to a troublesome Uruguay Round outcome. R eciprocity has been a motivating Reciprocity in GATT Rules principle of the GATT/WTO system. Although the economics of import restrictions recog- The thrust of the GATT/WTO system is that agree- nizes that the losses from a country's own restric- ment defines reciprocity (or balance), not the other tions exceed domestic gains, the politics has not way around. An agreed outcome from a negotiating found a way to enfranchise the domestic interests round, the system presumes, is an outcome that each that bear these domestic losses--users and con- member considers advantageous, by whatever stan- sumers of imports. When trade policy involves an dard the member chooses to apply. Beyond that, vari- exchange of domestic restrictions for foreign restric- ous provisions for adjustment, such as renegotiation tions, this amplifies the voice of export interests. The and safeguard actions,attempt to maintain the balance success of the GATT/WTO system manifests the that the agreement has established. The same holds for ingenuity of reciprocally agreed liberalization as a dispute settlement. In this section we look at how reci- means of transferring political power over domestic procity enters into each of these parts of the system. import restrictions to export interests, and it also manifests the power of these interests. Negotiations In this chapter we explore the role of reciprocity in GATT/WTO negotiations and in the processes of Reciprocity serves to motivate negotiations. Partici- making adjustments and settling disputes under or pants and commentors use reciprocity, or its func- within an agreement. We look at the role of reci- tional equivalent, "balance," as a standard against procity in past agreements, and we present evi- which to evaluate an outcome. The rules, however, dence suggesting that reciprocity is not the only do not define that standard; determining the stan- force that shapes the outcome of a negotiation. We dard is part of the evaluation itself.2 then turn to two issues that relate to reciprocity: The GATT, and the Marrakech Agreement that "credit" in reciprocal negotiations for unilateral lib- established the WTO, refer in their preambles to 50 Reciprocity in the WTO "entering into reciprocal and mutually advanta- my question of what it means in negotiating prac- geous arrangements directed to the substantial tice--what countries have interpreted as equivalent reduction of tariffs and other barriers to trade." concessions, and what they have not. We take up GATT Article XXVIII bis, the article that provides this topic below. for negotiations to be held, refers also to "negotia- tions on a reciprocal and mutually advantageous Treatment of Developing Countries in Negotia- basis." Neither the GATT nor the WTO provides tions. Part IV of the GATT provides elaborate com- further specification of what is "reciprocal" or of mitments to the developing countries. For what is "mutually advantageous." The logic of the example, Article XXXVI.8 states, "The developed GATT/WTO is that in the negotiations each mem- contracting parties do not expect reciprocity for ber is sovereign to determine for itself whether a commitments made by them in trade negotiations proposed agreement is to its advantage--to decide to reduce or remove tariff and other barriers to the the criteria by which to identify the pluses and trade of less-developed contracting parties." The minuses, and to apply those criteria by whatever commitments of Part IV, however, are not legally formula the member considers appropriate. The binding. Exhortations such as the one quoted are GATT's tradition of decision by consensus rein- qualified by other phrases: for example,"The devel- forces the idea that an agreement is an outcome that oped countries shall to the fullest extent possible-- each member considers to be to its benefit. If any that is, except where compelling reasons, including one member does not find the outcome advanta- legal reasons, make it impossible . . ."(Art. geous, the proposed agreement does not go into XXXVII.1), and "The adoption of measures to give effect. effect to these principles and objectives shall be a matter of conscious and purposeful effort on the GATT Elaboration on Reciprocity in Negotiations. part of the contracting parties both individually An early (1955) GATT working party, in response to and jointly" (Art. XXXVI.9). The operational a proposal to establish rules for how concessions meaning of such phrases is to make clear that they should be measured, concluded that "governments are not legal commitments. The commitment is to participating in negotiations should retain com- a nonmeasurable "conscious and purposeful plete freedom to adopt any method they might feel effort," not to a measurable result. most appropriate for estimating the value of duty Although such statements do not express legal reductions and bindings . . ." The working party obligations, they do have behind them the weight of went on to note that "there was nothing in the moral suasion; they are intended to influence behav- Agreement, . . . to prevent governments from adopt- ior without going so far as to regulate it. This moral ing any formula they might choose, and therefore suasion has not delivered much. For example, the considered that there was no need for the Contract- Trade-Related Aspects of Intellectual Property ing parties to make any recommendation in this Rights (TRIPS) agreement, the customs valuation matter" (GATT 1994a: 912­13). Similarly, Arthur agreement, the Sanitary and Phytosanitary (SPS) Dunkel, Director-General of the GATT from 1980 agreement, and several other Uruguay Round agree- to 1992, observed, "Reciprocity cannot be deter- ments suggested that industrial country members mined exactly; it can only be agreed upon" (GATT furnish technical assistance to developing country Press Release 1312, March 5, 1982). members that request it. This provision, however, is Since the GATT and the Marrakech Agreement not a binding commitment; the developing coun- are silent on how a member might measure the tries undertook to implement bound commitments advantage it draws from the agreement, they say in exchange for unbound commitments for assis- nothing about how much one country should gain tance. Although developing countries pressed hard from the negotiations relative to another. The word at the WTO for delivery on such promises, bilateral- "balance" does not appear in the GATT/WTO text ly or through an increased WTO technical budget, on negotiations. An agreement (the outcome of a the high-income countries have done little. The negotiating round) defines balance, not the other stalemate has prompted Rubens Ricupero (2000) to way around. Although the GATT/WTO rules make suggest that in the future, negotiations on topics no demands as to what reciprocity means in a nego- that will involve expensive implementation be tiation, there remains the practical political-econo- accompanied by an "implementation audit" that 51 T H E W O R L D T R A D E O R G A N I Z AT I O N will identify concretely what developing countries involved some degree of objectivity. In the end, will have to do and what it will cost. Short of a however, appropriate compensation or retaliation is bound commitment from the high-income coun- what the parties agree on, not what an objective and tries to meet such costs, statements about imple- exogenous standard dictates. mentation assistance should be omitted. There should be no more instances of creating the rhetoric Safeguard Actions (only) of reciprocity by exchanging bound commit- ments for unbound promises. GATT Article XIX, the "escape clause," or "safe- guard" article, includes a similar provision. (The article, roughly speaking, allows a country to Renegotiations restrict imports that cause injury to domestic pro- Political reality will require from time to time that ducers.) Implicitly, the article calls on the country adjustments be made in the outcome of an agree- that takes safeguard action to provide compensa- ment, and domestic politics will demand that some tion. Explicitly it provides that exporting countries "concession" be withdrawn.3 The GATT article on may retaliate if satisfactory compensation is not renegotiations states, "In such negotiations and offered: "If agreement among the interested con- agreement, which may include provision for com- tracting parties with respect to the action is not pensatory adjustment with respect to other prod- reached, . . . the affected contracting parties shall ucts, the contracting parties concerned shall then be free . . . to suspend . . . the application to the endeavor to maintain a general level of reciprocal trade of the contracting party taking such action . . . and mutually advantageous concessions not less of substantially equivalent concessions or other favorable to trade than that provided for in this obligations under this agreement the suspension of Agreement prior to such negotiations" (Art. XXVI- which the Contracting Parties do not disapprove II. 2). If the importing country wanting to raise a . . ." (GATT Art. XIX.3[a]).4 tariff above a previously bound level does not reach The Uruguay Round Agreement on Safeguards agreement with supplying countries on appropriate mentions compensation explicitly, and, in its Article compensation, the exporting countries will, in time, 8.1, it includes the exhortation in GATT Article be entitled to retaliate-- "to withdraw . . . substan- XXVIII (renegotiations) "to maintain a substantial- tially equivalent concessions" (Art. XXVIII.3a, 3b, ly equivalent level of concessions and other obliga- 4d, 5). Adjustments of concessions should maintain tions." Retaliation, as in the GATT safeguard article, the balance that the previous agreement established. is the suspension of the application of "substantially In practice, many renegotiations have eventually equivalent concessions or other obligations . . ." been taken up as part of a next round of negotia- (Art. 8.2). In practice, the determination of what is tions, and it is not possible to identify in these "substantially equivalent" has been determined instances whether the compensation that the parties strictly by negotiation among the interested parties. agreed to was appropriate. In other instances, deter- The GATT Contracting Parties have never disap- mining what is a "substantially equivalent conces- proved of a countermeasure to an action under sion" has centered on finding an equivalent amount Article XIX (GATT 1994a: 490). of trade and an equivalent change in the degree of protection. One of the less complex--but still not Dispute Settlement simple--parts of the process has been to agree on an appropriate base period in which to measure the Although compensation and retaliation are part of amounts of trade involved. Other parts were more the vocabulary of GATT/WTO dispute settlement, difficult; for example, often what was at issue was the process primarily has to do with maintaining not simple tariffs but more complicated tariff quo- behavior within an agreement rather than adjusting tas. Many renegotiations have stemmed from cre- what was agreed.5 The GATT text that is relevant to ation of a customs union, and here the task is to compensation and retaliation reads: "If the Con- compensate for discrimination, not merely for tracting Parties consider that the circumstances are change in a tariff rate. serious enough to justify such action, they may Maintaining balance--determining what is an authorize a contracting party or parties to suspend equivalent adjustment to an agreed outcome--has the application to any other contracting party or 52 Reciprocity in the WTO parties of such concessions or other obligations Control of Free-Riding under this Agreement as they determine to be appro- priate in the circumstances" (GATT Art. XIX.2; The initial GATT rounds consisted of bilateral emphasis added). negotiations on modification of most-favored- Robert Hudec (1978) has explained that the nego- nation (MFN) schedules, conducted among a limit- tiators of the proposed International Trade Organi- ed number of countries. (For example, in the 1947 zation (ITO) and the GATT were ambiguous as to round the United States negotiated with 16 coun- whether dispute settlement concerned compensa- tries that supplied about two-thirds of U.S. tion or compliance. In the 48 years of the GATT, imports.) Efforts were made in these negotiations to there was only one instance in which the dispute limit concessions to products imported in large part settlement mechanism authorized retaliation. In the from other participants. Table 7.1, a tabulation of WTO's first six years, the WTO dispute settlement U.S. experience in early rounds, shows, for example, mechanism has twice authorized retaliation: in the that at the Dillon Round 96 percent of U.S. tariff Bananas case and the Beef Hormone case. As of late cuts--all made on an MFN basis--were on imports 2001, both were still in political dispute. from countries that made concessions in return. At that time, 66 percent of U.S. imports came from these countries. The difference between the 96 per- Reciprocity and Other Influences on an cent and the 66 percent reflects the emphasis on Agreement limiting concessions to products imported almost One can find, in the results of negotiations, clear entirely from countries that reciprocated. Attention evidence of the influence of reciprocity. One can to internalizing the concessions (that is, to limiting also find evidence that more is at play than reaching free-riding) led, however, to low coverage of the tar- a mercantilist balance of concessions received ver- iff cuts--for the United States, the reductions cov- sus concessions given. In this section we review ered 15 percent of dutiable imports in the 1956 other influences that are likely to shape a negotia- round and 20 percent in the 1960­61 round. tion. We also present scattered evidence of the influ- At the Kennedy Round, bilateral bargaining over ence of reciprocity and of other factors. tariff cuts was replaced by formula cuts. The shift to Table 7.1 Control of Free-Riding in GATT Negotiations: U.S. Experience, 1947­67 Kennedy Round, 1964­67 Geneva Annecy Torquay Geneva Dillon Major All Round, Round, Round, Round, Round, partici- partici- 1947 1949 1951 1956 1960­61 pantsa pants Percentage of dutiable imports from all countries subject to cuts 35 37 26 15 20 -- 44 Percentage of dutiable imports coming from participants 65 6 34 67 66 68 72 Percentage of dutiable imports subject to cuts coming from participants 84 39 64 89 96 81 91 --Not available. a. Austria, Canada, Denmark, European Economic Community, Finland, Japan, Sweden, Switzerland, and the United Kingdom. Source: Finger (1979): 424­25. 53 T H E W O R L D T R A D E O R G A N I Z AT I O N a formula approach did lead to broader reduc- although public works and other benefits have also tions--U.S. cuts applied to 44 percent of imports. been used.6 Exclusion of free riders took the form of negotiation Another way in which governments attempt to over "exclusion lists" and, as the figures show, did minimize the problem of compensating losers is by limit spillover to free riders to only 9 percent of con- taking advantage of the large volume of intraindus- cession imports. try trade that characterizes the modern trading sys- tem. To the extent that concessions given by an industry can be offset by concessions received on Fewer Concessions Given, Fewer Received the products exported by the same industry, the Table 7.2 provides another indication that to receive government need not develop interindustry mecha- concessions, a country has to give concessions. The nisms for balancing losers against winners. Gilbert message is obvious: the lower the degree of partici- Winham (1986: 65) notes that from the Kennedy pation in the negotiations, the lower the share of Round forward, there has evolved a tendency to exports affected by the concessions of other partici- look for such "self-balancing sectors." In regional pants. agreements such balancing can be increased through the use of rules of origin. An example is the way in which the North American Free Trade Agree- Domestic Reciprocity ment (NAFTA) and other regional agreements in At the same time, there is more to the liberalization which the United States participates condition process than concessions given over the interna- access to the U.S. market for textile products on the tional table versus concessions received. The bar- use of U.S.-made fibers or fabrics. gaining process ties access to foreign markets to the granting of access to the domestic market and Noneconomic Objectives thereby mobilizes export interests to favor import liberalization. But the domestic politics of setting War, according to Clausewitz, is the pursuit of gains for export industries against losses for diplomacy through other means. Often, so is trade import-competing industries is not frictionless. For policy. Freedom of international commerce was the a government motivated toward trade liberaliza- third of U.S. President Woodrow Wilson's Fourteen tion, the tough tradeoffs are not between it and for- Points. To Cordell Hull, secretary of state for Presi- eign governments but between domestic winners dent Franklin D. Roosevelt, the link was straightfor- and losers. ward: "Unhampered trade dovetailed with peace; Overcoming such frictions has been, in practice, high tariffs, trade barriers and unfair economic partly a matter of power--in the simple case in competition with war" (Hull 1948: 81). After World which negotiating authority must be specifically War II, leadership in Europe and in the United granted, using export industries to win more con- States saw economic union in Europe and the con- gressional votes than the opposition can rally. It has struction of an open global trading system more as also been partly a matter of compensation. Adjust- strategic objectives than as economic ones. A gov- ment assistance is the straightforward example, ernment that can mobilize noneconomic motives Table 7.2 U.S. Imports Covered by Kennedy Round Tariff Concessions (Reductions plus Bindings) as a Share of Total U.S. Imports from the Country Group (percent) Country group Share (percent) Major participants 70 Other industrial country participants 49 Active developing country participants 33 Other developing countries 5 Note: Data are for 1994 imports. Source: Finger (1979): 435. 54 Reciprocity in the WTO into significant support for trade liberalization will and intellectual property rights unless high-income be in a position to play a hegemonic role (as the countries negotiated on agriculture and on textiles United States did in early GATT rounds) and to and clothing. make larger concessions than it receives in exchange. Equal Sacrifice, Softly Applied When, however, the time came for totting up, the Locking In equal sacrifice concept was the one that delegations Developing countries that have unilaterally liberal- used.7 Particularly in the last month of the negotia- ized sometimes view binding such liberalization tions (the mopping-up phase) the negotiators internationally as a defense against the risk of back- devoted significant attention to ensuring that each sliding should political authority shift or popular participating country had made an appropriate support wane. Such an objective, like a noneconom- contribution to the tariff reduction exercise. Dele- ic one, can motivate a government to accept what a gations widely but informally accepted that the tar- calculation based strictly on considerations of mer- gets were average reductions of one-third for cantilist market access would view as a bad bargain. industrial countries and one-fourth for developing countries.8 The negotiating guidelines lacked preci- sion; for example, was the one-third cut to be a Individual Sacrifice for the Common Good weighted or an unweighted average? Over all prod- Many of the participants in the initial ITO and ucts, or only over dutiable items? GATT negotiations viewed their task as the con- Tied up with achievement of these targets was the struction of a system from which all countries would question of how countries would receive "credit" for derive significant noneconomic benefits (perhaps unilateral tariff reductions and for extensions of economic gains as well, but the emphasis was usual- bindings that did not imply tariff cuts. The agricul- ly on noneconomic considerations). This view of tural negotiations established formal negotiating reciprocity differs from the mercantilist bargaining guidelines, not only on the amounts by which model in that the benefits a participant gets are not import restrictions and other agricultural supports unequivocally identified with the particular market were to be reduced by each country but also for how access concessions that the country receives; the link agricultural nontariff barriers were to be converted between contribution and benefit is amorphous, to tariffs. resulting from the collective nature of the system These percentages were negotiating guidelines, rather than from any particular element of it. not bound commitments. Even in agriculture, Robert E. Hudec, in his 1987 book Developing where the negotiating guidelines were circulated as Countries in the GATT Legal System, builds on the a GATT document (GATT 1993b), legal commit- common-good view to provide a convincing inter- ments were the rates notified on each country's pretation of how GATT members came to accept schedule. The GATT/WTO members seem to have "special and differential treatment" as the appropri- policed these guidelines rather softly. Interviews ate attitude toward developing countries. In con- with more than a dozen delegations after the round structing any system from the contributions of its found none that had attempted to calculate the members, Hudec notes, it is difficult to ask the less depth of cut by each country, or even for major well off to contribute proportionally with the bet- trading partners. Likewise, no delegation had tabu- ter-off members. lated concessions received--that is, the coverage of its exports by concessions scheduled by other coun- tries. A number of developing country delegations The Uruguay Round Tariff Reductions pointed to the agriculture and the textiles and cloth- Scorecard: What We Learn from It ing agreements as evidence that they had paid atten- Reciprocity--"get what you pay for," or, more tion to what they would receive, but in agriculture, aggressively, "pay for what you get!"--was clearly too, although the guidelines were more precise, the motivating principle of the Uruguay Round either the numbers were checked only casually or negotiations. For example, developing countries disregard of the guidelines was widely accepted. would not negotiate in "new areas" such as services After-the-fact examination has turned up a lot of 55 T H E W O R L D T R A D E O R G A N I Z AT I O N "dirty tariffication"--tariff rates considerably in ever established for what any country should "give" excess of those that the guideline formulas generate or could expect to "receive." It is hence impossible to (Hathaway and Ingco 1996). measure the extent to which credit for unilateral lib- The interviews revealed that as the negotiations eralization was given. were being completed, selling the agreement at On the question of how to take into account home--that is, gaining approval--was an impor- bindings that did not imply tariff cuts, such as ceil- tant consideration. The issue was not the overall ing bindings, not even an unofficial approach balance of concessions; it was to make sure that evolved. Toward the end of 1990 the Mexican dele- powerful domestic constituencies were accommo- gation circulated a nonpaper arguing that credit dated. The focus was on the effect on the big trees, should be given for expansion of the scope of bind- not the forest. ings, but it did not offer a method for measuring the "tariff cut equivalent."11 Later, the chair of the GATT Market Access Group provided guidelines for Concessions Given versus Concessions Received: such measurement, including a matrix of suggested Great Differences among Countries equivalents between depth of tariff cut and scope of "Concessions given" is a familiar concept. Its com- expansion of bindings. The view of the negotiators plement, "concessions received," refers to the con- with whom we spoke was that there never emerged cessions made by trading partners that apply to a even notional agreement on how to convert exten- given country's exports.9 In Table 7.3, columns (1) sion of bindings into a tariff cut equivalent. through (3) show the depth of tariff reductions, and Giving credit for unilateral liberalization by columns (4) through (6) attempt to introduce the developing countries is part of the standard scope as well as the depth of tariff changes. pro­developing country list of what a negotiation It is hard to find equal sacrifice in this subpart of the should do. But the fact that what is an "appropriate" Uruguay Round outcome. If all the reductions were outcome is defined by the agreement and not by an equal, column (2) would show identical numbers for exogenous standard means that calls for credit for each country, as would column (5). It is also hard to unilateral liberalization are an exercise in moral find mercantilist balance. If such balance had been suasion, not an application of economic or account- achieved--that is, if each country received conces- ing science. Finger, Reincke, and Castro (2002: table sions roughly in line with the concessions it gave--all 2) found that the suasion did have an effect. Bind- the numbers in column (6) would be zero, which is ings of unilateral tariff cuts (but not unbound uni- clearly not the case. The summary statistic in the last lateral cuts) do seem to have been counted toward row reports that, on average, a country's imbalance developing countries' fulfillment of their "obliga- (positive or negative) was over half as large as the tion" to reduce tariffs by one-fourth. value of the concessions the country received.10 In sum, calls for credit for unilateral liberaliza- tion--where that liberalization has then been bound under the GATT/WTO--have been effective. Credit for Unilateral Liberalization Calling for a "credit rule," however, reveals a misun- Many developing countries undertook unilateral derstanding of how the GATT/WTO works. liberalization in the 1980s and 1990s. Credit in the reciprocal negotiations for this liberalization was Apples versus Oranges part of the informal guidelines for meeting the equal sacrifice criterion. From what delegations told The "grand bargain," as Sylvia Ostry (2000) has us in interviews, informal practice was more or less labeled it, that was struck at the Uruguay Round was to calculate tariff cuts from the rate prevailing in that the developing countries would take on signifi- 1988 to the rate bound at the Uruguay Round. cant commitments in "new areas" such as intellectu- Developing countries were given credit for unilater- al property and services, where industrial country al liberalization by allowing them to count from the enterprises saw opportunities for expanding inter- rates applied earlier, at the beginning of the 1980s. national sales. The industrial countries, in In any case, no tabulation of country-by-country exchange, would open up in areas of particular tariff cuts was made, either by the GATT secretariat export interest to developing countries: agriculture, or by individual countries, and no formal target was and textiles and clothing. 56 Reciprocity in the WTO Table 7.3 Tariff Concessions Received and Given at the Uruguay Round Mercantilist balance Percentage tariff reductiona (percentage point dollars)b (1) ­ (2) as (4) ­ (5) as percentage Concessions Concessions percentage Received Given of (1) received given of (4) Economy (1) (2) (3) (4) (5) (6) High-income Australia 0.76 3.35 ­341 21,032 88,162 ­319 Austria 2.64 3.74 ­42 74,602 108,820 ­46 Canada 0.22 0.89 ­305 5,291 26,205 ­395 European Union 1.94 2.19 ­13 578,816 627,939 ­8 Finland 3.47 2.52 27 63,924 44,021 31 Hong Kong (China) 2.36 0.00 100 60,258 0 100 Iceland 1.59 0.20 87 2,151 299 86 Japan 2.06 1.06 49 481,006 143,142 70 New Zealand 0.84 0.83 1 5,126 4,155 19 Norway 1.15 2.17 ­89 24,250 44,263 ­83 Singapore 1.96 0.85 57 50,294 32,741 35 Switzerland 2.15 0.89 59 100,659 46,829 53 United States 1.21 1.07 12 214,791 283,580 ­32 Transition Czech and Slovak Union 2.06 1.05 49 9,773 7,312 25 Hungary 1.82 1.69 7 7,755 13,727 ­77 Poland 1.36 1.26 7 8,609 7,112 17 Developing Argentina 0.98 0.00 100 6,331 0 100 Brazil 1.37 0.00 100 38,037 98 100 Chile 0.50 0.00 100 3,291 0 100 Colombia 1.25 0.02 98 6,323 81 99 India 1.22 6.16 ­405 14,380 67,172 ­367 Indonesia 0.87 0.25 71 16,222 3,355 79 Korea, Rep. of 1.87 5.99 ­220 100,809 262,918 ­161 Malaysia 1.46 1.97 ­35 36,108 28,966 20 Mexico 0.16 0.00 100 960 3 100 Peru 0.57 0.03 95 1,586 58 96 Philippines 2.43 1.29 47 19,748 12,847 35 Sri Lanka 1.36 0.01 99 1,595 33 98 Thailand 1.33 5.93 ­346 20,564 95,953 ­367 Tunisia 1.42 0.02 99 2,506 72 97 Turkey 1.72 3.00 ­74 12,557 32,661 ­160 Uruguay 0.52 0.00 100 772 6 99 Venezuela, R. B. de 0.21 0.13 38 2,051 806 61 Summary Sum of absolute differences/ Sum of absolute differences/ statistics sum of received = 86 percent sum of received = 58 percent a. Weighted average of change measured as dT/(1 + Tavg) * 100, where Tavg is the average of the before- and after-change rates, calculated across all tariff lines, including those on which there was no reduction. Why this formula? Whereas cutting by half a tariff of 2 percent saves the importer only 1 cent, cutting by half a tariff of 50 percent saves the importer 25 cents. As a part of what the importer pays, the tariff reduction relates to the tariff charge plus the price received by the seller--to Ps(1 + T) rather than simply to T. Finger, Ingco, and Reincke (1996) provide a more detailed explanation. b. Tariff cut as measured in column (1) or (2) multiplied by the value (in millions of dollars) of the imports or exports to which the import- ing country applies MFN tariff rates. A percentage point dollar is a 1 percent tariff change on 1 dollar of exports or imports. 57 T H E W O R L D T R A D E O R G A N I Z AT I O N What the North gave in this exchange was tradi- The problems that developing countries face in tional market access, reduction of import restric- the new areas largely have to do with project design tions, and in agriculture, reduction of export and cost-benefit analysis--with development eco- subsidies and production subsidies. What the South nomics, not market access. The World Bank and the gave in the new areas was different. WTO obliga- GATT/WTO are different institutions that work in tions on services, on intellectual property rights, different ways. These differences are not arbitrary; and on standards basically have to do with the they reflect what the international community structure of the domestic economy. The industrial knows about how to deal with trade issues versus countries that wanted these areas to be covered by how to deal with development issues (see Finger the WTO rationalized their inclusion by reference and Nogués 2001). to their "trade-related" attributes (although the actual motivation was the trade interests of their The Uruguay Round Scorecard enterprises). Whatever the fig leaf, regulation here is, figuratively speaking, nine-tenths concerned with The outcome of the Uruguay Round was a good one the domestic economy and one-tenth about trade. for the North. Not only did the industrial countries The two sides of the grand bargain have funda- gain from the concessions they received; the eco- mentally different economics. In real economics, nomics of the concessions they gave was also posi- giving away an import restriction is not a cost; it is tive, through the opening up of their own something that enhances the national economic agriculture and textiles sectors. And for the South? interest. GATT bargaining is a response to the diffi- On the gain dimension--market access--develop- cult politics of liberalization, not to the good sense ing countries did not achieve a mercantilist surplus of its economics. The economics of new area (Table 7.4). Their tariff reductions covered as large a responsibilities are different in two respects (see also share of their imports as did those of the industrial Chapter 48, by Finger and Schuler): countries, and their tariff cuts, measured by how these reductions will affect importers' costs, were · Implementing such responsibilities will cost deeper than those of the industrial countries. This is money--for example, for laboratories to develop true even when we take into account the tariff and enforce standards.12 equivalent of the Multifibre Arrangement (MFA) · The result can be a substantial net cost rather quotas that the industrial countries have committed than a benefit. For many developing countries the themselves to remove. economics of TRIPS is the same as the economics For developing countries as well as industrial to oil importers of oil price increases. Just the countries, concessions made in the grand bargain patent changes required by TRIPS will cost some make for difficult domestic politics. For developing countries more than they gain from the whole of countries, these concessions will also mean real eco- the market access liberalization package (see Fin- nomic costs. The scorecard on the Uruguay Round ger 2001). grand bargain? The South's concessions in the new Table 7.4 Uruguay Round Tariff Concessions, All Merchandise Industrial Developing economies economies Percentage Depth of Percentage Depth of of imports cuta of imports cuta Includes tariffication and bound reductions on agricultural products 30 1.0 29 2.3 Includes the above plus the tariff equivalent of elimination of the Multifibre Arrangement 30 1.6 29 2.3 a. Depth of cut, dT/(1 + T), is a weighted average across all products, including those on which no reduction was made. Source: Finger and Schuknecht (2001): table T-1, based on Finger, Ingco, and Reincke (1996). 58 Reciprocity in the WTO areas are, as mercantilism, unrequited--and as real 5 The WTO Dispute Settlement Understanding (Art. 3.7) establish- economics, they are costly. es explicit priorities among different outcomes: "The aim of the dispute settlement mechanism is to obtain a positive solution to a dispute; A solution mutually acceptable to the parties to a dis- Conclusions pute and consistent with the covered agreements is clearly to be preferred; Withdrawal of the measures concerned if they are Reciprocity in negotiations is a motivation and an found to be inconsistent with . . . the covered agreements; objective, not a criterion. Within an agreement, rec- Compensation . . . only if the immediate withdrawal of the pro- iprocity--better known in this context by its other vision is impractical . . . a temporary measure pending the with- drawal of the measure . . . inconsistent with the covered name, balance--comes closer to having an opera- agreements; and, the last resort . . . suspending . . . concessions tional meaning. Still, it is to a large extent some- or other obligations . . ." thing that can be agreed on but not measured. 6 Zeiler (1992) provides examples of the trades U.S. President "Credit" is moral suasion--a useful rallying cry for John F. Kennedy made to win congressional approval of the driving a better deal for the South. It is, however, authority to negotiate in what came to be called the Kennedy futile, and a basic misunderstanding of the Round. Providing quota protection for the textile industry was GATT/WTO, to think that credit can be converted one; another was an extensive waterways project for the state into the "shall" language of obligation. of Oklahoma. Mistaking clean clothes for dirty was acceptable 7 Provided that each country's trade is balanced (that is, exports when that was what both sides brought to the table. equal imports) and that there are uniform cuts with complete Each took home what its politics saw as the other's coverage (of goods and countries), "equal sacrifice" comes to dirty laundry. In economics, each was doubly better the same arithmetic as "get what you pay for." off; trading market access "concessions" was good 8 These guidelines were cited by many of the Geneva delegations economics for the giver as well as for the receiver. that were interviewed by Finger and colleagues as part of the Bringing in the new areas changed things. What the research published as Finger, Reincke, and Castro (2002). As to the origin of the figures, at the July 7­9, 1993, G-7 summit in developing countries are now asked to put on the Tokyo, the "Quad" (Canada, EU, Japan, and the United States) table can have domestic economic costs as well as trade ministers announced a substantial market access agree- domestic political costs. Reciprocal bargaining over ment, as well as their goals for what they hoped to achieve the political dimension may not be enough. overall: reductions to zero for selected products or harmoniza- Progress in these areas may require management of tion at low levels; a 50 percent cut in tariffs 15 percent and above; and, for other tariffs, a negotiated reduction of at least their economic dimensions, as well. one-third. The one-third reduction for industrial countries may thus have come from this agreement; we have not identified Notes the origin of the one-fourth target for developing countries. 9 The formula dT/(1 + T), not dT/T, was used to measure tariff 1 According to the Oxford English Dictionary, the first recorded change. Concessions given for a country are the familiar sum use of "reciprocity" in this sense can be found in the Prelimi- for the country of all its MFN tariff cuts across all tariff lines, nary Articles for the Peace between the United States and weighted by imports. For concessions received, if Dij is the Great Britain, in 1782. reduction of the MFN tariff rate of country i on tariff line (that 2 Finger, Hall, and Nelson (1982) sort decision processes into is, "product") j, and Wijk is the share or weight (by value) of country k's total exports of product j to country i, then the "political" versus "technical" ones. At the technical end of the "reduction received" (column 1 of Table 7.3) by country k is spectrum, the criteria are given, and the decision turns on whether or not the criteria are met. Examples are an the sum, across countries and across products, of Dij multiplied antidumping determination or a jury's decision in a court trial. by Wijk. To calculate percentage point dollars of concessions At the other end of the spectrum, a "political" decision received, for Wijk in that formula we substitute Vijk, the value of k's exports of product i to country j. The countries in the table involves debate over what the criteria are, as well as about are those for which data were available from the WTO Inte- whether the criteria are met. A legislative decision on tax grated Database. See Finger, Ingco, and Reincke (1996) for a reform might be an example. In this framework, reciprocity in description of the database and a more detailed explanation of a negotiation is a political concept. What it means operational- the calculations. ly is not specified by the rules of the negotiation. 10 For some of these countries, counting only the concessions 3 Finger (1998) discusses in more detail GATT/WTO safeguards made at the Uruguay Round leaves out the unilateral liberaliza- and other pressure valve provisions. tion they implemented in the 1980s. If, however, we drop from 4 There are time limits as well as notification and consultation the table all the Latin American countries plus Sri Lanka and requirements. Tunisia, the average (absolute) imbalance is still 56 percent. 59 T H E W O R L D T R A D E O R G A N I Z AT I O N 11 In GATT/WTO usage, a nonpaper is a way of circulating an idea for discussion without proposing that the idea be adopt- ed; it is a means of promoting preliminary discussion. The nonpaper was cosponsored by 19 other developing countries. 12 Although considerable amounts of money will flow in different directions as a result of a tariff reduction and the political fall- out may be severe, implementing the reduction costs nothing. The signature of an executive or ratification by a legislature does it. 60 8 C O N S TA N T I N E M I C H A L O P O U L O S WTO Accession tries applying to join the WTO, many developing countries and economies face very similar chal- lenges in establishing the institu- tions needed to implement WTO commitments. Perhaps the most important of these challenges is the need to introduce laws and institutions for the operation of private enterprises and markets free from government controls-- other than those explicitly pro- vided under WTO regulations-- A regarding, for example, standards, ccession to the WTO is a complex, sanitary and phytosanitary (SPS) provisions, intellec- difficult, and lengthy process. In tual property rights, and state-trading practices. May 2001 it was a process being faced by 28 coun- Equally important to a country's economy is the tries, 9 of them transition economies and about half introduction of greater stability in commercial poli- of the remainder, least-developed countries (LDCs). cy, which is a consequence of adherence to WTO This chapter analyzes the WTO accession process rules and legally binding agreements. Stability is and identifies the main issues and challenges faced important both to domestic producers and to by acceding countries. exporters from other countries wishing to access these economies' markets.Adherence to WTO provi- sions--for example, by binding tariffs and by speci- Benefits of Membership fying conditions for foreign direct investment (FDI) There are three main benefits of WTO membership: in the services agreement--would improve the effi- (a) strengthening of domestic policies and institu- ciency and productivity of acceding countries. tions for the conduct of international trade in both WTO membership also offers the opportunity for goods and services, which is required before acces- new members to lock in existing, relatively liberal sion into the WTO can be accomplished; (b) trade regimes. Although the trade regimes in acced- improvements in the ease and security of market ing economies vary considerably, many have estab- access to major export markets; and (c) access to a lished regimes with relatively low tariffs and no dispute settlement mechanism for trade issues. significant formal nontariff barriers. For these countries, membership provides the opportunity to lock in these regimes by assuming legally binding Policies and Institutions obligations regarding tariff levels. This not only per- Although there are significant differences in the insti- mits them to enjoy the benefits of liberal trade but tutional and policy environment of the various coun- also gives them a first line of defense against the 61 T H E W O R L D T R A D E O R G A N I Z AT I O N domestic protectionist pressures that are present in cies and institutions that have a bearing on the con- all market economies. duct of international trade; (b) the members' fact- finding phase; and (c) the negotiation phase. The last two phases, while conceptually separate, tend to Market Access overlap in practice. Throughout, the applicant is Two main dimensions of market access are of faced with meeting WTO requirements and provi- importance to acceding economies. The first is the sions, as well as demands by existing members. With extension of permanent and unconditional most- very few exceptions, negotiation is in one direction favored-nation (MFN) status, which comes with only: the applicant is asked to demonstrate how it WTO membership. At present, economies that are intends to meet the existing WTO provisions--it not members of the WTO have been granted MFN cannot change them. Existing members can ask the treatment voluntarily by major trading partners, applicant to reduce the level of protection in its but there is nothing to guarantee that they will con- markets, but the reverse does not usually occur. tinue to be accorded such treatment. For example, in the United States extension of MFN to Russia and The Formalities several other economies in transition is contingent on the economies' adherence to the provisions of After a country sends a letter to the Director-Gener- the Jackson-Vanik amendment to the 1974 Trade al of the WTO expressing its desire to accede to the Act regarding freedom of emigration.1 The second organization, the request is considered by the WTO point is the substantial evidence that the incidence General Council, which consists of representatives of antidumping actions (both investigations and of all members and which meets frequently during definitive measures) is much higher against non- the course of the year. The General Council routine- WTO members than against members. ly decides to set up a working party, with appropri- ate terms of reference, to consider the accession application, and it nominates a chairman of the Dispute Settlement working party.2 Membership in the working party is Access to an impartial and binding dispute settle- open to all members of the WTO. In the case of ment mechanism, the decisions of which have a sig- applications by large countries such as China or nificant chance of being enforced, is an important Russia, many countries participate; in the case of potential benefit for the acceding economies, many smaller countries, the working party is usually made of which are small and heavily dependent on inter- up only of the "Quad" (Canada, the European national trade. The WTO's dispute settlement Union, Japan, and the United States) plus a number mechanism has proved successful in providing of other members, including neighboring countries opportunities for members to obtain satisfaction that are significant trading partners of the appli- regarding grievances stemming from practices of cant. The formalities phase can be quite short--no other members that cause trade injury. Although more than a few months. developing countries face some problems in access- ing this mechanism, membership provides an The Memorandum opportunity that, with proper assistance, can be beneficial to new members, especially in their rela- The preparation of the memorandum on the for- tionships with large trading partners. eign trade regime by the applicant explaining its policies and institutions can be a demanding task because of the range of issues that the memoran- The Accession Process dum has to address and the degree of detail The process of accession to the WTO is demanding required. The issues include much more than sim- and lengthy. It can be divided into an introductory ply trade in goods and services, although describing phase of formalities and three substantive phases. the trade regime for services, which encompasses The three substantive phases are (a) the applicant's the financial sector, insurance, telecommunications, preparation of a memorandum on the foreign trade professional services, and the like, is a large task in regime (hereafter referred to as the "memoran- itself. Relevant subjects also include various aspects dum"), which describes in detail the country's poli- of foreign exchange management and controls, 62 WTO Accession investment and competition policy, protection of ent rules and criteria, as opposed to administra- intellectual and other property rights, and enter- tive discretion. A key issue for enterprises that are prise privatization. The preparation of the memo- expected to remain state owned is whether they randum is solely the responsibility of the applicant, operate under market conditions or enjoy special and so is any delay in its preparation. monopoly rights and privileges. Even if the original memorandum is prepared · Some issues relate to the jurisdiction and capacity quickly, if it is incomplete in its details or if the leg- of national agencies to implement policies on islation and practices described are inconsistent which commitments are being made. The funda- with WTO provisions, the subsequent question- mental concern is one of governance: do the and-answer period can be protracted. At times, agencies have the authority and capacity to imple- members have asked the WTO secretariat to review ment the commitments that they are making in draft memoranda before their circulation to prevent the context of WTO accession regarding the laws incomplete documentation from being disseminat- and regulations that affect the conduct of interna- ed. The secretariat, however, assumes no responsi- tional trade? A related concern has to do with the bility regarding the contents of the memorandum. role and jurisdiction of local authorities and whether they have the right and opportunity to nullify commitments made by the national Questions and Answers authorities in the context of accession negotia- Once the memorandum has been circulated to tions. WTO members, the accession process enters the second stage, in which members ask questions and Negotiations obtain clarifications on the applicant's policies and institutions. This typically takes several months. (In At some point during the question-and-answer the case of Russia, it took more than a year.) The phase--after most, but frequently not all, the points working party usually does not meet until the mem- raised by working party members have been orandum and the initial questions and answers have answered--the applicant is requested to submit its been distributed. so-called initial schedule of offers in goods and The purpose of the detailed review that takes services. This consists of (a) the detailed schedule of place during this phase and that may involve several tariffs the applicant proposes to impose on goods working party meetings is to make sure that the leg- and the level at which the tariffs are "bound," and islation and institutions of the applicant are in con- (b) the commitments it makes (and the limitations formity with WTO provisions. The applicant is it sets) on providing access to its market for ser- requested to submit for the consideration of the vices.3 In addition, the applicant is requested to working party members relevant legislation on a make commitments regarding the level of support it variety of issues covered by the WTO. Delays during plans to provide to its agriculture in relation to a this phase are frequent; if a member feels that the base reference period (usually three representative answers submitted to a question or the actions years before the application for accession), as well as taken to remedy an inconsistency are inadequate, it other aspects of its support for agricultural trade, simply resubmits the question for the next round. such as export subsidies. Although the issues raised in each accession Once these offers are tabled, the accession process working party vary somewhat depending on the enters its final phase, which involves specific bilater- country, some common themes emerge in the dis- al negotiations between the applicant and each cussions of accession, especially, but not exclusively, WTO member that wishes to hold such talks in the case of countries in transition. regarding the tariff level or the degree of openness of the services sector proposed by the prospective · Within the context of laws and the operations of member. The actual timing of the original offers government institutions, two broad issues typi- varies considerably, and sometimes they are tabled cally receive special attention: the degree of priva- very early in the question-and-answer phase, as tization in the economy, and the extent to which happened, for example, in the case of Georgia. government agencies involved in the regulation of Often, bilateral negotiations take place in parallel economic activity do so on the basis of transpar- with formal meetings of the working party that 63 T H E W O R L D T R A D E O R G A N I Z AT I O N continue to deal with questions and answers regard- General Council (usually a formality), the country ing the foreign trade regime. The negotiations phase is invited to sign a protocol of accession. can also be lengthy, depending on the degree of openness the applicant proposes and the demands Progress in Accession for market access made by members. When these negotiations are in the process of As of May 1, 2001, 28 working parties had been for- being finalized and the applicant has provided mally established to consider the accession applica- assurances that the legislation and institutions that tions of prospective WTO members (Table 8.1). Most would permit compliance with WTO provisions are of the working parties were established some time in place, a draft report on accession, including the ago, with the oldest, those for Algeria and China, dat- schedule of agreed commitments on goods and ing from 1987. With few exceptions (that of Algeria services, is prepared by the secretariat for considera- being one) most of the working parties are active. In tion by the working party. After approval by the practice, accession has taken, on average, a little more working party, the report is forwarded to the Gener- than five years, from the establishment of the work- al Council. Following a favorable decision by the ing party to entry into force of WTO membership. Table 8.1 Accessions to the World Trade Organization as of May 1, 2001 Establishment of Tariff Services Draft working Economy working party Memorandum offers offers party report Algeria 06/87 07/96 -- -- -- Andorra 10/97 02/99 09/99 09/99 -- Armenia 12/93 04/95 01/99 10/98, 07/99 08/99 Azerbaijan 07/97 04/99 -- -- -- Belarus 10/93 01/96 03/98 05/99 -- Bosnia and Herzegovina 4/00 -- -- -- -- Bhutan 10/99 -- -- -- -- Cambodia 12/94 06/99 -- -- -- Cape Verde 07/00 -- -- -- -- China 03/87 02/87, 09/93 04/94 09/94, 11/97 12/94, 05/97 Kazakhstan 02/96 09/96 06/97 09/97 -- Lao PDR 02/98 -- -- -- -- Lebanon 04/99 06/01 -- -- -- Macedonia, FYR 12/94 04/99 -- -- -- Nepal 06/89 02/90, 09/98 -- -- -- Russian Federation 06/93 03/94 02/98 10/99 -- Samoa 07/98 -- -- -- Saudi Arabia 07/93 07/94 09/97, 06/99 09/97, 06/99 -- Seychelles 07/95 08/96 06/97 05/97 -- Sudan 10/94 01/99 -- -- -- Taiwan (China) 09/92 10/92 02/96, 08/99 09/94, 08/99 08/98 Tonga 11/95 05/98 -- -- Ukraine 12/93 07/94 05/96 02/98, 06/98 -- Uzbekistan 12/94 09/98 -- -- -- Vanuatu 07/95 11/95 11/97, 05/98 11/97, 11/99 11/99 Vietnam 01/95 09/96 -- -- -- Yemen 07/00 -- -- -- -- Yugoslavia, FR 01/01 -- -- -- -- -- Not yet done. Source: World Trade Organization. 64 WTO Accession A number of economies are at an early stage in result in long delays while governments ascertain the negotiation process. They include Azerbaijan, consistency between existing legislation and regula- Bosnia and Herzegovina, Cape Verde, FYR Macedo- tions and WTO requirements and while they design nia, and Yemen. Several others, such as Armenia, and put in place the amendments or new legislation Belarus, Kazakhstan, Russia, and Ukraine, have been or regulations needed. This is compounded by the involved in the accession process for a long time reality that legislative processes are themselves but, for various reasons, are not yet close to com- lengthy. The WTO secretariat can be of assistance pleting it. The process was finally completed for only in a very limited way in the accession process, China and Taiwan (China) by the end of 2001. as the WTO budget allocates very few resources to accession of new members. The five staff members in the WTO Accession Division are thinly stretched Why Does It Take So Long? in servicing even the procedural needs and paper- To understand why accession to the WTO is such a work generated by more than 25 active accession long process, it is necessary to look first at the vari- working parties. ous phases of the accession process and the reasons why delays may occur. Difficulties in the Negotiations Phase The negotiations phase can be and frequently has Weak Follow-Up been the most time-consuming phase of accession. In several cases a government has taken the initial Negotiations partly have to do with whether the step to apply for accession and have a working party acceding member's policies and institutions are set up, but then it fails to follow up the accession consistent with various aspects of the WTO agree- process by preparing a memorandum on its policies ments and partly with the specific tariff bindings or taking subsequent steps, or it does so after a long and commitments in agriculture and services. interval. Working parties for Uzbekistan and Sudan Delays can occur on both sides. The acceding gov- were set up in 1994, but the memoranda of foreign ernment may be unwilling to make needed liberal- trade policy were only submitted in September ization commitments--for example, it may not 1998, for Uzbekistan, and January 1999, for Sudan. offer to liberalize nontariff barriers, or it may pro- pose binding tariffs at levels much higher than existing ones. Members, for their part, may not be Political Issues satisfied with the level of liberalization proposed or In a few cases political issues between an applicant may be unwilling to accept delays in bringing the and one or more influential WTO members have laws and institutions of the applicant into conform- introduced delays. This happened in the past to ity with WTO provisions. Sometimes, as in the cases some extent with the accessions of China (and of of Albania, Croatia, Estonia, and Latvia, delays have Taiwan, China, which is linked to it) and of FYR stemmed not from the WTO accession process as Macedonia. such but from disagreements between the European Union (EU) and the United States over the commit- ments of acceding countries in the WTO (for exam- Inherently Time-Consuming Processes ple, in audio-visual services) and the possible future Even if the above problems did not exist, accession association of these countries with the EU. is inherently a time-consuming process. The prepa- ration of the memorandum presents serious diffi- Strategy and Tactics of the Applicant Country culties for governments that typically do not have sufficient human or material resources to address Within the rules and disciplines of the WTO, each the issues that have to be discussed in detail. Most country has considerable scope as to how restrictive countries have had to seek assistance from outside or liberal its trade regime will be. There are no spe- experts funded by bilateral aid agencies, from the cific rules as to the maximum level at which a coun- WTO itself, and from the World Bank. try has to bind its tariffs, how many services it will The question-and-answer process is also time liberalize, whether to establish antidumping legisla- consuming; applicants' institutional weaknesses tion, or how fast to liberalize its agricultural trade. 65 T H E W O R L D T R A D E O R G A N I Z AT I O N Countries thus have a strategic choice to make dur- duce uncertainty about trade policy among the ing the negotiations phase: how liberal their trade country's trading partners. regime will be, consistent with overall WTO disci- Several transition countries that have recently plines. become WTO members, such as Albania, Estonia, A strategy that some countries have pursued in Georgia, the Kyrgyz Republic, Latvia, and Mongolia, their accession negotiations is to try to liberalize as pursued a different strategy. In most respects their little as is necessary to ensure accession. Since such governments adopted a liberal trade strategy as part applicants cannot negotiate significant improve- of the process of accession.4 This entails (a) binding ments in their access to other markets, they try to tariffs at the usually low currently prevailing levels maintain significant levels of protection to use as or agreeing to reduce and bind tariffs at low levels as bargaining chips for obtaining improved access in part of the accession negotiations; (b) agreeing to a future negotiating rounds. Some of the countries liberal trade regime in agriculture and services; and that are using this strategy, such as China and Rus- (c) at an early date after accession, participating in sia, also feel that significant levels of protection are such agreements as the government procurement necessary during a transition period when ineffi- code, which increases competition and transparen- cient state-owned enterprises are being restructured cy in the operation of their markets. (see Gabunia 1998). These countries have typically The fundamental benefits of such a strategy are presented initial offers that propose to bind tariffs at economic: these countries reap the benefits of liber- rates much higher than those currently applied. al trade and investment. But the strategy has a num- Similar issues arise in services. Many transition ber of other advantages as well: it tends to facilitate and developing economies feel that their services negotiations for accession; it provides governments sectors are underdeveloped and would like to limit with political cover against domestic protectionist the commitments they make to open these sectors interests that may otherwise succeed in subverting to foreign competition. This is especially an issue in an existing liberal trade regime; and the legally such areas as financial services and telecommunica- binding WTO commitments lock in reforms by tions, in which countries frequently face requests making it more difficult for future governments to from WTO members to establish liberal policies reverse the liberalization. Increased protection to regarding commercial presence. Such policies "safeguard" against serious injury to domestic would permit foreign services suppliers to establish industry is permitted under WTO rules, but it is subsidiaries or joint ventures based on the principle based on a detailed and transparent investigation to of national treatment, which prohibits discrimina- demonstrate injury, which is then notified to the tion against foreign services providers and thus has WTO and subjected to the scrutiny of other mem- a direct bearing on foreign direct investment. bers. This is far more difficult than for a powerful There are significant dangers to a "minimum lib- domestic industry to simply seek government sup- eralization" accession strategy. Individual countries, port for raising tariffs beyond the applied level but especially small developing economies, have little below the higher bound level, which a government leverage in market access negotiations, and so the can do almost without any constraint. The point potential benefits they may be able to obtain about the WTO is not that it prohibits protection through such a strategy may be very small. At the but, rather, that it permits it only according to cer- same time, maintaining protection through rela- tain rules; obeying these rules makes protection tively high tariffs and protected agriculture and more transparent as well as more difficult to initiate services sectors imposes costs on the applicants' and expand. own economies: they forgo the benefits of a more The Chinese accession, concluded in 2001, has liberal trade regime, which, in the first instance, combined elements of both strategies and has raised accrue to the country itself. If countries bind tariffs a number of additional issues. First, China has used at levels higher than those applied and assume few the process of WTO accession to stimulate and commitments regarding agriculture and services make irreversible substantial trade liberalization (both of which are possible under WTO rules), they and more broadly based reforms. Second, China, in are subject to another risk: they create an opening many ways an economy in transition, considers for domestic interests to exert political pressure for itself a developing country and has been seeking to additional protection in the future, and they pro- obtain transition periods and other special and dif- 66 WTO Accession ferential treatment that WTO agreements extend to WTO agreements. When such weaknesses are developing countries. The latter includes nonreci- brought out in negotiations, members suggest that procity, preferential market access, and different the applicant seek technical assistance, available commitments and time limits in the implementa- from a variety of bilateral and multilateral donors, tion of the provisions of various aspects of the and that it present a detailed plan regarding the par- agreements, ranging from agriculture to subsidies ticular aspects of the relevant WTO provisions in and trade-related intellectual property rights which weaknesses exist and how and within what (TRIPS). China, because of its position as a large time period it proposes to remedy them. market, has also bargained on certain aspects of UNCTAD, the World Bank, the European Union, market access, such as on textiles and on issues Switzerland, and the United States, as well the related to its designation as a nonmarket economy. WTO, have programs that provide technical assis- tance on various aspects of the accession process, especially in the preparation of the initial country Attitudes and Policies of WTO Members memorandum. Anecdotal evidence about these The demands made on newly acceding countries are programs suggests a somewhat uneven perfor- greater than WTO disciplines on existing members. mance. Most countries report very helpful contri- Based on recent accession experience, the areas dis- butions by foreign consultants and advisers in the cussed below are ones in which members typically preparation of the memorandum. In some cases, request that acceding countries make more far- however, it appears that advice provided by outside reaching commitments than those made by many experts has actually slowed the accession process existing members at similar levels of development. because the consultants suggested, and the country agreed to, a "bargaining" strategy of tariff binding at Tariffs. Acceding countries are requested to bind all high levels and limited offers on services. In addi- tariffs, whereas many developing countries contin- tion, there have been problems of coordination ue to have a large portion of their tariff schedule among the various donors, as well as between the outside agriculture unbound. Ceiling bindings have bilateral aid agencies providing the assistance and been accepted, but there is pressure to bind close to their colleagues in the trade ministries who negoti- applied rates. ate the accession. Agriculture. In addition to binding the tariff sched- Plurilateral Agreements. There is pressure for ule, commitments are expected on aggregate mea- countries to begin examining the provisions of the sures of support (AMS), export subsidies, and the plurilateral agreements (for example, on govern- like. Since many acceding countries did not provide ment procurement and civil aviation) at the time of substantial support to agriculture but, rather, penal- accession and to commit to a timetable for complet- ized it, the requests they face for reductions in AMS ing negotiations soon after accession. may not be warranted, and in any case meaningful calculation of commitments in this area is subject to "Market Economy" Issues. Although there is no serious statistical difficulties. explicit requirement in the WTO agreements that a member have a market economy, a requirement Rules and Disciplines. Acceding countries are typi- that acceding countries have, fundamentally, such cally requested to meet all commitments at entry an economy is being pushed de facto by existing with regard to, for example, TRIPS, customs valua- members as part of their leverage in the accession tion, standards, and SPS regulations, without time process.5 The pressures have been felt by all acced- limits such as those available to existing members at ing countries, including China, where explicit similar levels of development, and regardless of understandings were reached with regard to the whether institutional weaknesses make it difficult existence of state trading in specific sectors. At the for them to fulfill such commitments. Such weak- same time the Quad countries have been unwilling nesses relate broadly to the operations of a market to modify their own antidumping procedures economy; it takes time to establish the institutional regarding the designation as "nonmarket infrastructure that would enable the applicants to economies" of transition countries that have discharge their responsibilities properly under the become WTO members. Under this designation, 67 T H E W O R L D T R A D E O R G A N I Z AT I O N different, less transparent, and potentially discrimi- It is politically difficult to adopt a liberal trade natory practices can be applied in the determina- strategy at accession, especially when major trading tion of whether dumping has occurred and, in the partners, which are WTO members, take advantage case of EU safeguard actions, against imports from of opportunities that are perfectly legal under the a number of these countries, including all the mem- WTO to limit market access--for example, by bers of the Commonwealth of Independent States maintaining high levels of protection in agriculture. (CIS) and China (Michalopoulos and Winters Even recognizing the political difficulties involved, a 1997). For this reason, the nonmarket economies strong argument can nonetheless be made that if label has been a major cause of trade friction developing countries and transition economies cur- between many transition economies, on the one rently applying for WTO accession adopt a liberal hand, and the United States and the European trade strategy at entry, they will maximize the bene- Union, on the other. fits and opportunities for integration in the interna- Legal justification for using such procedures can tional community that WTO membership offers be found in GATT provisions that permit different and will accede more quickly, as well. treatment "in the case of imports from a country It is fair to ask whether countries should not which has complete or substantially complete maintain some flexibility in their initial offers, as monopoly of its trade and where all domestic prices they are bound to face demands to liberalize by are fixed by the State" (Palmeter 1998: 116). These existing members almost irrespective of the level of practices were perhaps fully justified when practical- protection they initially propose. Although there is ly all trade was controlled by state trading enter- merit in this point, it probably should not be prises or ministries and prices were fixed by the pushed too far. Experience in recent accession nego- state. Countries in transition, however, have made tiations suggests that countries which make initial great progress in introducing market forces in recent offers to bind their tariffs at levels significantly dif- years. It would be difficult to argue that, say, China ferent from the applied level encounter serious diffi- or Russia has at present "a substantially complete culties in accession--even though the practice is monopoly on trade" or that all domestic prices in widespread among existing developing country these countries are fixed by the state. Thus, continu- members, many of which have not bound large por- ation of the traditional EU and United States tions of their tariff schedule. When such an initial antidumping practices no longer appears justified in offer is put on the table (as has happened with sev- the new setting (Michalopoulos and Winters 1997).6 eral countries of the former Soviet Union, as well as Because the GATT antidumping provisions accept other applicants for accession), working party national legislation and practices as decisive, the odd members basically refuse to consider it or to enter situation can arise in which countries become new negotiations on that basis. They simply ask the WTO members but are still designated as nonmar- country to submit a revised offer with bound rates ket economies for antidumping purposes. closer to the applied ones before serious negotia- tions take place. China's accession is unique, for both political and Lessons of Experience and Issues for the economic reasons, and lessons from it have to be Future drawn with extreme care. Undoubtedly, China has The first important lesson of experience is that each used WTO accession to promote and lock in wide- accession case involves a different negotiation, with ranging reforms. China differs from most recently different dynamics. This makes it difficult to gener- acceding countries in that it has been able to negoti- alize. Nevertheless, the cases of a number of small ate a number of transition periods--for example, countries that have recently concluded the accession for eliminating quantitative restrictions, licensing, process suggest that the smaller the country and the and state trading--as well as the maintenance of more liberal its regime, the faster the accession tariff quotas in agriculture. It probably has much process. There are two reasons for this: smaller more bargaining power than all the recently acced- countries realize that the costs of protection are ing countries taken together. It is a moot point high for them, and the small size of their economies whether the time limits and extensions obtained by poses fewer market access issues for major WTO China (which were much less than it requested) are members. compatible with its economic interests or whether a 68 WTO Accession faster liberalization of its trade regime would have complex, and challenging for all countries, especial- been more conducive to the country's longer-term ly the LDCs. The process is inherently time consum- development. To WTO members negotiating ing, but there are a number of steps that acceding China's accession, it is almost irrelevant what Chi- countries and WTO members could take which nese protection does to China's economy; they are would facilitate and expedite accession. concerned about the impact of such protection on Governments seeking accession need, first, to their exports to the Chinese market and its large establish a central coordination point to provide potential. direction and manage the multiplicity of legislative At the same time, China has had to accept limita- and regulatory changes in their foreign trade regime tions on its market access that other developing that are necessary for accession. Second, they need countries have not. For example, it has agreed to be to adopt liberal trade policies, which will both con- subject to product-specific selective safeguards; it tribute to their effective integration into the inter- accepted three more years of restrictions in the national economy and facilitate WTO entry. Third, implementation of the Agreement on Textiles and governments need to focus on and identify those Clothing (ATC), from which it had been completely areas of the WTO agreements in which weaknesses excluded; and it agreed to be designated a nonmar- in their institutional infrastructure require that they ket economy for 15 years. In this context, the more delay implementation of WTO provisions. They exceptions and the longer transitions a large coun- should actively solicit technical assistance, prepare a try such as China seeks to obtain, the more WTO realistic plan for implementing remedial actions, members will strive to maintain provisions that and seek agreement to obtain suitable delays in the inhibit full access of the acceding country's prod- implementation of the agreements as part of the ucts in their markets. accession process (despite the apparent reluctance It can be argued that WTO members' insistence of members to agree to such extensions so far). on a liberal commercial policy at entry is likely to WTO members can also take steps to help expe- serve the acceding countries' long-term develop- dite the accession process. It is in their interest that ment interests, as well as WTO members' commer- the organization achieve universal membership cial objectives. But insistence on adherence to all sooner rather than later, as current members would WTO commitments at entry and without transition benefit if all countries adhered to the rules and pro- periods in areas (such as customs valuation, TRIPS, visions of the WTO. In this regard, members should standards, and SPS) where there are obvious institu- attempt to ensure that accession is not delayed on tional weaknesses in LDCs and transition account of high-income countries' own disagree- economies raises serious problems. That is, acced- ments or disputes. ing countries, motivated by their strong desire for WTO members also need to consider the institu- membership, may agree to obligations that they tional weaknesses of acceding governments and to cannot implement, leaving them open to subse- moderate their demands by agreeing to suitable, quent complaints. Alternatively, providing generous time-bound extensions in meeting WTO obliga- transition periods at a time when the transition tions. This should not mean lowering the require- periods for other countries that are already mem- ments but, rather, allowing more time to meet bers are expiring would create inequities between them. If such extensions are not provided, either the existing and new members. The solution to this negotiations become stalled or the acceding country problem may be the substantial extension of some ends up accepting obligations that it cannot imple- of these transition periods both for existing low- ment. In particular there is merit to extending and income members of the WTO and for acceding standardizing transition periods for acceding coun- LDCs and transition economies; after all, these tries in the areas of standards, TRIPS, and custom transition periods were arbitrarily set in the first valuation, where countries invariably face serious place. constraints in meeting requirements at accession. Industrial countries should continue to provide assistance to developing countries and countries in Conclusion transition that are not members in order to For a variety of reasons, the process of WTO acces- strengthen their institutional capacities so that they sion has been and is likely to continue to be lengthy, are better able to meet the requirements for WTO 69 T H E W O R L D T R A D E O R G A N I Z AT I O N accession. Such assistance needs to be better coordi- ance and receive "permanent" conditional MFN treatment. nated. The Integrated Program of Trade-Related When, however, Georgia, the Kyrgyz Republic, and Mongolia became WTO members, the United States exercised its right of Technical Assistance to the Least-Developed Coun- nonapplication under WTO Article XIII; that is, it did not pro- tries has the potential to benefit a number of acced- vide these countries with unconditional MFN status and thus, ing countries. Consistent with preserving the WTO de facto, had not accepted their accession. Subsequently, leg- as a member-driven institution, industrial country islation was enacted permitting the United States to notify the members should also consider substantially increas- WTO that it has accepted these countries' membership. In the ing the resources available to the WTO secretariat case of China, the United States had to address this important issue through amended legislation before membership negoti- for assisting acceding governments in the prepara- ations were concluded. tion of the original memorandum and in the design of legislation and regulations that would enable the 2 Usually the chairman is an ambassador, a permanent represen- tative to the WTO. Countries often request and obtain observ- applicants to meet WTO obligations. Channeling er status at the WTO to familiarize themselves with the more resources through the WTO would permit the institution before they make a formal request for accession. secretariat to play a more active role in coordinating 3 Services commitments are typically more general and open- assistance efforts in support of accession and would ended than commitments in the sphere of goods. For a discus- give greater assurance that the outside experts who sion, see Part IV of this volume. assist governments with the preparation of the 4 The trade regimes in Croatia and Jordan, which also acceded needed documentation and the modification of leg- recently, were somewhat less liberal. islation and regulations do so in ways that more effectively meet WTO requirements. A reasonable 5 GATT Article XVII calls for notification of enterprises engaging objective would be to cut the processing time of in state-trading practices. The article, however, was never intended to address problems that come up when the bulk of accessions to no more than two years, a time frame external trade is controlled by the state. Indeed the GATT that is feasible if the above steps are taken. If that accommodated several countries, such as Romania and schedule were attained by all acceding countries, the Czechoslovakia, that at the time had centrally planned WTO would be able to achieve universal member- economies. ship in the next five years--a worthwhile objective 6 In 1997 the European Commission announced proposals for for the international community. liberalization of EU policy on this issue vis-à-vis Russia and China, which would terminate their designation as nonmarket Notes economies at the country level and would permit determina- tions to be made case by case, taking into account the market 1 Belarus is subject to annual waivers (as was China, until recent- conditions prevailing in each commodity in which dumping ly); the other countries have been found to be in full compli- had been alleged (Croft 1997). This is similar to U.S. practice. 70 9 VA L E N T I N A D E L I C H Developing AB report. In these instances, only "negative" consensus can stop the process; that is, all Countries and members must agree not to pro- ceed or not to adopt panel and the WTO Dispute AB recommendations or rul- ings. This reversal of the consen- Settlement System sus rule led to a radical change in the dynamics of dispute set- tlement, making it more auto- matic and less dependent on the power of the countries involved in a dispute. Since there is an W extensive literature comparing ith the creation of the WTO, devel- the GATT and WTO systems, we confine ourselves oping and industrial countries to briefly summarizing the salient features of the became subject to the same set of rules and to simi- DSU before turning to developing country experi- lar commitments. A new Dispute Settlement ence and concerns.1 Understanding (DSU) was negotiated to enforce multilateral disciplines. The DSU is widely regarded The WTO Dispute Settlement System as one of the positive outcomes of the Uruguay Round, marking a move toward a more "automatic" The Dispute Settlement Body (DSB), which com- and rule-oriented system (Jackson 1997). This prises all WTO members, has the authority to estab- chapter evaluates the functioning of the DSU from a lish panels, adopt panel and AB reports, maintain developing country perspective. surveillance of the implementation of rulings and Although the cornerstone of the WTO dispute recommendations, and authorize suspension of settlement mechanism remains Articles XXII and concessions and other obligations under WTO XXIII of the GATT, the DSU brought about a sub- agreements (Art. 2 DSU). A member must first stantial change in the workings of the system. A request bilateral consultations if it considers that a major improvement was to remove the consensus benefit accruing to it directly or indirectly under the requirement at key stages of the process. The DSU WTO agreements is being nullified or impaired states that "where the rules and procedures of this (Art. 4 DSU). If consultations fail to settle the dis- Understanding provide for the Dispute Settlement pute, the complaining party may request the estab- Body to take a decision, it shall do so by consensus," lishment of a panel, which must be created unless but this general rule does not apply to the establish- the DSB decides by consensus not to do so (Art. 6 ment of a panel of experts, the adoption of its DSU). report, or, if the report was subject to an appeal A panel is generally composed of three panelists. before the Appellate Body (AB), the adoption of the Its deliberations are confidential, and the opinions 71 T H E W O R L D T R A D E O R G A N I Z AT I O N expressed in the panel report by individual panelists may seek to suspend concessions under another are anonymous (Arts. 6, 14 DSU). Nationals of covered agreement (Art. 22 DSU). countries that are parties to the dispute may not Suspension of concessions (retaliatory action) is serve on a panel unless the parties to the dispute the last recourse for countries in enforcing compli- agree otherwise. Panels must conduct examinations ance with DSB recommendations and rulings. Of within six months (Art. 12 DSU). Within 60 days of course, effective retaliatory measures are available to the date of circulation of a panel report to WTO countries with economic power. As noted by Hoek- members, the report must be adopted at a DSB man and Mavroidis (2000: 531): meeting unless a party to the dispute formally noti- fies the DSB of its decision to appeal or the DSB Those WTO members that can afford to either decides by consensus not to adopt the report (Art. take countermeasures or to incur the costs of 16 DSU). action being taken against them are in a better The Appellate Body, a standing tribunal created in position. When acting as complainants they the Uruguay Round, considers any appeals. The tri- will use threat and/or imposition of counter- bunal consists of seven members, of whom three measures in order to induce compliance; when serve on any given case. The members are appointed acting as defendants, they will have at least the for four years and may not be connected with any luxury of weighting the pros and cons between government. Appeals are limited to issues of law changing the domestic policies at stake (in covered in the panel report and to legal interpreta- order to avoid imposition of countermeasures) tions developed by the panel. AB proceedings are or simply keeping the domestic policies at stake not to exceed 60 days and are confidential. The intact (and see countermeasures imposed reports are drafted in the absence of the parties to against them). the dispute, and the opinions expressed in them are anonymous (Art. 17 DSU). When a panel or the AB A key principle, nevertheless, is that multilateral concludes that a measure is inconsistent with a cov- authorization of retaliatory action is required. In ered agreement, it must recommend that the mem- this regard there is a "sequencing problem." Article ber concerned bring its measures into conformity 22.6 prescribes that retaliation must be authorized with the WTO agreement (Art. 19 DSU). within 30 days of the time a country is supposed to Article 21.5 of the DSU states that "where there is comply with a WTO ruling. This deadline, however, a disagreement as to the existence or consistency does not allow enough time for completion of a with a covered agreement of measures taken to compliance review under Article 21.5.2 Valles and comply with the recommendations and rulings, McGivern (2000) conclude that three different such dispute shall be decided through recourse to precedents have been established for determining these dispute settlement procedures, including the consistency of implementing measures and the wherever possible resort to the original panel." In suspension of concessions: the Bananas model, in turn, Article 22 of the DSU foresees that "if no satis- which the arbitrators first determined the WTO- factory compensation has been agreed, (. . .) any consistency of the implementing measures before party having invoked the dispute settlement proce- assessing the level of the suspension of concessions; dures may request authorization from the DSB to the Salmon model, in which the parties provided for suspend the application to the member concerned "sequencing" on an ad hoc basis; and the Subsidies of concessions or other obligations under the cov- and Countervailing Measures (SCM) model, in ered agreements." The complaining party should which the parties used a provision of the SCM first seek to suspend concessions or other obliga- agreement to extend Article 22 retaliation deadlines. tions with respect to the same sector as that in In each case, the complaining party requested the which the panel or the Appellate Body has found a establishment of an Article 21.5 panel on the basis violation or a nullification or impairment of bene- of a bilateral agreement to extend the Article 22 fits. Then, if the party considers that such action is deadlines until after completion of the Article 21.5 not practicable or effective, it may seek to suspend review. concessions or other obligations in other sectors A number of provisions in the DSU relate to under the same agreement. Finally, if the circum- developing countries. Article 4.10 calls for members stances are serious enough, the complaining party to give special attention to the particular problems 72 Developing Countries and the WTO Dispute Settlement System and interests of developing countries in consulta- Developing Countries' Experience with the tions, and Article 12.10 allows for the extension of WTO Dispute Settlement System the consultation period in cases of measures taken by developing countries if the parties agree. Article Most of the clauses in the DSU regarding develop- 8.19 provides that a developing country involved in ing countries have proved to be more declarative a case can request that the panel include at least one than operative. For instance, the concept, in Article person from a developing country, and Article 12.11 4.10, of giving "special attention" to the particular states that in such cases the panel report is to indi- problems and interests of developing countries cate how account was taken of relevant provisions during consultations has no operative content and concerning differential and more favorable treat- has not been developed in panel or AB reports. ment for developing countries that are embodied in Although in one case this article was mentioned in the WTO agreements referred to in the dispute. a DSB meeting to support a developing country If a case is brought by a developing country, the position, there was no substantive discussion of the DSB, in considering what appropriate action might "special attention" concept. A similar problem be taken, is to take into account not only the trade arose with special and differential treatment (S&D) coverage of the measures complained of but also clauses in such agreements as that on antidumping their impact on the economy of the country con- (see Box 9.1). Although some panels have dealt cerned (Art. 21.8 DSU). Article 27.2 provides for with S&D clauses, given that these clauses have neutral legal advice (technical assistance) to be fur- been invoked in fewer than 10 cases involving nished by the WTO secretariat to developing country developing countries, it seems that they are not members. Finally, Article 24.1 calls for due restraint very relevant for these countries, either in defend- in invoking the DSU against least-developed coun- ing or in claiming their rights. 3 (See Chapter 49, by tries (LDCs), in asking for compensation, or in seek- Oyejide, in this volume; see also Whalley 1999.) ing authorization to suspend the application of DSU provisions related to the surveillance of concessions or other obligations to these countries. implementation of DSB recommendations and rul- B O X 9 . 1 S P E C I A L A N D D I F F E R E N T I A L T R E AT M E N T A N D T H E D S U : S O M E E X A M P L E S The Scallops Case cles 3(7), 4(2) and 4(5) of the DSU. This In 1995 Chile and Peru requested that a panel be request had been disregarded by the Com- established on the trade description of scallops munities thus discriminating against and drawn up by the European Communities (EC). impairing Chile's interests in deviation from The EC asked that the item be removed from the the provisions of Article 4(10) of the DSU DSB agenda, arguing that the time periods for which stated that Members "should give spe- consultations and for the inclusion of those items cial attention to the particular problems and on the agenda stipulated in the DSU had not interests of developing country Members." This been respected. According to Chile, however, was a discrimination against Chile which was not being granted the same treatment as the Communities did not take into account Canada and was not in conformity with the that consultations with Chile had already obligations of WTO members towards a been initiated when the Communities had developing country. agreed to Chile's participation in consulta- tions held with Canada on the same subject The Bed Linen Case. . . . [or that] [w]hen Canada-EC consultations In European Communities: Antidumping Duties on had been completed, Chile had requested Imports of Cotton-Type Bed Linen, India contended further consultations to resolve this matter in that the EC had not taken into account the spe- accordance with the letter and spirit of Arti- cial situation of India as a developing country. (continued) 73 T H E W O R L D T R A D E O R G A N I Z AT I O N B O X 9 . 1 ( C O N T I N U E D ) India asserted that the EC had not acted consis- provides procedural safeguards, and thus, tently with Article 15 of the Agreement on does not require any particular substantive Antidumping, which recognizes that "special outcome, or any specific accommodations regard must be given by developed country to be made on the basis of developing members to the special situation of developing country status. In the United States' view, country members when considering the applica- [Article 15] does not impose anything tion of anti-dumping measures" and calls for other than a procedural obligation to exploring constructive remedies before applying "explore" possibilities of constructive anti-dumping duties in instances where they remedies. The word "explore" cannot fair- would affect the essential interest of developing ly be read to imply an obligation to reach a country members. India asserted that the EC did particular substantive outcome; it merely not explore any such possibilities prior to the requires consideration of these possibilities. imposition of antidumping duties and did not react to detailed arguments from Indian The panel's view was that exporters pertaining to Article 15: "[D]espite repeated and detailed arguments by the Indian the imposition of a lesser duty or a price parties stressing the importance of the bed linen undertaking would constitute constructive and textile industries to India's economy, the EC remedies but we come to no conclusions failed to even mention India's status as a develop- as to what other actions might in addition ing country, let alone consider or comment on be considered to constitute constructive possibilities of constructive remedies." India sug- remedies as none have been proposed to gested that "such remedies may consist of, us. . . . In our view, Art. 15 imposes no obli- among others, the non-imposition of anti dump- gation to actually provide or accept any ing duties or undertakings." India rejected the constructive remedy that may be identified notion that any procedural mechanism, such as and/or offered. It does, however, impose simplified questionnaires or extensions of time, an obligation to actively consider, with an satisfied the requirements of Article 15. The EC open mind, the possibility of such a reme- agreed in principle and accepted that undertak- dy prior to imposition of an anti-dumping ings could be a remedy, but it argued that Indian measure that would affect the essential exporters did not offer undertakings within the interests of a developing country. time limits set by the EC regulation. The United States, a third party in this dispute, Source: WTO, WT/DSB/M/7 (scallops); WTO, argued that Article 15 WT/DS/141 (India). ings are too weak to imply any difference between transparent liaison between the state and industry the possibilities open to industrial and developing in order to obtain up-to-date information on trade countries. Article 21.7, however, mandates that problems in which developing countries have a when a matter is raised by a developing country, stake. Developing countries lack the high-level the DSB is to consider what further action might be expertise and resources to devote to such activities. appropriate to the circumstances. To date, this pro- International financing for training public officials, vision has not been used by a developing country, screening industrial countries' trade policies, and perhaps because a precondition is that the country building a network with other developing countries devote resources to analyzing and following cases. with the aim of jointly presenting cases could help This involves checking arguments, issues, and pos- address some of these problems. sibilities and comparing experiences and results; The technical assistance called for in Article 27.2 exploring new legal as well economic arguments; is provided by only a few consultants and is inade- and, domestically, building up an efficient and quate, given the large number of cases. In addition, 74 Developing Countries and the WTO Dispute Settlement System since the WTO secretariat must be impartial, its lat- ity. These proposals are supported by many devel- itude for helping developing countries with legal oping countries in particular, as there is a common strategic issues is limited. In this context, the Advi- concern regarding the costs associated with submit- sory Centre on WTO Law (described in Box 9.2) ting, pursuing, and defending cases and the scarcity could play an important role in helping developing of human resources for dealing with increasingly country governments present and pursue cases. complex issues. Venezuela has noted the need to increase the num- Finally, provisions related to least-developed ber of legal assistants to the secretariat to help countries have not been invoked at all because no developing countries and has called for the creation least-developed country has been involved in a dis- of a trust fund to establish alliances with private law pute, as a complainant or as a respondent. firms to augment developing countries' legal capac- B O X 9 . 2 T H E A D V I S O R Y C E N T R E O N W T O L AW Claudia Orozco ratification process, and the Advisory Centre On the side of the Third WTO Ministerial Confer- became operational in October 2001. ence, held in Seattle in 1999, ministers from 29 The Advisory Centre provides legal advice on WTO member countries signed an agreement WTO law to developing countries and to establishing the Advisory Centre on WTO Law economies in transition. This legal advice might (ACWL). The establishment of the ACWL is a con- take the form of advisory opinions on particular crete action toward addressing the needs of questions of law, analysis of situations involving developing countries for advice and training on trade concerns, or legal advice provided WTO law. The contractual nature of the WTO throughout a dispute settlement proceeding. In requires that members have a full understanding recognition of the differences among developing of the content and scope of their rights and obli- countries, the extent of the support to be provid- gations and that they are able to access the dis- ed will depend on the needs and requirements pute settlement mechanism. Otherwise, the of each member in each case. Examples might ever-growing complexity and breadth of the sys- include outlining the legal questions of a case, tem, coupled with the relative scarcity of special- drafting submissions, and commenting on drafts ized human resources in developing countries prepared by government officials. In addition, and the costs of specialized external legal coun- the ACWL holds regular in-house seminars on sel, would marginalize many members. jurisprudence for Geneva-based officials and To help address these needs, a legal aid facility regional yearly seminars for officials based in was proposed, with two goals: (a) training gov- capitals. Finally, and most important, the ACWL ernment officials in WTO jurisprudence, and (b) offers on-the-job training for government offi- providing specialized legal advice on WTO law, to cials in charge of a particular case and intern- include support throughout legal proceedings. ships for government lawyers responsible for The response was the establishment of the Advi- WTO issues. sory Centre as a small, independent international organization based in Geneva and open to all Note: Claudia Orozco was minister counselor at the WTO members. By March 31, 2000, the final mission of Colombia to the WTO between August 1994 date for becoming a founding member, the and July 2000. She served as legal counselor and was a treaty had been signed by 9 industrial and 22 panelist in several cases. On February 1998 she submit- developing countries. The 38 least-developed ted the project proposal for the Advisory Centre on countries that are members of the WTO are prior- WTO Law to the Netherlands, the United Kingdom, and ity beneficiaries of the ACWL's services. The Norway and led consultations with an informal group agreement entered into force in July 2001, after of WTO members. The result was a proposal that was the requisite number of countries completed the offered to all WTO members. 75 T H E W O R L D T R A D E O R G A N I Z AT I O N Developing Country Participation in matters ranging from patent protection under the the DSU [Trade-Related Aspects of Intellectual Property Rights, or TRIPS] Agreement, to balance-of-pay- As of September 2000, 207 complaints had been ments restrictions, safeguard measures, and the tax- notified to the WTO (Table 9.1). Of these, 16 were ation of alcoholic beverages" (Lacarte-Muro and active, 40 had concluded with the adoption of either Gappah 2000). The exception is intellectual proper- an AB or a panel report, 34 had been settled bilater- ty rights (IPRs), with many complaints alleging vio- ally or were inactive, and 12 were being implement- lation of the TRIPS agreement by both developing ed (WTO 2000).4 Industrial countries brought the and industrial countries. As of 1999, the number of most cases, and their share of total complaints (74 TRIPS cases had already reached 16, equivalent to percent) was greater than their share of world 10 percent of all filings under the DSU. Eleven of exports. Among the different categories of cases, these filings were brought by the United States those brought by industrial countries against devel- (Geuze and Wager 1999). As regards developing oping countries appear to have increased the most countries, Correa (2001) notes that "although the between the GATT period and the WTO era, from adoption by another Member of unilateral trade 10 to 31 percent. Over 40 percent of industrial sanctions would be incompatible with the multilat- country cases were against developing countries-- eral rules, developing countries have continued to higher than the developing countries' 27 percent be under unilateral demands by some developed share of industrial countries' exports in 1998. The countries, notably the United States in the area of proportion of cases by developing countries against IPRs, in some cases aiming at ensuring protection of industrial countries was also higher than might such rights beyond the minimum standards set have been expected (66 percent of all developing forth by the Agreement" (Correa 2001: 22). country complaints) and was higher than the share Although developing countries are likely targets of industrial countries in developing country for intellectual property rights cases, IPRs may also exports, 57 percent (Weston and Delich 2000). become their most effective means of exerting pres- Latin America and Asia are the developing coun- sure and eventually retaliating. Ecuador's threat to try regions most involved in the dispute settlement suspend its TRIPS concessions in the Bananas case process. To date, African countries have not initiat- (see Chapter 10, by Hudec, in this volume) and the ed or been respondents in any case, although sever- strategy used by Brazil with regard to public health al, including Nigeria and Zimbabwe, have made and patents would appear to be the first steps in this presentations as third parties. direction (Dyer 2001; see also Chapter 36, by No single theme dominates the substance of the Maskus, in this volume).5 Subramanian and Watal cases involving developing countries. "As respon- (2000) have proposed that "developing countries dents, developing countries have been involved in convert their TRIPS obligations into instruments of Table 9.1 Number of Dispute Settlement Cases, 1995 through September 2000 Complaint by Share of Industrial Developing Total total cases countries countries complaints (percent) Complaint against: Industrial countries 89 35 124 60 Developing countries 65 18 83 40 Total 154 53 207 100 Share of total cases (percent) 74 26 Memorandum: Share of cases under GATT (percent) 84 16 Note: Based on number of cases brought by each country. The European Union and its member countries are counted jointly. Source: Weston and Delich (2000); WTO (2000b); IMF, Direction of Trade Statistics, various issues. 76 Developing Countries and the WTO Dispute Settlement System multilateral enforcement embodying the retaliation The compliance panel would comprise the mem- possibilities in domestic legislation" (p. 415). bers of the original panel, if its report had not been According to these authors, "domestic legislation appealed, or the members of the Appellate Body implementing the TRIPS agreement must clearly that considered the appeal if the report of the origi- specify that the country's executive reserves the nal panel had been appealed. The compliance panel right to revoke or dilute these rights in the event would be required to circulate its report within 90 that partner countries are found to be in non-com- days of the date of its establishment, after which any pliance with commitments that affect the country's party to the compliance panel proceeding would be interest" (p. 411). In addition, they hold that "if permitted to request a meeting of the DSB to adopt designed with care, retaliation in TRIPS can be fea- the report within a period of 10 days. The report sible, effective, and legal. Further, it has one really would be subject to the negative consensus rule: it attractive attribute that distinguishes it from con- would be automatically adopted unless the DSB ventional trade retaliation in the area of goods: decided by consensus not to adopt. retaliation in TRIPS can be genuinely welfare Compliance panel reports would not be subject to enhancing in a way that conventional retaliation--a appeal. If the compliance panel found that the case of shooting oneself in the foot to shoot at the member concerned had failed to bring its measures other person's foot--is not" (p. 405). into compliance within the reasonable period of time determined by the original panel, the com- plaining party could request authorization from the Proposals for Reforming the Dispute DSB to suspend the application of concessions to Settlement System the member concerned or to suspend other obliga- A number of proposals have been made by develop- tions under the covered agreements. ing countries and by scholars to improve the func- The joint proposal also modifies Article 22.2 to tioning of the dispute settlement system. This entitle the complaining party to request authoriza- section briefly summarizes these suggestions. tion to suspend concessions if a compliance panel report pursuant to Article 21 bis finds that the member concerned has failed to bring its measures On Implementation of Recommendations and into compliance with the ruling of the DSB. If the Rulings and Suspension of Concessions member concerned objects to the level of suspen- Three proposals have been made: amend the system sion proposed, the proposal states that "the matter to resolve procedural problems such as the sequenc- shall be referred to arbitration. The arbitration shall ing issue described earlier; allow financial compen- be completed and the decision of the arbitrator sation for developing countries; and turn retaliation shall be circulated to Members within 45 days after into a collective action. the referral of the matter. The complaining party A large number of WTO members have made a shall not suspend concessions or other obligations joint proposal that Article 21.2 of the DSU be during the course of the arbitration." reformed to address the sequencing problem.6 The In regard to financial compensation, Pakistan has proposal foresees the creation of Article 21 bis, enti- commented that "[i]t would be useful to clarify that tled "Determination of Compliance," that would the term `compensation' used in Article 22 includes establish the following procedures. A complaining grant of financial compensation to the complaining party may request the establishment of a compli- party by the country which has been found to be in ance panel (a) any time after the member concerned violation of the rules. Panels should be authorized states that it does not need further time for compli- to recommend payment of such financial compen- ance; (b) any time after the member concerned has sation in disputes between developed and develop- submitted a notification that it has complied with ing countries where they find that as a result of the recommendations or rulings of the DSB; or (c) WTO inconsistent measures taken by developed 10 days before the date of expiration for the "rea- countries, the developing country has lost its trade sonable period of time" to comply. While consulta- in the affected product" (WT/GC/W/162). It is not tions between the member concerned and the the first time that a developing country has called complaining party are desirable, they are not for the inclusion of financial compensation in the required prior to a request for a compliance panel. dispute settlement system: a similar proposal was 77 T H E W O R L D T R A D E O R G A N I Z AT I O N made during the GATT era (see Chapter 10, by porary developments." In particular, Pakistan pro- Hudec, in this volume).7 posed that such clarification make clear that panels Proposals have also been made to make violation or the Appellate Body were not permitted to take of WTO rules a collective problem and, accordingly, into account "unsolicited information" including to require collective retaliatory actions. Pauwelyn "amicus curiae briefs from private parties" (2000: 6), for instance, argues that (WT/GC/W/162). In November 2000, at a special WTO General with the advent of the WTO--its legal refine- Council meeting, developing countries called for ment, quasi-judicial dispute settlement system, the Appellate Body to exercise extreme caution in and, in particular, major expansion into new inviting amicus curiae briefs from nongovernmental fields that directly affect individuals--it may be organizations (NGOs). (The context was the devel- time to move away from the idea of the oping countries' reaction to the AB ruling on the GATT/WTO only as a package of bilateral bal- Asbestos case.) Developing countries sought to limit ances between governments. Has the time not the Appellate Body's "interpretation powers" and to come to introduce the WTO as a truly multilat- prevent NGOs from participating in the dispute set- eral construct providing legal rules as public tlement system. Brazil, Egypt, India, Pakistan, goods that merit collective enforcement for the Uruguay, and the Association of Southeast Asian good of governments and economic operators? Nations (ASEAN) countries argued that a decision . . . [T]he enforcement of WTO rules can and to admit amicus curiae briefs was a substantive and should be seen as a collective rather than a not a procedural one and was therefore something mainly bilateral exercise. for WTO members to decide. Moreover, "develop- ing countries took the view that non-governmental In addition, Pauwelyn has proposed that "coupled organizations are not accountable to sovereign par- with countermeasures, a broad scheme of compen- liaments and have no contractual rights and obliga- sation--additional market access offered by the los- tions in the WTO. The AB had let itself be unduly ing party to WTO members--would provide influenced by the campaign of NGOs of major trad- genuine leverage to induce compliance, a move ben- ing entities. In effect, NGOs were being accorded eficial to all WTO members, and not just `compensa- privileges greater than those enjoyed by WTO tion' to the one or few that brought the case" (p. 9). Members."9 Finally, Pakistan has presented a proposal that As for the composition of the Appellate Body, Article 22.3 be amended to eliminate the possibility India has proposed that, to promote an atmosphere of cross-retaliation by industrial against developing conducive to impartial and independent function- countries. This would impede, for instance, retalia- ing of the Appellate Body, all future appointments tion against trade in goods if a developing country of AB members should be for a nonrenewable fixed has been found to be in violation of the TRIPS term of five or six years, to ensure that members agreement (WT/GC/W/162, p. 3). have no incentive to seek support for their reap- pointment (WT/DSB/W/117). On the Appellate Body On Time Lines The role of the Appellate Body--in particular, the extent to which it has gone beyond its mandate and The joint proposal mentioned earlier would shorten undertaken to "make rules" through interpretation the consultation period from 60 to 30 days; the peri- of WTO agreements--has been severely questioned od could be extended by up to 30 additional days if by developing countries.8 Pakistan has called for an one or more of the parties in the dispute were a interpretation of "the relevant provisions in the developing country and the parties agreed. In addi- DSU to make it clear that the responsibility for clar- tion, in the Working Procedures, the proposal ifying or modifying the provisions of the WTO would reduce the time for receipt of the complain- Agreements clearly rests with the WTO member ing party's first written submissions to three­four countries and that it would not be appropriate for weeks (currently, it is three­six weeks), while the Appellate Body to usurp these functions under increasing the time for the party complained against the guise of interpreting law on the basis of contem- to respond to four­five weeks instead of two­three 78 Developing Countries and the WTO Dispute Settlement System weeks, as at present. Since the proposal unifies the provisions relating to enforcement of S&D language reports (there would be a single report, including in WTO agreements are ineffective, developing the descriptive sections and the panel's findings and countries do not enjoy a "neutral" playing field. conclusions), it eliminates the period during which Although the DSU is not biased against any party in parties submit their comments on the descriptive a dispute, developing countries are less well report and, consequently, the possibility that at the equipped to participate in the process: they have request of a party the panel would hold a further fewer people with the appropriate training, they are meeting on the issues identified in the written com- less experienced, and they can bring fewer financial ments. After making other adjustments on time resources to bear. Therefore, although the DSU is an lines, the proposal states that "the total reduction of asset, developing countries must work to obtain time is up to approximately 47 days, and the time international financing for training and capacity frames in Article 20 (the reference to 9 months and building and for the establishment of a joint mech- to 12 months), and the periods in Art. 21.4 (the ref- anism among developing countries to screen indus- erence to 15 months and to 18 months) shall be trial country trade policies of interest to them--not reduced by one month" (WT/MIN[99]8, p. 7). only to reduce the costs of the screening but also to coordinate the submission of joint cases. In addi- tion, developing countries could use cases in which On Third Parties they are involved as a way to identify gaps in WTO In relation to third parties, the joint proposal retains agreements that need to be addressed through the obligation contained in Article 10 that a copy of negotiations. all documentation submitted in a case be given to Reform of the dispute settlement system does not third parties. It allows exclusion, however, of certain appear to be a priority on the negotiating agenda of factual confidential information (designated as such developing countries. Their efforts are mainly by the disputing party) and sets a period of 15 days directed toward defending their interests as best for the party to provide a nonconfidential summary they can in current cases, bridging the gap with that can be disclosed to the public of the informa- industrial countries in terms of legal expertise, and tion contained in the confidential submission. establishing more effective enforcement and retalia- tory devices. A Special Prosecutor, "Light" Procedures, and Customs Unions Notes Hoekman and Mavroidis (2000) have proposed a 1 See for example, Komuro (1995); Lafer (1996); Jackson kind of "special prosecutor," able to act on an ex (1997); Montaña Mora (1997). officio basis, to detect illegalities. They also suggest 2 On the sequencing problem, see the Salmon dispute between "light" procedures for cases involving less than Austria and Canada; the U.S.-Australian dispute over leather US$1 million of exports; in such cases a single pan- subsidies; and the Bananas case. See also O'Connor and elist would be asked to address the dispute within Vergano (2000); Rhodes (2000); Valles and McGivern (2000). three months. Turkey has proposed amending Arti- 3 Cases in which S&D clauses were invoked by a party in a dis- cle 10 of the DSU to grant all parties to customs pute were European Communities: Antidumping Duties on unions the right to participate in panel and AB pro- Imports of Cotton-Type Bed Linen, complaint by India; Korea: ceedings in disputes concerning measures intro- Measures Affecting Imports of Fresh, Chilled, and Frozen Beef, complaint by the United States; India: Quantitative Restrictions duced pursuant to a common trade policy of the on Imports of Agricultural, Textile and Industrial Products, com- union (WT/MIN[99]/15). plaint by the United States; Brazil: Export Financing Programme for Aircraft, complaint by Canada; Canada: Measures Affecting the Export of Civilian Aircraft, complaint by Brazil; and Indone- Conclusion sia: Certain Measures Affecting the Automobile Industry, com- The Dispute Settlement Understanding brought plaints by the United States, the European Communities, and Japan. There were also cases in which S&D clauses were about a positive and beneficial change for develop- invoked by a third party or some kind of statement was made ing countries. Weaker states have a better chance to about a developing country's preferential treatment on the defend their interests in a rule-oriented than in a basis of its status as a developing country. Examples include power-oriented system. However, since the DSU Guatemala: Antidumping Investigation Regarding Imports of Port- 79 T H E W O R L D T R A D E O R G A N I Z AT I O N land Cement from Mexico, complaint by Mexico; European Com- 6 See "Proposed Amendment of the DSU," WT/MIN(99)8, sub- munities: Measures Affecting the Prohibition of Asbestos and mitted by the government of Japan on behalf of cosponsors Asbestos Products, complaint by Canada; and Mexico: Canada, Costa Rica, the Czech Republic, Ecuador, the Euro- Antidumping Investigation of High-Fructose Corn Syrup, com- pean Communities (and its member states), Hungary, Japan, plaint by the United States. These cases may be found on the Korea, New Zealand, Norway, Peru, Slovenia, Switzerland, WTO Website, . Thailand, and Venezuela. 4 There was one active case on implementation of WTO rulings, 7 For a detailed discussion on the Uruguay-Brazil plan to reform as well as six adopted AB and panel reports on implementation the dispute settlement system, including financial compensa- of WTO rulings (Art. 21.5 DSU), one active arbitration on the tion, see Dam (1970): 368­73. level of suspension of concessions (Arts. 22.6­7), and four authorizations of suspension of concessions (under Art. 22.7 8 Industrial countries have also questioned panel rulings on the DSU and Art. 4.10 of the Subsidies Agreement). same grounds. For instance, when the DSB, in a case involving subsidies to an automotive leather manufacturer, adopted a 5 Brazil linked developing countries' right to access affordable panel ruling that required, for the first time, that a private medicines to patents. First, Brazil put forward a very broad and company repay in full an illegal export subsidy, the countries ambitious plan to fight HIV in its territory and pressed corpora- involved--the United States and Australia--commented that tions to reduce drug prices. Simultaneously, Brazil obtained a "the ruling should not set a precedent for future disputes"; the declaration at the World Health Organization on the virtues of European Union said that the implications needed more dis- its HIV program. Finally, at the TRIPS Council, Brazil submitted cussion; and the United States stated that "the payback reme- a document highlighting the need to interpret the TRIPS dy went beyond that sought by the US." "Canada and Brazil agreement in a way that did not impede countries' ability to expressed their serious concerns about the decision and Japan implement health policies. Brazil received rapid and wide- and Malaysia voiced misgivings" (Financial Times, weekend, spread support from developing countries (and from public February 12­13, 2000). opinion as well). As a result of this strategy, the United States withdrew the panel against Brazil on intellectual property 9 The question of amicus curiae briefs is part of a broader debate rights, and at the Doha ministerial meeting a separate declara- on the governance of the trading system; see Chapter 47, by tion was made asserting that the TRIPS agreement does not Tussie and Lengyel, in this volume. See also "Developing Coun- and should not prevent members from taking measures to tries Make Their Mark on WTO Appellate Body Controversy," protect public health. World Trade Agenda, no. 00/22 (December 4, 2000), p. 11. 80 10 R O B E R T E . H U D E C The Adequacy WTO system allow larger coun- tries to exert significantly stronger enforcement pressures of WTO Dispute against developing countries than developing countries can Settlement exert in the reverse situation. The shortcomings of the WTO legal Remedies system in this regard thus raise a legitimate issue for developing country governments when they must decide whether to employ A Developing Country Perspective the dispute settlement procedure against larger countries. In addi- A tion, these shortcomings raise ccording to conventional wisdom, the question of whether it would be worthwhile for it is a waste of time and money for developing countries to expend negotiating capital developing countries to invoke the WTO's dispute in an effort to remedy these shortcomings, and, if so, settlement procedure against industrial countries. what particular reforms should be sought. Even if, the argument runs, a developing country In this chapter, I examine the facts behind the con- obtains a clear legal ruling that an industrial coun- ventional wisdom more carefully than is usually try has violated its legal obligations, the developing done. My purpose is not to prove that the conven- country has no effective way to enforce the ruling. tional wisdom is entirely wrong. Rather, I hope to The only enforcement sanction provided by the show that the issue here, as in most issues, is not quite WTO dispute settlement procedure is trade retalia- the open-and-shut proposition that is usually tion--the imposition of discriminatory trade sanc- advanced. I think it is important that officials respon- tions by the complaining country against the trade sible for deciding these issues have an accurate of the defendant country. And trade retaliation by understanding of what is and is not wrong with the smaller developing countries, it is argued, simply remedies offered by the existing WTO system, and of does not inflict any significant harm on larger how well or how poorly the system works in practice. industrial countries. In the end, the argument con- My purpose is to outline at least some of the infor- cludes, retaliation will harm the developing country mation needed to arrive at such an understanding. imposing it far more than it will harm the industri- al country it is supposed to punish. Enforcement under the GATT Dispute The conventional wisdom has a great deal of truth Settlement System to it. The "law" of the WTO does not, in fact, give weaker countries the same protection that well- The weaknesses of the GATT dispute settlement sys- developed domestic legal systems usually afford tem, which operated from 1948 to 1994, have been their weaker citizens. The remedies provided by the described so often that they need little elaboration. 81 T H E W O R L D T R A D E O R G A N I Z AT I O N The entire system was based on consensus decision- response to legal rulings. In the 1980s, when gov- making, which meant that the consent of the defen- ernments began to use the dispute settlement sys- dant was required before the procedure--creating a tem to deal with more politically controversial panel, defining its terms of reference, appointing its matters, the success rate dropped to about 81 per- members, adopting its ruling, and authorizing retali- cent--not up to the standards of most domestic ation--could move forward at all. The central reform legal systems, but still a very impressive perfor- made by the WTO Dispute Settlement Understand- mance for an international legal regime, especially ing (DSU) was to make the procedure go forward in the politically sensitive area of trade policy automatically on the request of the complainant, (Hudec 1993: 285­94). with or without the consent of the defendant. Although complaints by developing countries did In addition to its central weakness, the formal not achieve the same level of success as those remedies provided by the GATT legal system when brought by larger countries, the results were still legal violation was found were also rather limited. A favorable in a significant percentage of the cases. ruling of violation entitled the complaining govern- Over the GATT's entire history, 28 complaints were ment to a rather general "recommendation" calling brought by developing countries. Of these, 17 ended on the defendant government to comply with its in legal rulings, 11 of which were rulings of legal obligations. The recommendation was directed only violation, and 10 of the 11 (91 percent) had a suc- toward future conduct, with no compensation for cessful outcome. Of the 22 complaints known to be harm done while the violation was in force. There based on a valid legal claim, satisfaction was was no time limit on the order to comply, and the achieved in 18 of the cases (82 percent). Even in the process of seeking compliance could drag on for more contentious cases of the 1980s, legally valid years. The complaining government could at some complaints by developing countries achieved a 73 point request authorization to retaliate by imposing percent success rate (Hudec 1993: 315­26). approximately equal trade barriers in return, but The paradoxical contrast between the voluntary the request could be vetoed by the defendant. In procedures and weak remedies of the GATT dispute modern GATT practice, only two requests for retal- settlement system, on the one hand, and its rather iation authority were made, both against the United strong record of success, on the other, contains a les- States, and both were vetoed.1 son. It teaches that the enforcement of international Despite the defendant's ability to block the proce- legal obligations cannot be explained by superficial dure, the GATT disputes procedure produced a con- analysis of dispute settlement procedures and reme- siderable number of dispute settlement complaints dies. Enforcement requires that governments be during its almost 50-year history. My own study of persuaded to reverse decisions they have taken in GATT cases from 1948 to the end of 1989 counted violation of the agreement. Governments are not 207 cases filed during that period, of which 88 pro- private litigants. They are complex institutions that duced legal rulings; of the 88, 68 were rulings of vio- make decisions in their own peculiar, often irra- lation. In the decade of the 1980s, when the GATT tional, manner, which we call "politics." Even small- system had matured, there were 115 complaints er governments are strong enough to be able to yielding 47 legal rulings, of which 40 were rulings of resist coercive forces that would move private liti- legal violation (Hudec 1993: 277­78). Provisional gants. Governments, however, usually have a data from a continuation of that study list 71 more longer-term interest in the efficacy of the legal rela- complaints in the final five years of GATT opera- tionships they have established with other govern- tions (1990 to 1994), with 22 legal rulings, 20 of ments, and so they are more inclined to act in ways which were rulings of legal violation.2 designed to preserve those relationships. Ultimately, Notwithstanding the defendant's ability to block the compliance decisions of governments are deter- adoption of adverse rulings, the great majority of mined more by calculated self-interest than by the violation rulings were in fact adopted. More- force. over, the bulk of these violation rulings, including In my view, government compliance with legal many of those not adopted, did produce a satisfac- rulings is usually the product of at least three inter- tory correction of the practice at issue. In its first related factors that influence the way in which gov- three decades the GATT system achieved almost a ernments make trade policy decisions. First, some 100 percent success rate in producing a satisfactory parts of the defendant government's decisionmak- 82 The Adequacy of WTO Dispute Settlement Remedies: A Developing Country Perspective ing apparatus usually want the conduct called for by effective, in many cases it may well be more effective GATT legal obligations to be pursued for its own than the other practical alternatives. Policy deci- sake, simply because it is good policy. Such officials, sions that focus only on the availability of retalia- and the private interest groups that share this view, tion thus run the risk of ignoring the other, quite constitute an existing political force within the valuable, gains that can be achieved from a legal rul- defendant government, and the effect of GATT legal ing alone. rulings is to give them greater influence in the national decisionmaking process. Second, many The WTO Reforms officials and private interest groups within the national government's decisionmaking process per- In the Uruguay Round, WTO member governments ceive a value in the legal system itself, believing that agreed to establish a more rigorous dispute settle- both they and their country will gain more over the ment system. As noted above, they began by making long term from an effective legal system than they the disputes procedure move forward automatically. will gain from noncompliance in this or that indi- The automaticity of the procedure makes it more dif- vidual case. Although these actors may not want to ficult for larger countries to bully smaller countries tie their country's hands through rigid commit- into giving up their legal complaints. If developing ments to a particular legal system--or may not have countries want to have a legal ruling, it will now enough political support to go that far--they will require less diplomatic confrontation to get one. nonetheless argue strongly against noncompliance The remedies granted for enforcing a binding in individual cases that would damage respect for ruling were also strengthened. In addition to mak- the system. Finally, one should not underestimate ing retaliation more readily available, the Uruguay the influence of active pressure by other govern- Round reforms adopted a number of reforms ments. If a majority of member governments intended to strengthen the effect of the ruling itself. believes in the value of an existing legal system, The primary remedy set forth in the DSU is still the those governments will have the same incentive to legally binding "recommendation" ordering the discourage noncompliance with legal rulings and defendant to bring its conduct into compliance. will express their views in the form of collective Although the recommendation is still only future- condemnation of noncompliance. directed (it provides no remedy for the harm done But even if all three factors existed in cases where by the violation so far), some steps were taken to no legal rulings were provided, legal rulings sharpen make that future-directed order more effective.3 the focus on the issue of compliance, and the nor- Panels were given explicit power to make nonbind- mative force of such rulings increases the power of ing suggestions for how compliance can be those participants who favor compliance. This is so achieved--a power that, if used, could sharpen the whether or not the ruling is enforced by coercive focus of compliance pressures. Quite a bit more sanctions. was added to the procedure for following up a rec- As noted above, the GATT dispute settlement ommendation after it has been issued. There is now procedure almost never employed retaliation as an a procedure for establishing a time limit for com- enforcement device. The fact that the GATT pliance, which, so far, has ranged from 6 to 15 nonetheless produced a large number of successful months. During that period, the illegal measure is legal rulings indicates, therefore, that the internal under periodic review, and it remains under peri- government forces just described frequently did odic review, without further action by the com- play a significant role in bringing about successful plainant, as long as noncompliance lasts. For the outcomes. This is not to say, of course, that enforce- government that does not, or cannot, retaliate, ment would not have been even more effective if these changes make it easier for the complainant to more retaliation had been employed. Other things focus and maintain community pressures for com- being equal, one would expect a better chance of pliance. compliance with a retaliation tool than without it. Retaliation is still the final remedy for eventual The key point, however, is that a legal ruling with- noncompliance. In contrast to the GATT disputes out retaliation can still be an effective policy tool for procedure, under which retaliation was a vague and a developing country seeking to reverse a legal vio- seldom-used remedy, the new WTO procedure lation by a larger country. Although not invariably appears to make retaliation the central objective of 83 T H E W O R L D T R A D E O R G A N I Z AT I O N the remedy structure. The defendant no longer has utility of the dispute settlement procedure for a the power to veto retaliation requests, making it cer- developing country complainant. tain that retaliation will be authorized whenever noncompliance is established. The use of retaliation Proposals for Additional Reforms to Cure is subject to time-limited procedures to resolve dis- the Imbalance putes over whether the defendant has in fact failed to comply with the ruling, and over the amount and This section contains background information nature of the retaliation, but these procedures only about some of the reforms that have been advanced delay the remedy slightly. in the past to cure the perceived imbalance in GATT In the light of past experience, the WTO's greater and WTO dispute settlement procedures. emphasis on retaliation as an enforcement tool would appear to be somewhat misguided. The Compensation for Harm Done emphasis on retaliation seems to have been the result of an effort to accommodate pressures for For most of the GATT's history, the prevailing view stronger enforcement. The U.S. negotiators had to has been that the only remedy for violation of a persuade the U.S. Congress that the WTO dispute legal obligation is a forward-looking order directing settlement procedure had strong enforcement pow- the defendant to comply in the future. Except for a ers. To do this, the negotiators had to satisfy the leg- string of antidumping and countervailing duty islators' rather simplistic view that enforcement can (AD/CVD) cases in the late 1980s and early 1990s, be achieved by retaliation. As is usually the case defendant governments have not been required to when negotiators try to convince domestic legisla- compensate for harm done before the illegal mea- tures to support a new trade agreement, the sure was brought into conformity. Uruguay Round negotiators probably promised The most important challenge to the exclusively more enforcement power from retaliation than forward-looking view of GATT remedies was a 1965 retaliation can deliver. Threats of retaliation can be effort by GATT developing countries to add mone- useful, but they can also become counterproductive tary compensation to the list of dispute settlement if used too forcefully or too often. Governments remedies. This occurred at a time when developing must remember that enforcement is a more com- country GATT members were using the threat of plex process than mere retaliation, involving the abandoning the GATT for UNCTAD to ask for a bet- generation of the political forces needed to bring ter deal from GATT. The developing country caucus about the desired compliance decision. made several proposals to improve the operation of The new emphasis on retaliation probably makes the dispute settlement procedure. One proposal the WTO dispute settlement system even more one- resulted in the adoption of a special accelerated pro- sided than before. Retaliation by larger countries cedure for complaints by developing countries, tends to be most effective when used against smaller which is still in force today.4 Two other proposals countries, and so, by making access to retaliation concerned the improvement of remedies: a proposal more available, these new reforms give larger coun- for monetary damages to be paid to developing tries a still greater advantage over smaller countries countries injured by GATT-illegal trade restrictions, that cannot effectively retaliate. (Needless to say, the and a proposal for collective retaliation.5 increased one-sidedness would be viewed as an The theory behind the developing country pro- advantage by industrial country governments.) posal for monetary compensation was that GATT- Once again, however, it must be stressed that the illegal trade restrictions caused serious harm to the greater one-sidedness of the procedure does not fragile economies of developing countries and that mean that legal complaints by developing coun- this harm would be multiplied by its retarding effect tries--that is, legal complaints without the retalia- on the development process. In these circum- tion option--cannot be a useful and effective policy stances, the developing countries argued, forward- tool. To the contrary, the smaller reforms made in looking remedies were not enough to remedy the the Uruguay Round do make legal complaints with- harm already done. Instead, they proposed, devel- out retaliation quite a bit more effective than they oping countries should be entitled to collect were before. One-sidedness is a problem, but it is a retroactive damages in the form of money awards. separate problem that has nothing to do with the The money would compensate the government's 84 The Adequacy of WTO Dispute Settlement Remedies: A Developing Country Perspective economic development program, rather than pri- validity of the proceedings under review. Finally, it vate interests, which would have removed many of must be recognized that the disreputable character the problems in calculating the harm. It was under- of AD/CVD measures makes them a natural target stood that the obligation to extend monetary com- for aggressive regulation. pensation would remain in force until the measure At first, it seemed that GATT governments them- was corrected. selves agreed that a refund remedy should be avail- Developing countries strenuously advocated the able in such cases. The first such ruling, in a 1985 money damages proposal through a long series of case brought by Finland against New Zealand, was committee meetings in 1965. Industrial countries adopted by the GATT Council, and New Zealand opposed the proposal with equal conviction, assert- did in fact issue a refund.8 Afterward, six subse- ing that money damages were simply outside the quent GATT panel decisions ordered refunds, but in realm of the possible. In effect, they were saying, the each case the result was inconclusive. Two of these GATT was never meant to be taken that seriously. panel reports were adopted, but only after the issues The proposal was not adopted. had become moot, and over the express reservation The GATT practice of denying compensation for of the defendant as to the panel's rulings on past wrongs clearly reflects a view of GATT law as refunds.9 The other four rulings were blocked having a lower status than domestic law. Under the entirely by the defendant government, with at least domestic law of most GATT members, taxes or part of the objection to adoption being the remedy other charges imposed in violation of national law order.10 The principal opponent of such refund are a legal nullity, and government authorities are orders was the United States, joined later by the required to refund any monies so collected. The European Community (EC).11 trade laws of such countries similarly provide for Although governments renegotiated many provi- refund of tariffs collected in violation of national sions of the GATT Antidumping Code and the tariff law. To my knowledge, however, few if any GATT Subsidies Code in the Uruguay Round, the countries authorize such refunds when tariffs or negotiations yielded no answer to the impasse over other charges are found GATT-illegal.6 By limiting refund orders. The United States then cast its posi- GATT law in this fashion, governments are saying tion in cement when the U.S. Congress adopted a that they do not want GATT legal obligations to statutory provision, in the 1994 legislation imple- have such direct legal effect. Ultimately, it is a state- menting the Uruguay Round agreements, that ment that governments do not want (or do not have AD/CVD or safeguards duties already paid in "liq- sufficient political support) to make trade agree- uidated" entries would not be refunded, although ments that binding. the GATT-illegal duties could be revoked for all The one exception to the GATT's consistent prac- "unliquidated" entries.12 The issue of AD/CVD tice of issuing only forward-looking remedies was a refund orders has come up only once so far under series of GATT panel decisions between 1985 and the new WTO dispute settlement procedure, in the 1995, all involving antidumping and countervailing Guatemala Cement case.13 The panel in that case did duties, in which panels ordered refunds of duties not rule on the issue, however, and its entire opin- imposed in violation of GATT rules. It is not clear ion was set aside by the Appellate Body on other why these panels singled out antidumping and grounds. countervailing duties for more demanding remedies In conclusion, it bears repeating that the call for than those normally employed against other GATT- refund of GATT-illegal antidumping and counter- illegal charges on imports. One GATT panel vailing duties is an exception to the perfectly consis- referred to provisions in the 1979 Antidumping tent GATT practice of denying refunds of Code requiring government to refund overcharges. GATT-illegal tariffs and all other kinds of GATT- This obligation applies only to overcharges as illegal charges. One evident reason for the absence defined by national antidumping laws, and national of a refund remedy has been that many govern- governments do seem to comply with it when they ments have lacked domestic legal authority to find that, under national law, overcharges have been refund taxes or charges in such cases. Given the made.7 Furthermore, AD/CVDs rest entirely on usual controls on contingent government expendi- specific proceedings against specific firms, and thus tures, one might expect that few governments their validity seems more clearly contingent on the would be eager to seek such authority, particularly if 85 T H E W O R L D T R A D E O R G A N I Z AT I O N it were to extend to refunds of such things as GATT- This compensatory theory of trade retaliation has illegal tariff charges. Developing countries them- run through GATT law since the days of the negoti- selves might wish to think twice about whether they ations on the International Trade Organization wish to shoulder such a refund responsibility. (ITO) in 1947­48. 14 That theory is, of course, a pol- Meanwhile, the weight of joint U.S.-EU opposition icy choice; GATT governments could always adopt a to refunds in AD/CVD cases promises to be formi- different standard if they wanted to. The signifi- dable, especially since U.S. opposition is now cance of the history of the compensatory theory is required by statute. simply that it shows a steadfast desire on the part of leading GATT members not to have a law with stronger sanctions. Other Forms of Trade Retaliation The second objection to proportional retaliation The standard complaint of developing countries is a practical one: an individual developing country about the remedy of trade retaliation is that it is too usually does not have a large enough market to weak to be effective against large countries. The assemble the amount of trade retaliation that would amount of retaliation is limited to the trade loss be needed to cause noticeable pain in a large indus- caused by the illegal trade measure in question. trial country--at least not without shutting down Since individual developing countries tend to have most of its own economy. The only way to achieve only a small share of the defendant country's mar- significantly greater retaliation would be to develop ket, their retaliation measures can affect only a small some form of collective retaliation by many coun- amount of that country's trade--usually not tries at the same time. Thus, the proposal usually enough, the argument runs, to cause any significant shifts to one for collective retaliation, asking that the hardship for the large industrial country or its pro- GATT put aside both the compensatory limit to ducers. retaliation and the notion that only the com- Academic discussions of GATT/WTO remedies plainant is entitled to retaliate. usually arrive at the question of whether In 1965 the developing country proposals for GATT/WTO trade retaliation should not be mea- reform of the GATT dispute settlement remedies sured according to a scale that would make certain included a proposal calling for collective retalia- that the amount of retaliation is large enough to be tion.15 The justification for the proposal was the meaningful against larger countries. Such proposals same as the one advanced today: individual devel- are usually justified by arguments that the law must oping countries could not impose sufficient retalia- have sanctions large enough to accomplish its task. tion to cause noticeable pain in larger industrial Two objections are usually interposed against countries. The idea was that in such cases a number such proposals for proportional retaliation. The of countries would be authorized to deny market most important is the assertion that the purpose of access to the large-country defendant. By definition, retaliation has never been to serve as a punitive this retaliation would also have been punitive in sanction; on the contrary, the right to retaliate has amount, although there were some arguments that always been viewed as a right to maintain the bal- higher retaliation levels could be based on a "devel- ance of reciprocity in GATT obligations. The start- opment multiplier" that inflated the measurement ing assumption has been that the obligations of the harm developing countries suffered from undertaken by each country involve a balance of GATT-illegal trade restrictions. benefits--the benefits granted to others in the form Industrial countries strongly resisted this propos- of a country's own obligations, balanced against the al. Beyond the objections based on unwillingness to benefits that country obtains from the obligations change the "compensatory" limit to retaliation, undertaken by others. The theory is that a breach of there were also objections based on an assertion legal obligations reduces the benefits being received that multiple retaliations would soon produce so by the complaining country and that, if the breach many new restrictions that they would choke the is not cured, the complaining country must be channels of commerce. In informal conversations, allowed to reestablish the balance by withdrawing industrial country delegates tried to point out that, obligations of its own. Such balancing, however, solidarity notwithstanding, countries not involved requires only retaliation equal to the amount of the in the dispute would soon tire of being asked to benefits lost. harm their own citizens for this purpose. Even far- 86 The Adequacy of WTO Dispute Settlement Remedies: A Developing Country Perspective ther behind the scenes, of course, was the awareness Cross-Retaliation under the TRIPS Agreement by industrial countries that the existing limitations on remedies suited them quite well, for the very In 1999 the several strands of argument claiming same reasons that developing countries did not like that trade retaliation is not a practical policy instru- them. Viewing things from the perspective of their ment for developing countries were brought togeth- role as potential defendants, industrial countries er in a new type of retaliation proceeding initiated were quite content with membership in a legal sys- by Ecuador in the Bananas case (Box 10.1).16 tem in which they could hurt others but some of the Ecuador tried to take advantage of the "cross-retali- others could not really hurt them. ation" provisions found in Article 22.3 of the WTO B O X 1 0 . 1 T H E B A N A N A S C A S E The European Union (EU) import regime for Operators that traditionally exported bananas bananas has long been a bone of contention. In from former British and French Caribbean effect, the EU maintains a system that gives pref- colonies were granted 30 percent of all import erential market access to bananas produced by licenses for non-country-specific quotas. These African, Caribbean, and Pacific (ACP) countries. licenses could be used to import ACP bananas or As a result Caribbean producers have always had could be sold to firms desiring to import from a substantial share of the EU market, to the detri- Latin America. In the latter case, which often ment of Central and South American countries. occurred, the quota allocation system resulted in Preferences predated the formation of the Euro- a transfer of rents from the (mostly U.S.-based) pean Economic Community (EEC) and, in fact, firms buying the licenses to those granted the caused problems between France and Germany quota rights. Borrell (1997) estimated that the during the negotiations leading to the creation of new regime was worse than the national ones it the EEC in 1957; Germany had a free trade replaced: total costs to EU consumers were about regime for bananas and imported from Latin US$2 billion, while ACP suppliers obtained American countries, while France maintained US$150 million--a cost per consumer of over very high barriers to support French colonial pro- US$13 for each dollar transferred. ducers (Messerlin 2001). These differences led to Latin American producers brought two cases to the imposition of national intra-EU trade barriers, the GATT contesting the national systems (in reserving the U.K., French, and Spanish markets 1992) and the new common EU regime (in for former colonies. The policies were a very inef- 1993). They won both. In 1994 the EU conclud- ficient way of assisting the former colonies: every ed a Banana Framework Agreement with four dollar transferred cost EU consumers US$5, of countries (Costa Rica, Colombia, Nicaragua, and which US$3 went to distributors and US$1 was Venezuela) under which these countries were wasted (Borrell 1997). allocated specific quotas on the understanding In 1993 the EU adopted a complex import that they would not bring a case to the WTO licensing and distribution system for the union as before 2002. In 1996 four Latin American pro- a whole, as part of its effort to create a single mar- ducers that had been left out of this agreement ket. The common market organization that was (Ecuador, Guatemala, Honduras, and Mexico), imposed was based on historical trading relation- joined by the United States on behalf of U.S. ships and was designed to continue to provide multinational fruit firms, contested the EU import preferential access for ACP countries (signatories regime in the WTO, claiming that it discriminated of the Lomé Convention). It involved two tariff against their producers and banana marketing quotas--one for traditional ACP suppliers and one companies. The object of the attack was not so for nontraditional ACP and Latin American grow- much the tariff preferences that were granted to ers--and four categories of suppliers. Out-of- ACP countries--for which the EU had obtained a quota imports were subject to high specific tariffs. waiver--but the allocation of quotas. (continued) 87 T H E W O R L D T R A D E O R G A N I Z AT I O N B O X 1 0 . 1 ( C O N T I N U E D ) The WTO panel report, published in June 1997, ened imposition of tariffs on their goods (Finan- found the EU banana import regime in violation cial Times, August 26, 2000, p. 5). of WTO nondiscrimination and market access Toward the end of 1999, Ecuador also sought rules. The dual tariff rate quota regime was found and obtained authorization to retaliate. Its request to be inconsistent with GATT Article XIII (requir- was a double first in the history of the trading sys- ing nondiscrimination), and the 30 percent allo- tem: the first request for retaliation by a develop- cation of import licenses to traditional sellers of ing country, and the first time approval for ACP bananas was ruled inconsistent with GATS cross-retaliation had been sought. Ecuador nondiscrimination rules. On appeal, the Appellate argued that its merchandise imports from the EU Body endorsed most of the panel's conclusions. were too small to allow full retaliation (set at In 1998 the EU revised its regime. It continued US$200 million by the arbitrators) against imports to maintain two tariff rate quotas, but it assigned of EU goods. It obtained authorization to suspend import quotas for non-ACP countries on the basis concessions under other agreements, including of historical market shares and abolished the TRIPS, after having exhausted the possibilities for operator categories for allocation of licenses. retaliating against imports of EU consumer goods. Consultations regarding the WTO-consistency of (The panel concluded that retaliation against the new measures were inconclusive. Just before imports of intermediates and machinery would be the January 1999 deadline for implementation, "ineffective"--that is, too costly for the economy.) the United States sought authorization to retali- This use of cross-retaliation was not foreseen by ate. To this, the EU responded that the United negotiators in the Uruguay Round, who had States should first obtain a panel finding that the envisaged cross-retaliation as an instrument to new mechanism did not conform to WTO rules. enforce the TRIPS agreement (since developing Ecuador did request that the original panel exam- countries were not major exporters of intellectual ine whether the EU measures were in compli- property­intensive goods) rather than as a vehicle ance. The Dispute Settlement Body (DSB) for developing country retaliation. reconvened the original panel to examine both At the time of writing, Ecuador had not imple- requests. Concurrently, the United States request- mented retaliatory actions, and negotiations con- ed authorization from the DSB to retaliate against tinued. In October 2000 the EU proposed a the EU in the amount of US$520 million. The EU system of three tariff quotas, to be allocated on a responded with a request for arbitration. first come, first served basis, with the adoption of The panel rejected the EU argument, given that a tariff-only regime by 2006. Latin American Ecuador had challenged it. Responding to countries objected to the move away from histor- Ecuador, the panel found that the new EU meas- ical market shares, while the United States object- ures were not fully compatible with the WTO. ed to the tariff quota for ACP bananas. The same panel determined the level of nullifica- The Bananas case illustrated that disagreements tion suffered by the United States to be equiva- between parties on the adequacy of implementing lent to US$191.4 million. Subsequently, the measures have the potential to give rise to a recur- United States was authorized to raise duties ring series of panels dealing with essentially the against the EU by that amount. The U.S. retalia- same issue. It also revealed the weakness of the ulti- tion included a provision allowing for a "carousel mate enforcement threat that is available, although approach" under which a different set of exports the innovative use of cross-retaliation threats by from the EU was subjected to retaliatory tariffs of Ecuador suggests that small countries do have up to 100 percent each six-month period. This mechanisms for putting pressure on large players. procedure was designed to maximize the political "pain" of the retaliation. In this it was successful, Source: Prepared by the volume editors, based on WTO as illustrated by the lobbying by the U.K. cash- data available at ; Porges (2000); Hoek- mere wool products industry against the threat- man and Kostecki (2001); Messerlin (2001). 88 The Adequacy of WTO Dispute Settlement Remedies: A Developing Country Perspective Dispute Settlement Understanding (DSU). These possibility of substituting input suppliers from other provisions allow a country to retaliate against a vio- countries with the observation that if inputs were in lation of obligations under one WTO agreement fact available from other suppliers on equal terms, (say, violations of the TRIPS agreement) by sus- they would already be in the market. In the end, how- pending obligations under another agreement (for ever, the panel backed away from this rather summa- example, obligations under the GATT) in cases ry dismissal of the substitute-supplier issue and where suspension of obligations under the violated relied on the conclusion that the EC had not suffi- agreement (TRIPS, in this case) would not be "prac- ciently rebutted Ecuador's claim that substitution ticable or effective." The cross-retaliation provision would involve "transitional costs" of significant mag- was originally demanded by industrial countries to nitude for a developing country (paras. 93­94). In allow them to impose trade retaliation under the sum, the panel's superficial analysis did not add GATT to sanction violations of the TRIPS or GATS much authoritative weight to the traditional develop- agreements and was adopted over strong objections ing country argument, except perhaps to suggest that by developing countries. the panel members themselves had been conditioned In the Bananas case the EC had violated GATT to accept it without much analysis. and GATS obligations. The EC failed to comply The panel then employed an equally superficial within the time allowed, and so, after the United analysis to "split the baby" by ruling that developing States had retaliated against the EC by raising tariffs country retaliation against consumer goods was not on EC goods, Ecuador decided to seek authority to "impracticable." It recognized that retaliation on retaliate as well. At this point Ecuador turned the imports of consumer goods would impose welfare tables on the proponents of cross-retaliation by losses on developing country consumers but then arguing that retaliation against EC exports of goods rather summarily dismissed Ecuador's hardship or services was not "practicable or effective" under argument by saying that Ecuador had not presented DSU Article 22.3. Consequently, it asked for author- enough evidence of hardship to justify a conclusion ity to cross-retaliate by suspending certain of its of impracticability (para. 100). Consequently, for obligations under TRIPS. GATT violations involving trade in goods, Ecuador Ecuador's arguments in support of the assertion was required to retaliate on consumer goods before that trade retaliation was not "practicable or effec- being allowed to retaliate in other sectors. tive" were the same arguments that had been made The panel also interpreted the word "effective," by developing countries for decades: that trade stating that the "effectiveness" criterion included the retaliation harms the developing country more than issue of whether the retaliation would have a mean- it harms the industrial country defendant and that, ingful political impact on the defendant country moreover, the retaliation is too small to inflict any (para. 72). This interpretation, incidentally, can be meaningful pain on the industrial country.17 In the viewed as a formal recognition of the post-WTO arbitration proceeding to review Ecuador's request tendency to view retaliation as a sanction designed for cross-retaliation, the arbitration panel was to induce compliance by economic pain, rather than required to render a judgment about Ecuador's the original view of retaliation as a form of tempo- argument--in effect, to render a formal legal judg- rary compensation for an imbalance of benefits. ment about the long-maintained claim that trade The facts of the Bananas case did not constitute a retaliation was an inadequate legal remedy for very good argument for inadequate effectiveness. developing countries. The panel's decision was less Since the value of Ecuador's lost banana exports to illuminating than it might have been, but on the the EC was uncommonly large, the dollar value of whole it gave a certain degree of official approval to Ecuador's trade losses (US$201.6 million) was actu- that argument. ally larger than the allowed amount of U.S. retalia- The arbitration panel began by interpreting the tion in the Bananas case and also larger than either word "practicable." That term was interpreted to the U.S. or Canadian retaliation in the Beef Hor- include consideration of whether retaliation would mone case. Nonetheless, the panel did make a find- harm the developing country itself (WT/DS27/ ing of ineffectiveness, albeit not a very clear one. It ARB/ECU, paras. 70, 73). The panel ruled that retali- gave two different, and apparently inconsistent, ation which increases the cost of industrial inputs answers. With regard to EC exports of industrial would not be "practicable." It tried to brush aside the inputs, the panel ruled that Ecuador's retaliation 89 T H E W O R L D T R A D E O R G A N I Z AT I O N against such industrial inputs would be ineffective I would end on a note of caution with regard to "given the fact that Ecuador, as a small developing TRIPS retaliation. As pointed out earlier in this country, only accounts for a negligible proportion chapter, there is considerable evidence that the of the EC's exports of these products" (para. 95). power of retaliation, although helpful, is not really With regard to consumer goods, the panel ruled the key ingredient in enforcement of GATT/WTO that Ecuador had failed to demonstrate that retalia- legal rulings. To reiterate the point, enforcement of tion against such goods would be ineffective, with- legal rulings is a political process involving the culti- out explaining why Ecuador's similarly "negligible" vation of a government decision to change a previ- proportion of the EC market for consumer goods ous decision. The U.S. Congress was wrong to insist was not as probative in that case (para. 100). on retaliation as the key to its enforcement The rather superficial and inconsistent answers demands. The U.S. negotiators were wrong to play on this point suggest that the panel did not have up to that misconception by trying to persuade the enough time to develop a fully coherent analysis of Congress that easier retaliation would make WTO the long-standing developing country claims about enforcement as effective as the Congress wanted. the inadequacy of trade retaliation. But the answers Developing countries would be just as wrong to do show the panel's inclination to support those think that practicable TRIPS retaliation will bring claims of inadequacy, and they also show that the about a decisive change in the political fundamen- support may be limited to the clearer case of retalia- tals of WTO enforcement. More effective retaliation tion against industrial inputs. It would not be wise will make the system work somewhat better for to read much more into the decision than this, espe- developing countries, but it is not wise to invest all cially since, in the absence of appellate review, the the eggs in that basket. next panel will not be bound by anything said in this report. On balance, the panel's rather unclear Notes response was encouraging enough to make it worthwhile for developing country officials to think 1 Both requests came in the Superfund case, United States: Taxes about the possibility of cross-retaliation in dispute on Petroleum and Certain Imported Substances, GATT, BISD, 34th Supp. 136­66 (1988). The follow-up proceedings are dis- settlement cases involving uncured violations. cussed in Hudec (1993): 210­11, 535­37. Potentially, the most significant aspect of Ecuador's retaliation proposal was the possibility 2 The continuation study has not yet been published. that retaliation under TRIPS could be both more 3 The next section discusses the single exception to the state- "practicable" and more "effective" than trade retalia- ment that the recommendation provides no remedy for harm tion. In theory, at least, denying the intellectual incurred in the past: the effort in several cases to make defen- property rights of foreign owners results in assets dant governments refund antidumping and countervailing duties imposed in violation of GATT law. See text at notes being made available to developing countries at 7­14, below. cheaper prices, which is usually a benefit to econom- ic development rather than a burden on it. Likewise, 4 See GATT, BISD, 14th Supp. 18 (1966). The procedures call for mediation and fact-gathering by the secretariat, the automatic although the amounts of retaliation in most cases establishment of a panel (a significant advance in those days), will still be "negligible," at this time in the WTO's and a considerably accelerated time schedule. history the ripple effects of even small-scale denial of 5 For a brief description of the negotiations, see Hudec (1990): intellectual property protection could cause consid- 242­43. The GATT document series recording the negotiations erably more political discomfort than the usual is COM.TD/F. The main proposals are COM.TD/F/W.1 (April small-scale case of trade retaliation. As the arbitra- 27, 1965) and COM.TD/F/W.4 (October 11, 1965). I partici- tion panel itself made clear, however, TRIPS retalia- pated in these negotiations as a U.S. delegate, and some of tion will involve a number of distinctive legal, the information given here is based on my personal recollec- practical, and economic problems for the retaliating tions. state.18 Ecuador's retaliation request in the Bananas 6 Normally, the only way to secure refunds is to try to persuade case is therefore only a very tentative first step in a national authorities to revise their interpretation of national much longer journey.A great deal more analysis, and law in light of the GATT/WTO ruling so that as a matter of national law the refund is owing. The refunds carried out after considerably more practical experience, will be adverse panel rulings made under the North American Free needed before it is clear whether TRIPS retaliation is Trade Agreement (NAFTA) Chapter 19 review of AD/CVD the key to this long-troubling problem. occur only because national legislation makes dispute settle- 90 The Adequacy of WTO Dispute Settlement Remedies: A Developing Country Perspective ment under Chapter 19 part of the domestic AD/CVD pro- 14 The text of GATT Article XXIII.2, based on the August 1947 draft ceeding and thus binding on national authorities as a matter of the ITO Charter, states that the contracting parties may of domestic law. authorize such retaliation "as they deem to be appropriate in the circumstances." The final ITO Charter text adopted in March 7 See the explanation in note 6. 1948 changed this passage to read "appropriate and compensa- 8 New Zealand: Imports of Electrical Transformers from Finland, tory, having regard to the benefit which has been nullified or GATT, BISD, 32nd Supp. 55­70 (1986). impaired," to make clear that retaliation was not to exceed the amount needed to compensate for the harm done (ITO Charter, 9 The two cases are as follows. (1) United States: Countervailing Article 95.3). See also Havana Conference, Reports of Committees Duties on Fresh, Chilled and Frozen Pork from Canada, GATT, and Principal Subcommittees, UN, ICITO 1/8 (September 1948), BISD, 38th Supp. 30­47 (1992). After the U.S. CVD had been p. 155. withdrawn, the United States agreed not to block adoption of During the history of the GATT, the only GATT panel to discuss the panel ruling, but it reserved its position on the merits. (2) the issue was the 1952 panel that adjudicated the level of the United States: Measures Affecting Imports of Softwood Lumber Netherlands retaliation against U.S. dairy restrictions. The panel from Canada, GATT, BISD, 40th Supp. 358­517 (1995). The claimed that the word "appropriate" in the (1947) text of Article United States announced that it would be refunding deposits XXIII.2 gave the panel a certain flexibility to take into account and bonds for other reasons, but it expressly reserved its posi- other factors that might aggravate the harm. But the panel tion on the validity of the panel's order that deposits and found it "appropriate" to reduce the level of retaliation by 20 per- bonds be refunded; GATT, SCM/M/67 (meeting of October cent from the amount submitted by the Netherlands, suggesting 27­28, 1993). that flexibility cut in both directions. See Netherlands: Action under Article XXIII:2, GATT, BISD, 1st Supp. 32, 62-64 (1953), dis- 10 The four cases were as follows. (1) Canada: Countervailing Duty cussed in detail in Hudec (1990), ch. 16. A similar interpretation on Boneless Manufacturing Beef, GATT, SCM/85 (October 13, of "appropriate" was offered by the GATT secretariat's legal 1987); panel report not adopted. The case is discussed briefly adviser during discussions of the Superfund case. See GATT, in Hudec (1993): 221­22, 533­34. (2) United States: C/M/220 (GATT Council meeting of April 8, 1988), p. 35. Antidumping Duties on Stainless Seamless Pipes and Tubes from In the WTO Dispute Settlement Understanding, para. 22.4 Sweden, GATT, ADP/47 (August 20, 1990); panel report not clearly states that the retaliation shall be equivalent to the adopted. The impasse over remedies in this case is discussed amount of the nullification and impairment caused by the meas- briefly in Hudec (1993): 253­54, 572­73. (3) United States: ure at issue. Thus, it returns to the original meaning of the ITO Antidumping Duties on Gray Portland Cement and Cement Clink- Charter. The WTO arbitration panels that have ruled on the er from Mexico, GATT, ADP/82 (July 7, 1992); panel report not amount of retaliation have all followed this instruction. See Euro- adopted. (4) European Communities: Antidumping Duties on pean Communities: Regime for the Importation, Sale and Distribu- Audio Tapes in Cassettes Originating in Japan, GATT, ADP/136 tion of Bananas--Recourse to Arbitration by the European (April 28, 1995); panel report not adopted. Communities under Article 22.6 of the DSU, WT/DS27/ARB/USA 11 The European Community appears to have shifted toward the (April 9. 1999) (U.S. retaliation); id., WT/DS27/ARB/ECU (March U.S. position during a 1993 complaint against Brazilian coun- 24, 2000) (Ecuador retaliation); see also European Communities: tervailing duties; it initially asked for a refund order in its com- Measures Concerning Meat and Meat Products (Hormones -- plaint but withdrew its request during the panel proceeding. Recourse to Arbitration by the European Communities under Article See Brazil: Imposition of Provisional and Definitive Duties on Milk 22.6 of the DSU, WT/DS26/ARB (July 12, 1999) (U.S. retaliation); Powder and Certain Types of Milk from the European Economic id., WT/DS48/ARB (July 12, 1999) (Canada retaliation). Community, GATT, SCM/179 (December 27, 1993), para. 200. 15 See sources cited in note 5. The EC opposed a request for refunds in the Audiocassette case (see note 10). 16 European Communities: Regime for the Importation, Sale and Dis- 12 Section 129 of the Uruguay Round Agreements Act, 108 Stat. tribution of Bananas -- Recourse to Arbitration by the European 4813, 4836, 19 U.S.C. 3501, 3538 (1994), provides partial Communities under Article 22.6 of the DSU, WT/DS27/ARB/ECU authority to revoke AD/CVD and safeguards measures in order (March 24, 2000) (decision by panel of arbitrators). to comply with WTO panel rulings. Although this is an 17 The objection to the possibility of GATS retaliation by Ecuador advance over prior law, subsection (c)(1) limits the effect of in other services sectors was based on essentially different such revocations to "unliquidated" entries that enter or are arguments, resting primarily on the nature of Ecuador's limited withdrawn from the warehouse on or after the date of the GATS obligations (WT/DS27/ARB/ECU, paras. 103­20). The order revoking the measure. issues raised by this defense, and the panel's response, are not 13 Guatemala: Antidumping Investigation Regarding Portland treated in this chapter. Cement from Mexico, WTO, WT/DS60/R (June 19, 1998) (panel report), reversed on appeal WT/DS60/AB/R (November 2, 18 The panel delivered a lengthy lecture on the prospective perils 1998) (Appellate Body report). of such retaliation (WT/DS27/ARB/ECU, paras. 130­65). 91 III SELECTED TRADE POLICIES AFFECTING MERCHANDISE TRADE T rade barriers imposed at the border Alessandro Nicita, offer the interested reader more remain high in some parts of the world. detailed data on the patterns of trade and protec- Average (unweighted) tariffs in South tion prevailing at the end of the 1990s. Asia are in the 25 percent range or higher, well Protection in industrial countries currently imposes above the 10 percent average found in East Asia, costs on developing countries that exceed the approx- Latin America, and Eastern Europe and Central Asia. imately US$45 billion in official development aid flows Nontariff barriers remain a serious obstacle in many received by these countries each year. Protection countries. At the same time, industrial countries imposed by developing countries carries a cost to the maintain high tariffs on certain "sensitive" prod- world economy of over US$60 billion per year. Global- ucts--mostly labor-intensive items that are pro- ly, tariff barriers to trade in merchandise cost the world duced by developing countries, as well as many economy about US$250 billion. This ignores the effect agricultural products. In Chapter 11 Sam Laird pro- of contingent protection (antidumping and safe- vides an overview of the remaining tariffs and non- guards) and of the red tape involved in customs clear- tariff barriers and their impact. The appendixes to ance. It is evident that the benefits of reducing market this volume, by Francis Ng, Marcelo Olarreaga, and access barriers are enormous. 93 S E L E C T E D T R A D E P O L I C I E S A F F E C T I N G M E R C H A N D I S E T R A D E Much attention has centered recently on granting and looks at how these might be made more least-developed countries (LDCs) duty- and quota- responsive to the needs of developing countries. In free access to industrial country markets. This is Chapter 16 Brian Rankin Staples examines the important for these countries because existing pat- organizations and instruments involved in trade terns of protection discriminate against them, as is facilitation initiatives and summarizes the lessons demonstrated in Chapter 12, by Olarreaga and Ng. from cross-country experience in this area. Preferential access to markets will be beneficial to Many countries have sought to use export-pro- LDCs, but it comes at a cost to other developing moting policies either as a vehicle to offset the anti- countries. This cost, however, is limited, given the export bias created by other policies (overvalued small size of most LDC economies. Of greater signif- exchange rates, transactions costs, and so on) or as icance is the evidence that preferences are of limited a way of supplementing trade reform efforts. Indus- value. One reason is that they are generally condi- trial policies of various types are common in many tional on stringent rules of origin. Chapter 13, by countries: examples include subsidies, export pro- Luis Jorge Garay and Rafael Cornejo; Chapter 14, by motion, and creation of export-processing zones Stefano Inama; and Box 13.1, by Gomi Senadhira (EPZs). Two issues arise: What makes sense from a (on the U.S. African Growth and Opportunity Act) development viewpoint? And to what extent does show that rules of origin can be restrictive and can the WTO restrict the use of efficient policies for pro- give rise to high compliance (red tape) costs. moting industrialization and export development? Red tape is also an important factor in customs There may, in fact, be a good case for pursuing clearance procedures in general. Developing coun- EPZs and promoting exports; these mechanisms can tries therefore confront a large and important trade be effective ways of offsetting the high transactions facilitation agenda. In part, this agenda has to do costs that prevail in developing economies and that with market access--for example, with simplifying inhibit investment. It is important, however, to rules of origin, which can be pursued through the design such schemes in ways that limit the scope for WTO (see Chapter 14)--but it is mostly domestic. rent-seeking and reduce the likelihood of invest- The domestic part of the agenda is the most impor- ment occurring in sectors in which the country does tant and requires institutional strengthening, as well not have a potential comparative advantage. As as policy change. Of particular relevance for the dis- Mari Pangestu explains in Chapter 17 in the East cussion in Part III are customs administration reform Asian context, WTO rules do not significantly con- and trade facilitation. These are areas in which strain the ability of developing countries to pursue numerous international bodies are active and in welfare-enhancing policies. The agreements do, which the private sector can become part of the however, have implications for industrial policies, solution--for example, through the provision of cer- especially export subsidies and local content tification or inspection services. requirements. Country experience suggests that strengthening Philip English and Luc De Wulf, in Chapter 18, customs regimes and their administration to reduce examine experiences with trade promotion organi- transaction costs, antiexport bias, and corruption is zations, EPZs, subsidies, duty drawbacks, and other important in harnessing trade reform for develop- export promotion policies and mechanisms and ment. Transactions costs related to customs clear- review the options for developing countries. In ance can be a major impediment to investment in countries where tariff revenues continue to be need- tradable sectors, especially in activities that are time- ed, it is crucial that exporters have access to import- sensitive or where it is important to be integrated ed intermediate inputs at world market prices in into global production networks that operate on the order to be competitive. This requires well-function- basis of just-in-time supply chain management. ing customs regimes that efficiently refund duties Streamlining customs procedures and eliminating paid on imported inputs or, preferably, allow red tape require a concerted effort that involves exporters to import inputs duty-free without run- exploiting potential partnerships and synergy with ning afoul of WTO subsidy rules. Implementing organizations that have expertise in this area, such systems requires training and institutional including the private sector (for example, express strengthening. For example, many African countries carriers). Vinod Rege, in Chapter 15, reviews inter- lack well-functioning drawback regimes, and this national efforts to standardize customs valuation increases antiexport bias. 94 Selected Trade Policies Affecting Merchandise Trade Trade-related investment measures (TRIMs) are provisions and outline how firms should respond to sometimes used in an effort to promote industrial the various stages of the trade litigation process. development. Among them are local content Although this Handbook is not a legal reference to requirements, which, as Bijit Bora notes in Chapter the WTO, many of the chapters in this part refer to 19, have become controversial following the adop- GATT provisions. For ease of reference for those who tion of WTO rules that apply to developing coun- are not familiar with the WTO, the Glossary provides tries. (These rules were already embodied in the a summary of key GATT rules and articles. GATT but were not enforced against this country group.) Although, in principle, a case for such poli- Further Reading cies can be made--they may be appropriate for off- setting specific distortions--experience in many UNCTAD, Duty and Quota Free Market Access for countries reveals that great care must be taken in LDCs: An Analysis of QUAD Initiatives (Geneva, 2001), their use. The case study on Australia in Chapter 20, is a comprehensive and detailed discussion of cur- by Garry Pursell, illustrates that the use of TRIMs can rent initiatives to provide LDCs with preferential mean a very high cumulative cost to society. access to major industrial country markets. Rolf The final chapters in this part look at the elimina- Langhammer and André Sapir, Economic Impact of tion of remaining quotas on imports of apparel and Generalized Tariff Preferences (London: Trade Policy textiles and at the use of safeguard actions. These Research Centre, 1988), although somewhat dated, topics are closely linked: many observers expect the is a useful analysis of the economic effects of Gener- incidence of safeguard measures to increase once alized System of Preferences (GSP) schemes. The quotas under the Multifibre Arrangement (MFA) are authors argue that these schemes largely benefit fully eliminated, as required by the WTO Agreement countries which pursue export-oriented policies and, on Textiles and Clothing (ATC). Hanna Kheir-El-Din, for the most part, do not need preferences to com- in Chapter 21, discusses the implications of the ATC pete. Edwin Vermulst, Jacques Bourgeois, and Paul for developing countries. Waer, Rules of Origin in International Trade: A Compar- The so-called instruments of contingent protec- ative Study (Ann Arbor: University of Michigan Press, tion--antidumping and emergency measures--that 1994), provides a comprehensive discussion of ori- are permitted under the WTO if imports are deemed gin rules. John Raven, Trade and Transport Facilita- to injure domestic industries are a major source of tion: An Audit Methodology (Washington, D.C.: World uncertainty regarding market access conditions. Bank, 2000), is a useful set of tools for those seeking Antidumping, traditionally used by industrial coun- to identify trade facilitation bottlenecks and priori- tries, is increasingly being employed by developing ties. Gerald K. Helleiner (ed.), Non-Traditional Exports nations. As J. Michael Finger explains in Chapter 22, and Development in Sub-Saharan Africa: Experience some of these instruments, especially antidumping, and Issues (Helsinki: World Institute for Development make no economic sense and are best avoided by Economics Research, 2001), provides surveys and developing countries. More efficient instruments are assessments of the instruments used to promote available that are preferable from a development exports in low-income countries and reviews experi- perspective. A key element of such a mechanism is ence with them. Theodore Moran, Foreign Direct that it takes into account the interests of all parts of Investment and Development (Washington, D.C.: society, not just a subset of the domestic industry Institute for International Economics, 1998), offers that confronts competition from imports. This is a an extensive discussion of experience with TRIMs policy area in which further multilateral rule-making and related policy measures. J. Michael Finger (ed.), can be important for developing countries. Howev- Antidumping: How It Works and Who Gets Hurt (Ann er, as Finger notes, domestic actions to improve the Arbor: University of Michigan Press, 1993), brings economic content and rationality of these measures together case studies illustrating how antidumping is are likely to be more beneficial. In the meantime, applied in practice and assessing the implications for exporters have to live with the threat of being con- users and targets. Neil Vousden, The Economics of fronted with contingent protection. In Chapter 23 Trade Protection (Cambridge, U.K.: Cambridge Uni- Gary N. Horlick and Eleanor Shea, two practicing versity Press, 1990), is a good academic textbook on trade lawyers, discuss the relevant U.S. trade law the instruments of trade policy. 95 blank 11 S A M L A I R D Market Access importation of even minimum quantities. . ." Negotiations may be directed toward reductions in Issues and the applied tariffs or the binding of duties. They may concern select- WTO: An Overview ed products or may take place under agreed multilateral proce- dures. They may take account of the individual needs of mem- bers and industries, and flexibil- ity is afforded to developing countries to assist their econom- ic development. The application M of these guidelines, however, arket access negotiations in the depends on their acceptance by the partners in a WTO encompass trade in goods and negotiation, and it can be difficult for any one coun- services. Negotiations on goods--the subject of this try to rely on these guidelines to escape making and subsequent chapters in Part III--are essentially commitments. The results of negotiations are listed concerned with tariff reductions and the elimination in each member's schedule of commitments, or reduction of certain nontariff barriers to imports. recorded in the WTO's Integrated Data Base, which WTO rules covering contingency protection, stan- is not public. dards, and so on are not part of market access nego- The launching of multilateral negotiations has tiations per se, although they can have an important historically been decided at ministerial meetings. effect on the conditions of market access.Acceptance Modifications of scheduled concessions, however, of improved WTO rules can contribute to the secu- do not have to wait for a round to be launched but rity and predictability of market access. This chapter can be negotiated under the provisions of Article provides an overview of the key market access issues XXVIII of the GATT 1994 with those WTO mem- that confront developing countries. bers with which the concessions were originally negotiated, as well as other members deemed to have a "principal supplying interest." Such negotia- Tariff Negotiations tions are subject to consultation with any other Under the provisions of Article XXVIII bis of the member with a "substantial interest" in the product GATT 1994, tariff negotiations in the WTO are car- or products concerned. ried out "on a reciprocal and mutually advanta- The common belief that tariffs are unimportant geous basis" with the aim of achieving "substantial for future negotiations stems from the fact that after reduction of the general level of tariffs and other seven rounds of multilateral trade negotiations, charges on imports and exports and in particular industrial countries' import-weighted industrial the reduction of such high tariffs as discourage the most-favored-nation (MFN) tariffs will average 97 S E L E C T E D T R A D E P O L I C I E S A F F E C T I N G M E R C H A N D I S E T R A D E about 3.5 percent when Uruguay Round commit- escalation in industrial countries at the completion ments are fully implemented. But the devil is in the of Uruguay Round implementation. Developing details; there is considerable variation across coun- countries also have relatively high tariffs on labor- tries and across sectors. Simple average tariffs can be intensive manufactures and agriculture. In general, twice as high as import-weighted rates--the higher their tariffs are typically higher than those of indus- the tariff, the less tends to be imported. On the trial countries and also show a pattern of escalation other hand, the existence of various preference (Michalopoulos 1999a). schemes means that even applied MFN rates may As mentioned in Chapter 6, by Hoekman, in this overstate the tariffs on much trade. volume, what matters in the WTO is the level at Particularly high tariffs and tariff peaks (several which tariffs are bound. In the case of developing times the average and, in some cases, well over 100 countries, bound rates are often much higher than percent) prevail in some sectors. Many of these high the applied rates. For example, WTO members from rates are in areas of export interest to developing North Africa and the Middle East have bound rates countries--textiles and clothing, footwear, and that average 26.8 percent, whereas applied rates agriculture (Table 11.1). Agricultural tariffs are gen- average 14.4 percent (Table 11.3). This creates a erally higher than those on manufactures. The degree of uncertainty about market access in such impact of tariffication of agricultural nontariff bar- countries. riers (NTBs) in the Uruguay Round was so great in The stylized facts are, therefore, sectoral patterns some cases as to increase average tariffs. of tariffs that remain highly dispersed, with signifi- Industrial countries' tariff escalation, by which cant gaps between applied and bound tariff rates. tariffs are increased at later stages of processing in The reasons for this relate partly to evolution in sec- order to encourage domestic processing, may nega- toral policy and partly to the participation of WTO tively affect industrialization in developing coun- members in negotiations. Agricultural policy in tries. Table 11.2 provides a picture of tariff many countries is rooted in the history of food Table 11.1 Post­Uruguay Round Applied and Bound Rates of Industrial and Developing Economies by Major Product Group (percent) Industrial economies Developing economies Product group Applied Bound Applied Bound 1. Agriculture, excluding fish 5.2 7.2 18.6 19.9 2. Fish and fish products 4.2 4.9 8.6 25.9 3. Petroleum 0.7 0.9 7.9 8.4 4. Wood, pulp, paper, and furniture 0.5 0.9 8.9 10.3 5. Textiles and clothing 8.4 11.0 21.2 25.5 6. Leather, rubber, and footwear 5.5 6.5 14.9 15.4 7. Metals 0.9 1.6 10.8 10.4 8. Chemical and photographic supplies 2.2 3.6 12.4 16.8 9. Transport equipment 4.2 5.6 19.9 13.2 10. Nonelectrical machinery 1.1 1.9 13.5 14.5 11. Electrical machinery 2.3 3.7 14.6 17.2 12. Mineral products; precious stones and metals 0.7 1.0 7.8 8.1 13. Manufactures, not elsewhere specified 1.4 2.0 12.1 9.2 Industrial goods (rows 4­13) 2.5 3.5 13.3 13.3 All merchandise trade 2.6 3.7 13.3 13.0 Note: Weighted averages, excluding trade within free trade areas. The applied rates are those for the base period; the bound rates are those applying after implementation. In some instances this means that the applied rates are higher than the bound rates. Source: Finger, Ingco, and Reincke (1996). 98 Market Access Issues and the WTO: An Overview Table 11.2 Tariff Escalation on Products Imported by Industrial Economies from Developing Economies Product Post­Uruguay Round bound tariff (percent) All industrial products (excluding petroleum) 4.3 Raw materials 0.8 Semimanufactures 2.8 Finished products 6.2 All tropical products 1.9 Raw materials 0.0 Semimanufactures 3.5 Finished products 2.6 Natural resource­based products 2.7 Raw materials 2.0 Semimanufactures 2.0 Finished products 5.9 Source: GATT (1994a). Table 11.3 Post­Uruguay Round Import-Weighted Applied and Bound Tariff Rates (percent) Country group or region Applied tariff rate Bound tariff rate Industrial economies 4.0 4.7 Developing economies 13.1 20.8 Latin America and the Caribbean 10.1 18.6 East Asia and Pacific 9.8 16.6 South Asia 27.7 56.1 Other Europe and Central Asia 9.6 14.9 Middle East and North Africa 14.4 26.8 Sub-Saharan Africa 16.5 19.8 Note: Unweighted averages, excluding trade within free trade areas. The applied rates are those for the latest year available, generally 1997, 1998, or 1999. The data on applied rates cover 96 developing countries and 23 industrial countries. Data on bound rates were available for only 65 developing countries. See Appendix A, Table A.2, of this Handbook for country details on applied tariffs. Sources: WTO, IDB CD-ROM 2000; WTO, Trade Policy Review, various issues; World Bank (2000e). security and a perceived need for self-sufficiency. As undertaken by developing countries in the 1980s a result, agriculture was effectively excluded from and 1990s. It is also an outcome of the increasing negotiations before the Uruguay Round. In devel- prevalence of regional trade agreements (Crawford oping countries high tariffs reflected import-substi- and Laird 2000), as well as the application of unilat- tution industrialization policies. By virtue of eral preferences such as those under the Generalized provisions for special and differential treatment, System of Preferences (GSP), the Cotonou Agree- these countries were not required to make conces- ment (successor to the Lomé Convention), the sions in the early GATT rounds. As a result, they Caribbean Basin Initiative, and special preferences received little in return, so that many of their to improve market access for the least-developed exports continue to face high tariffs. countries. When MFN bound rates are reduced in The gap between bound and applied rates has multilateral negotiations, the value of such prefer- much to do with autonomous reform programs ences is decreased, and this may have led some 99 S E L E C T E D T R A D E P O L I C I E S A F F E C T I N G M E R C H A N D I S E T R A D E countries to resist MFN tariff reductions or to take by 40 percent, although the uncertainty in the degree less interest in multilateral negotiations. Negotiated to which such producer payments are linked to pro- MFN tariff rates, however, are more secure than duction decisions makes such analysis difficult. The preferences, and in the longer term it is desirable to percentage real income gains associated with this adapt industrial structures to freer trade in order to liberalization, reported in the first bar for each coun- benefit from comparative advantage. try or region shown in Figure 11.1, are largest in developing regions such as South Asia (other than India) and Southeast Asia (other than Indonesia). Gains from Further Liberalization Virtually all developing regions except the net food- Developing countries have a large stake in the importing Other Middle East region are expected to achievement of significant agricultural liberaliza- experience overall gains from multilateral reduc- tion. Hertel and others (forthcoming) build a model tions in agricultural protection. The bulk of these of the world economy in 2005--at which time gains derives from efficiency improvements generat- Uruguay Round commitments will have been fully ed in the developing countries themselves (the sec- phased in--and estimate that another 40 percent ond bar in each set in the figure), reflecting the fact reduction in agricultural tariffs and export subsidies that most of the potential gains from liberalization will bring about an increase in global real income of arise from removal of own protection. around US$60 billion per year. This figure increases There has been a sweeping change in the structure by US$10 billion if domestic support is also reduced of international trade in the past two decades. In the 11.1 Implications of a 40 Percent Reduction in Agricultural Trade Barriers Percentage of 2005 income 1.2 1 0.8 0.6 0.4 0.2 0 EIT ROW Japan BrazilIndia China OthSEAAusNZLWEurope OthSSA TurkeySoAfrCU(China) NAmerica Indonesia ­0.2 OthSoAsia OthLatAmOthNICs OthMENA Taiwan ­0.4 Region Real income Efficiency Source: Hertel and others (forthcoming). 100 Market Access Issues and the WTO: An Overview mid-1960s manufactures exports accounted for A simulation analysis of the impact of a 40 per- only around a quarter of developing country cent cut in applied tariffs on manufactures by all exports, and by the early 1980s they had only risen countries suggests that global trade volume would to around a third. Since then, growth has accelerat- expand by about US$380 billion in 2005, or about ed. As of the mid-1990s, the share was about three- 4.7 percent of projected merchandise and nonfactor quarters, and it is projected to go on rising (Figure services trade (Hertel and Martin 2000). The largest 11.2). The share of exports of developing countries efficiency gains (as a share of income) occur in going to other developing countries has also risen developing economies, and the countries or regions sharply as the importance of developing countries where tariffs are highest in the 2005 base (China, in the world economy has increased. Developing Other South Asia, and India) gain the most. countries therefore have a strong interest in seeking further reductions in tariffs on industrial products. Tariff Negotiating Issues and Modalities The average OECD tariff on imports from develop- ing countries is four times higher than on those It is recognized that developing countries should originating in the OECD (Table 11.4), reflecting receive credit for autonomous liberalization under high tariffs on products such as textiles and cloth- their own reform programs. Some countries take ing. Estimates of the implied tariffs paid suggest the view that for developing countries to qualify for that the barriers developing countries face in other such credit, the low, reformed rates need to be developing countries account for more than 70 per- bound. This is particularly difficult in a request- cent of the total tariffs levied on their industrial offer approach to negotiations, where developing exports (Hertel and Martin 2000). countries have relatively small markets and little 11.2 Share of Manufactures in Developing Country Merchandise Exports, Actual and Projected, 1965­2005 Percent 0.9 0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0 1965 1967 1969 1971 1973 1975 1977 1979 1981 1983 1985 1987 1989 1991 1993 1995 1997 1999 2001 2003 2005 Year 1 Agriculture 2 Minerals 3 Manufacturers Source: Hertel and Martin (2000). 101 S E L E C T E D T R A D E P O L I C I E S A F F E C T I N G M E R C H A N D I S E T R A D E Table 11.4 Patterns of Protection in Manufacturing, 1995 Importing region High-income Developing Exporting region economies economies Import-weighted average tariffs (percent) High-income economies 0.8 10.9 Developing economies 3.4 12.8 World 1.5 11.5 Implied tariff paid (billions of U.S. dollars) High-income economies 16 93 Developing economies 23 57 World 40 150 Source: Hertel and Martin (2000). negotiating power. Formula approaches that call for NTBs. One way to prevent such exceptions is to proportionately higher cuts on high tariff rates can agree on a minimal cut on each tariff line. In the help reduce tariff peaks and escalation (see Laird, Uruguay Round agriculture tariff cuts were made 1999b; see also Chapter 53, by Panagariya, in this on the basis of simple averages. Import-weighting volume). A formula approach can also help over- was not a practical proposition because prohibitive come difficulties related to how to grant credit. nontariff barriers mean that some products are not Another option may be to carry out early reduc- imported. tions (that is, before the conclusion of a negotiating Many tariff types are legitimate under the WTO. round) and to make deeper cuts in MFN tariffs on In addition to percentage or ad valorem rates, duties products of particular export interest to developing may be specific (for example, US$1.00 per kilo- countries or to the least-developed countries, as was gram), alternative (US$1.00 per kilogram or 10 per- done on tropical products in the Uruguay Round. cent, whichever is higher), or mixed (US$1.00 per Prior to the Seattle Ministerial, some Asia-Pacific kilogram plus 10 percent). Switzerland is exception- Economic Cooperation (APEC) countries proposed al in that all its rates, other than zero rates, are an "early harvest" of products for accelerated liber- expressed in specific terms. Specific tariffs are often alization (see the analysis by Dee, Hardin, and designed to offset low international prices for the Schuele 1998). Proposals have been made to reduce affected product, in lieu of variable levies, which are very low rates (nuisance tariffs) to zero. But deeper prohibited under the WTO Agreement on Agricul- tariff cuts on raw materials and components than ture.1 This practice is sometimes said to impart a on finished goods can increase effective protection, bias against imports from low-cost suppliers--in producing a perverse result for resource allocation. most cases, developing countries. Ad valorem tariffs Simplification of the structure of tariff rates can would be more transparent. Requiring members to increase transparency and help reduce distortions provide information on the ad valorem incidence of in trade and production, so that a negotiation can other rate types would be useful. be used to restructure sectoral and fiscal policy. Tariff quotas or tariff-rate quotas are tariffs that Most WTO members have a range of tariff rates-- increase above a certain value or volume of imports. typically, zero for raw materials, a low to moderate They are used for agricultural imports of commodi- rate for intermediate products, and higher rates for ties subject to minimum import requirements. For finished goods--but some countries have hundreds example, the duty for the first 1,000 tons of a prod- of different rates. uct imported in a fiscal year may be 10 percent, but If agreed average cuts are import-weighted, as has the duty after the first 1,000 tons might be 50 per- been the case for industrial products in the past, cent.2 Some duties of this kind are also expressed as countries may be able to avoid cutting rates on specific rates. The WTO Agriculture Agreement products subject to prohibitively high rates or does not stipulate how tariff quotas are to be 102 Market Access Issues and the WTO: An Overview administered--that is, who gets to import at the in- Agriculture quota or lower rate and who must pay the higher rate. To put the issue in perspective, among all WTO The main NTBs that directly affect market access members 1,371 tariff quotas are in operation. primarily concern the agricultural sector and Although in some instances the out-of-quota rate is involve subsidies and tariff rate quotas (TRQs). For- not being applied even when imports exceed the mally, under Article IV of the WTO Agreement on quota amount, and average quota fill rates in 1999 Agriculture, market access negotiations are strictly were only about 50 percent, there are estimates that defined as the tariff negotiations, but market access the out-of-quota rate exceeds 100 percent for some will also be directly affected by further reductions in affected products (Elbehri and others 1999). the use of domestic subsidies, which are already covered in the negotiations mandated by the Uruguay Round agreement. Nontariff Barriers To put the mandated negotiations into context, it Strictly speaking, market access negotiations in the is important to understand that before the Uruguay WTO are concerned only with tariffs. In the case of Round there had been little discipline in the agricul- NTBs, which certainly affect market access, the tural sector. As a result of the round, agriculture was main focus of negotiations is in the area of rules, largely brought under the main WTO disciplines. which set conditions for the use of such measures. Import measures had to be eliminated or converted Examples include contingency protection (safe- to tariffs ("tariffied"), and the tariffs were then sub- guards, antidumping, and countervailing mea- ject to progressive reduction commitments, except sures), technical barriers (including sanitary and for rice and some staples that were subject to mini- phytosanitary measures), local content require- mum access commitments--that is, TRQs. It was ments, subsidies, import licensing, state trading, also agreed to reduce the level of domestic support, and rules of origin.3 except for exempted "green-box" policies and de Under the WTO Agreement on Safeguards, vol- minimis amounts. Industrial countries were to untary export restraints were to be eliminated in reduce domestic support (the aggregate measure of return for some flexibility in the use of safeguards. support, or AMS) by 20 percent over 6 years, while There remain, however, some measures with very developing countries were to reduce their domestic similar effects: production restraints (e.g., on alu- support by 13 percent over 10 years. The agreement minum and petroleum), sectoral consultations also included reductions in outlays on export subsi- (automobiles), and the use of price restraints dies (for industrial countries, a reduction of 36 per- ("undertakings") as the outcome of antidumping cent over 6 years, and for developing countries, one investigations. Technically, any discussion of these of 24 percent over 10 years) and in the volume of cases would also come under the rules negotiations, subsidized exports (reductions of 21 percent over 6 one of the functions of which is to ensure that tariff years by industrial countries and 14 percent over 10 liberalization is not undermined by NTBs. years by developing countries). Special safeguards There are a number of areas in which negotia- (increased duties) can be triggered by increased tions are designed specifically to reduce or eliminate import volumes or price reductions (by comparison NTBs rather than establish how they may be used. with average 1986­88 prices expressed in domestic In the case of industrial products, the main NTBs currency). A peace clause, intended to constrain the are currently in the textiles and clothing area, and use of countervailing measures until 2003, is some- these are being eliminated as the sector is progres- times seen as setting a time limit on the current sively integrated into GATT 1994.4 In principle, no negotiations. further negotiations on quota elimination should Before the official launch of the mandated negoti- be required in this sector, but the fact that the main ations in agriculture, work had already begun in the liberalization has yet to take place has given rise to WTO under an exercise on the analysis and fears that the industrial countries may be unable to exchange of information. The formal negotiations meet their obligations under the WTO Agreement are conducted in special sessions of the WTO Com- on Textiles and Clothing (ATC). (For further dis- mittee on Agriculture, which was established by the cussion on textiles and clothing, see Chapter 21, by WTO General Council in February 2000. The terms Hanna Kheir-El-Din, in this volume.) of the negotiations are laid out in Article 20 of the 103 S E L E C T E D T R A D E P O L I C I E S A F F E C T I N G M E R C H A N D I S E T R A D E Agreement on Agriculture and were revised in the Conclusion Doha Ministerial Declaration to acknowledge that "without prejudging the outcome . . . we commit Clearly, much remains to be done to liberalize ourselves to comprehensive negotiations aimed at: access to markets in both industrial and developing substantial improvements in market access; reduc- countries. Further reduction in tariffs remains tions of, with a view to phasing out, all forms of important--the tariff agenda is by no means a mar- export subsidies; and substantial reductions in trad- ginal one. The market access agenda also includes ing-distorting domestic support." Account is also to nontariff measures such as agricultural subsidies, be taken of nontrade concerns, special and differen- tariff rate quotas, antidumping, and restrictive tial treatment for developing country members, and product standards. The market access agenda in the objective of establishing a fair and market-ori- services, not discussed in this chapter, is large (see ented agricultural trading system. Part IV of this Handbook). Making significant Through March 2001, 47 negotiating proposals progress will be a major challenge, yet, given the had been submitted by 125 members. (A full list of magnitude of the remaining barriers to trade in these proposals is given on the WTO Website, and, goods, there is substantial scope for "trade conces- in a sign of new transparency in this area, the docu- sions" in the market access areas (Hoekman 2002). ments themselves can be downloaded.) Briefly, the proposals cover market access (tariffs, tariff quotas, Notes food quality, and special and differential treatment for small economies), export competition (subsi- Helpful comments on an earlier version of the chapter were dies, credits, and export taxes), domestic supports received from Rolf Adlung, Bernard Hoekman, Costas Michalopou- (blue-box, green-box, and transitional issues), non- los, Christopher Moir, and Peter Tulloch. trade concerns, development issues (including spe- 1 Some WTO members assert that only rates in excess of com- cial and differential treatment for developing mitments are prohibited, rather than the systems per se. countries), and state trading. Putting aside tariff issues, nontariff measures, other than "pure" rules 2 Seasonal tariffs are sometimes used to protect domestic agri- cultural production during the growing season. Since the in- issues that impinge directly on market access, season high rates cannot exceed bound levels, they are usually include the use of domestic support, the operations expressed as temporary reductions in the bound MFN rate in of state trading enterprises, and export taxes. Con- the off season. tentious issues relating to domestic support include 3 WTO rules specify that charges related to trade, other than tar- the possible extension of coverage of "green-box" or iffs, are to be based on the cost of the service provided. Other- permitted subsidies to attain "multifunctionality" wise, they may be construed as a tariff and included within the or nontrade objectives; a possible development box tariff commitment. In practice, a number of such charges are that would allow subsidies for economic develop- levied as a percentage of the unit value and are unrelated to ment; and the elimination or reduction of export the cost of the service. Examples include consular or visa fees, port handling charges, customs processing fees, lighthouse subsidies in value or volume terms. Rules issues charges, statistical taxes, and the like. Antidumping and coun- include environmental issues, sanitary and phy- tervailing duties and surcharges for safeguard or balance of tosanitary (SPS) measures, the operations of state payments purposes are not covered by WTO commitments on trading enterprises, the virtual exclusion of agricul- tariff bindings. ture in many regional trading agreements, and the 4 Exporters complain that the main liberalization has been possible elimination of special safeguards. Many of delayed until near the end of the transition period ("back-load- these topics are addressed at greater length in the ing") and that liberalized sectors have been subject to special chapters that follow. In-depth analyses of the agri- safeguards, antidumping measures, and so on. cultural trade policy and negotiating agenda can be found in Ingco and others (forthcoming). 104 12 M A R C E L O O L A R R E A G A F R A N C I S N G Tariff Peaks and da, the European Union (EU), Japan, and the United States. The preferential access grant- Preferences ed by Quad members to devel- oping countries through the Generalized System of Prefer- ences (GSP) and related schemes, as well as through reciprocal trade agreements such as the North American Free Trade Agreement (NAFTA), should, in principle, help devel- oping country exporters over- D come these high tariffs. In espite generally low average most- practice, preferences tend to be limited in that "sen- favored-nation (MFN) import duties, sitive" products are often excluded from the the tariff structure in many industrial countries still schemes or some type of quantitative limitation is contains rates above 100 percent. These tariff peaks imposed. The latter may restrict the amount that are often concentrated in products that are of can be imported under the preferential rates (a tar- export interest to developing countries, including iff rate quota) or constrain the countries that are eli- major agricultural staple food products such as gible (Michalopoulos 1999c; Hallaert 2000). sugar, cereals, and fish; tobacco and certain alco- Tariff peaks are generally defined by UNCTAD holic beverages; fruits and vegetables; food industry and the WTO as duty rates that exceed 15 percent. products with a high sugar content; clothing; and Understanding the prevalence and pattern of tariff footwear. peaks is important for a number of reasons. First, In part, the existence of these peaks and the result- peaks affect commodities that account for a signifi- ing dispersion in tariff rates reflect the fact that, as cant share of total exports from least-developed Finger and Winters note in Chapter 7 of this volume, countries. Second, from a political-economy point developing countries did not participate in the of view, tariff peaks are where the "action" is--they reciprocal exchange of liberalization commitments are the products with the highest protection in the under the GATT. The Uruguay Round of multilater- Quad, and they therefore have the greatest impact on al trade negotiations increased tariff dispersion, as exports to Quad markets. Tariff peaks are among the tariffication of nontariff barriers (NTBs) in agricul- priority trade policy issues that need to be addressed ture led to the imposition of very high duties on in a negotiating context by developing countries. agricultural products that had previously been In large part, the significance of tariff peaks reflects quota-constrained. As a result, tariffs that are more the success achieved in the Uruguay Round in elimi- than three times higher than the average MFN tariff nating NTBs. Only 1.2 percent of tariff lines remain are not uncommon in the Quad economies--Cana- subject to NTBs in Canada; the share is 4.2 percent in 105 S E L E C T E D T R A D E P O L I C I E S A F F E C T I N G M E R C H A N D I S E T R A D E Europe, 2.6 in Japan, and 2.9 in the United States oping countries for these tariff peak products, and (OECD 1997a). These NTBs, however, apply to cloth- the prevailing pattern of developing country ing--a sector that is of great interest to developing exports.1 It concludes with an assessment of the countries, and one that will remain constrained by impact on developing countries of the elimination quotas until 2005 (see Chapter 21, by Kheir-El-Din, in of tariff peaks by the Quad and with an evaluation this volume). In the case of agriculture, although the of the initiatives by some OECD members to grant Uruguay Round led to tariffication of all NTBs (with least-developed countries duty- and quota-free the exception of rice in Japan), tariff rate quotas are access to their markets. Examples include the EU often used; these involve two-tier tariff systems with "Everything but Arms" initiative discussed in Box out-of-quota imports subject to higher tariffs. 12.1 and U.S. actions to provide improved access to This chapter provides a brief description of the U.S. markets to Caribbean and Sub-Saharan African extent and importance of existing tariff peaks in the countries. The challenge will be to extend such ini- Quad, the preferential treatment granted to devel- tiatives to a broader set of poor countries. BOX 12.1 THE EUROPEAN UNION'S "EVERYTHING BUT ARMS" INITIATIVE In February 2001 the European Union (EU) grant- African, Caribbean, and Pacific (ACP) states, was ed duty- and quota-free access for all goods origi- eliminated by the Cotonou Agreement. The EU nating in least-developed countries (LDCs), with can now offer better market access to least-devel- the exception of armaments. The "Everything but oped ACP states without extending it to ACP Arms" (EBA) initiative was enacted by Council countries that are not in the least-developed cate- Regulation 416/2001--amending European gory, as Article 174(2)(b) would have required. Community (EC) Regulation 2820/98--which The EBA, like the existing GSP scheme, also allows applied a multiannual scheme of generalized tariff for diagonal cumulation of origin between the preferences for the period July 1, 1999, to Decem- LDCs, on the one hand, and, on the other, Associ- ber 31, 2001. The amendment extended duty- ation of Southeast Asian Nations (ASEAN) mem- free access without any quantitative restrictions to bers, South Asian Association for Regional 919 agricultural products originating in LDCs; Cooperation (SAARC) members, and the EU. more than half of these items were meat and dairy There are several ways in which the EBA differs products, beverages, and milled products. The from the EU's GSP scheme. First, in contrast to the EBA entered into force on March 5, 2001. GSP, the EBA is not subject to renewal and revision The EBA was adopted as an amendment to the and has no time limitation. The European Com- existing Generalized System of Preferences (GSP) mission will review the functioning of the EBA in scheme to ensure its compatibility with WTO 2005, when amendments can be introduced if rules. The basis for the EBA under the WTO is necessary. Second, new provisions allow the EU to paragraph 2(d) of the Enabling Clause of 1979, introduce safeguard measures when imports of which allows special treatment to be granted to products originating in the LDCs increase massive- LDCs in the context of any general or specific ly in relation to the usual levels of production and measures in favor of developing countries. Thus, export capacity. Specific safeguard measures apply at least from this legal point of view, the EBA ini- especially with regard to sensitive products such as tiative is tied to the existing GSP scheme. This bananas, rice, and sugar should imports cause seri- fact, however, does not impose any constraint on ous disruptions to the EU mechanisms regulating the EU with regard to the scope and nature of the these products, in particular, the Common Agricul- LDC preferential trade regime. tural Policy (CAP) and the ACP-EU protocols. The EU also had to ensure the WTO-compatibil- ity of the EBA by avoiding a constraint imposed by Product Coverage Article 174(2)(b) of the Lomé Convention. This All products are included in the EBA initiative. article, which enjoined nondiscrimination among Only three products are not liberalized immedi- 106 Tariff Peaks and Preferences B O X 1 2 . 1 ( C O N T I N U E D ) ately: bananas, rice, and sugar. Duty-free access withdrawal are manifest cases of unfair trading for these products will be phased in as follows: practices on the part of a beneficiary country or manifest infringements of the objectives of inter- · Bananas. Duties are to be eliminated gradually national conventions concerning the conservation over five years in equal 20 percent annual and management of fishery resources. reductions starting in January 2002. All duties A safeguard clause in Article 28 states that MFN are to be eliminated by January 1, 2006. duties on a product may be reintroduced if that · Rice. Liberalization will occur over four years, product originating from a developing country is starting in September 2006 with a 20 percent imported on terms that cause or threaten to cause reduction, to be increased to 50 percent on serious difficulties to a EU producer of like or directly September 1, 2007, and to 80 percent on Sep- competing products. In examining the possible tember 1, 2008. Elimination of duties is to be existence of such serious difficulties, the European complete by September 2009. During the tran- Commission takes the following factors, among sition period, LDC rice exports will benefit from others, into account: reduction in market share of a tariff rate quota (TRQ). The initial quantities of EU producers, reduction in their production, this quota are to be based on best LDC export increase in their stocks, closure of their production levels to the EU in the recent past, plus 15 per- capacity, bankruptcies, low profitability, and low cent. The quota will grow every year by 15 per- capacity utilization, employment, trade, and prices. cent, from 2,517 tons (husked-rice equivalent) The EBA initiative modifies this scheme by: in the 2001/02 September-to-August market- ing year to 6,696 tons in 2008/09. · Adding to the grounds for the possible tempo- · Sugar. The arrangements for sugar are similar rary withdrawal of preferences massive to those for rice. Full liberalization will be increases in imports into the EU of products phased in between July 1, 2006, and July 1, originating in LDCs, in relation to their usual 2009. During the transition period, LDC raw levels of production and export capacity. This sugar can be exported duty-free to the EU addition will allow the European Commission within the limits of a tariff quota, which will be to "react swiftly when the Communities' increased from 74,185 tons (white sugar financial interests are at stake." equivalent) in 2001/02 to 197,355 tons in · Inserting a new paragraph in Article 28 of the 2008/09. The provisions of the ACP-EC Sugar GSP allowing for the suspension of the prefer- Protocol will remain valid. ences provided by this regulation for bananas, rice, and sugar "if imports of these products Safeguard Provisions cause serious disturbance to the Community While the EBA initiative clearly breaks new ground markets and their regulatory mechanisms." in granting full market access for the least-devel- Here it becomes clear that while the EU is gen- oped countries, it also provides for mechanisms to erally ready to extend preferential market avoid disruptions to the EU market. Under the access to sensitive products, it also wants to EU's current GSP scheme, preferential tariff treat- provide for special safeguards regarding the ment may be temporarily withdrawn, in whole or three most sensitive ones. The Commission in part, in the case of certain activities such as slav- announced that whenever LDC imports of ery, forced labor, export of goods made by prison bananas, rice, and sugar exceed or are likely to labor, manifest shortcomings in customs controls exceed the previous year's level by more than on export or transit of drugs, failure to comply 25 percent, the Commission will automatically with international conventions on money launder- examine whether the conditions for applying ing, fraud, or failure to provide the cooperation GSP safeguard measures are met. required for the verification of certificates of ori- gin. Other circumstances qualifying for such a Source: European Commission (2001): UNCTAD (2001). 107 S E L E C T E D T R A D E P O L I C I E S A F F E C T I N G M E R C H A N D I S E T R A D E Tariff Peaks and Imports in the Quad than 85 percent of the tariff peaks are for industrial Economies products, whereas in the EU and Japan most peaks affect agricultural products (91 percent in the EU Between 6 and 14 percent of Quad tariff lines at the and 77 percent in Japan). The maximum tariff rates six-digit level of the Harmonized Commodity at this level of aggregation in the Quad economies Description and Coding System (HS) are above 15 are, for Canada, 340 percent (butter); for the EU, percent (Table 12.1).2 There are 200 to 300 such 250 percent (edible bovine offal); for Japan, 170 lines in the United States, the EU, and Japan; Cana- percent (raw cane sugar); and for the United States, da has more than 700 tariff peaks. The average tariff 120 percent (groundnuts in shell). in the Quad over all tariff peak products is 28 per- In 1999 the value of Quad imports of products cent, or 4.5 times the unweighted total average tariff subject to tariff peaks was US$92.8 billion. More of 6.2 percent. The highest average tariff for peak than 60 percent (US$55.2 billion) of Quad imports products, 40.3 percent, is found in the EU. (The EU of these products originate in developing countries average for the entire tariff universe is a much lower and potentially face an average tariff of 28 percent.3 7.4 percent.) In the United States and Canada more This represents about 5 percent of total developing Table 12.1 Tariff Peaks and Imports, Quad Economies, 1999 Tariff peak product United All Quad (at HS six-digit level) Canada EUa Japan States economies Number of tariff peak products (MFN rate 15 percent) 732 317 233 307 1,077b Agricultural products 85 290 178 48 364b Industrial products 647 27 55 263 713b Tariff peak products as percentage of all tariff lines 14.3 6.2 4.6 6.1 7.8c Average unweighted MFN tariff rates (percent) Tariff peak products 30.5 40.3 27.8 20.8 28.0 All products 8.3 7.4 4.3 5.0 6.2 Maximum rate 342.7 251.9 170.5 121.0 221.5 Total imports of tariff peak products (billions of U.S. dollars) 8.7 27.1 15.8 41.2 92.8 All preferential and GSP countries 7.6 16.5 4.8 26.3 55.2 Least-developed countriesd 0.09 0.3 0.03 0.9 1.3 Share of tariff peak products in total imports (percent) 4.6 3.4 4.9 4.6 4.2 All preferential and GSP countries 4.8 4.9 2.8 6.6 5.2 Least-developed countriesd 30.2 2.8 2.6 15.0 11.4 Import revenue collection in tariff peak products from world (billions of U.S. dollars) 1.6 8.9 6.3 5.4 22.2 All preferential and GSP beneficiaries 0.7 4.3 1.4 4.6 11.0 Least-developed countriesd 0.02 0.03 0.001 0.2 0.2 Note: MFN, most-favored-nation. a. Excludes all intra-EU trade in world totals. b. Number of nonoverlapping categories. c. This is the simple (unweighted) average across Quad countries. Note that of the 5,032 tariff lines at the six-digit level of the Harmonized System, for 21 percent (1,077/5,032) there is a tariff peak item in at least one Quad member. d. Based on the UN classification of 48 countries. Sources: For MFN tariff, OECD data; for preferences, WTO tariff files; for trade, UN COMTRADE Statistics. 108 Tariff Peaks and Preferences country exports to the Quad. LDC exports are dis- beneficiaries are only 9 percent in Canada, 18 per- proportionately affected by tariff peaks in the Quad; cent in Japan, and 23 percent in the United States. products subject to tariff peaks represent 15 to 30 For LDCs, the margins fall to 25 percent in Canada percent of these countries' total exports to the Unit- and 30 percent in the United States and Japan. The ed States and Canada. EU, by contrast, has a 50 percent margin for GSP beneficiaries and a 70 percent margin for LDCs in tariff peak items. Tariff Peaks and Preferences for Developing Thus, although existing schemes grant significant Countries preferences to developing countries, preferences are Most developing countries enjoy preferential access concentrated in products that already enjoy low tar- to Quad markets, either through unilateral schemes iffs (between 0 and 15 percent) rather than on tariff such as the GSP or through free trade agreements peaks. In other words, preferential schemes offer lit- such as NAFTA or EU association agreements. In tle protection against tariff peaks in the Quad, the cases of Canada, Japan, and the EU around 170 except in the EU.5 developing countries benefit from GSP (or better) preferences.4 In the case of the United States 29 Tariff Peaks and LDC Exports developing countries are excluded from the GSP, so that only 140 developing countries benefit from Total LDC exports in 1999 were US$22.7 billion, of some sort of preferential access. which US$17 billion went to the Quad economies. Preferences granted by the Quad are of a cascad- More than US$5.5 billion of LDC exports to the ing nature: countries with free trade agreements world--25 percent of their total exports--are (FTAs) generally get the best treatment, followed by potentially affected by tariff peaks in Canada; that LDCs and other developing countries (Table 12.2). is, tariff peaks in Canada affect product categories The United States grants preferences to the mem- that account for 25 percent of the global exports of bers of the Andean Pact and the Caribbean Com- LDCs. Most of these affected exports are in apparel munity and to Mexico under NAFTA. For the EU, and clothing (HS 61 and 62). More than 99 percent we report both Cotonou preferences (ACP) and the of LDC exports of apparel to the world are affected FTA preferences granted to Eastern European and by the average tariff peak of 22 percent in Canada. Mediterranean countries. Two different groups of There is almost no preferential access for LDCs in LDC countries are constructed in the EU case: these items (the preference margin is only 8 per- LDCs that are not ACP members, and LDCs that cent). Exports of other developing countries poten- are. For Canada, developing countries are grouped tially affected by Canadian tariff peaks are also into several categories: those benefiting from LDC, concentrated in apparel, and preference margins are GSP, or Caribbean preferences, and Mexico and even smaller, around 3 percent. Chile, which benefit from FTA status. Finally, for Similarly, more than US$3 billion of LDC Japan, developing countries are divided into GSP exports to the world, or 14 percent, is potentially beneficiaries and LDC beneficiaries. affected by tariff peaks in the United States. These On average, these preferential schemes are quite exports are again concentrated in apparel, which generous. In the EU the average tariff faced by LDCs accounts for US$2.6 billion. They face an average or ACP members is below 1 percent, compared with tariff of 19 percent and do not benefit from prefer- the 7.4 percent average MFN tariff. GSP preferences ential access. Tariff peaks in Japan affect about in the EU are lower but still imply more than a 50 US$500 million in LDC exports to the world, and percent margin. Japan and the United States follow tariff peaks in the EU affect about US$800 million. with a 50 percent preference margin under their LDC exports affected by EU tariff peaks are con- GSP regimes and an average 60 percent preference centrated in meat and fish products (HS 16), fish for LDCs. Canada gives a 25 percent preference to and crustaceans (HS 03), sugar (HS 17), tobacco GSP countries and 45 percent to LDCs. (HS 24), and footwear (HS 64). The EU's "Every- Preferences are much less generous for tariff peak thing but Arms" initiative ensures that with the products. Except in the EU, preference margins are exceptions of sugar, rice, and bananas, all of these significantly below the average across all products. exports now benefit from full duty-free access to Preference margins on tariff peak items for GSP the EU. In the case of sugar the preference margin 109 S E L E C T E D T R A D E P O L I C I E S A F F E C T I N G M E R C H A N D I S E T R A D E Table 12.2 Tariff Peaks and Preferential Duty Rates, Quad Economies, 1999 Average unweighted preferential duty rate (percent) Tariff All goods at Preferential trade Number of peak HS six-digit agreements/GSP countries products level Canada United States 1 7.1 1.6 Australia 1 28.2 7.8 New Zealand 1 28.2 7.8 Mexico 1 15.9 3.1 Chile 1 12.2 2.4 Israel 1 11.8 2.5 Caribbean countriesa 18 23.3 4.3 GSP-only beneficiariesb 108 28.2 6.2 Least-developed countriesc 47 22.8 4.4 Other countries (MFN rate) (30.5) (8.3) European Union 15 Eastern Europe and Middle Eastd 30 20.1 1.8 GSP-only beneficiariese 42 19.8 3.6 Least-developed ACP countriesf 37 11.9 0.8 Other ACP countriesg 32 12.4 0.9 Other least-developed countriesh 11 12.6 0.9 Other countries (MFN rate)i (40.3) (7.4) Japan GSP-only beneficiariesj 127 22.7 2.3 Least-developed countriesk 42 19.0 1.7 Other countries (MFN rate) (27.8) (4.3) United States Canada 1 0.6 0.1 Mexico 1 1.6 0.3 Israel 1 0.6 0.1 Andean Pactl 4 14 1.7 Caribbean Communitym 22 13.5 1.6 GSP-only beneficiariesn 80 16 2.4 Least-developed countrieso 38 14.4 1.8 Other countries (MFN rate) (20.8) (5.0) a. Includes 18 Caribbean countries or territories under Commonwealth Caribbean Countries Tariff. b. Excludes eight developing countries: Albania, Aruba, Bosnia and Herzegovina, FYR Macedonia, Mongolia, Oman, Saudi Arabia, and Yugoslavia. c. Excludes Myanmar. d. Includes countries with reciprocal and nonreciprocal trade agreements with the EU. e. Includes most developing economies in Latin America and Asia; excludes Hong Kong (China), Republic of Korea, and Singapore, which are non-GSP economies. f. Includes 37 ACP and least-developed countries under the Lomé Convention. g. Includes 32 ACP countries under the Cotonou Convention but not under the group of least-developed countries. h. Includes 11 least-developed countries that are not ACP members. i. Includes all industrial countries as well as Hong Kong (China), Korea, Singapore, and 14 transition countries. j. Includes 127 countries; excludes Albania, Bosnia and Herzegovina, Estonia, Latvia, Lebanon, Lithuania, FYR Macedonia, Moldova, Viet- nam, and Yugoslavia. k. Excludes three LDCs: Comoros, Djibouti, and Tuvalu. Three others (Democratic Republic of Congo, Kiribati. and Zambia) are included in the GSP group. l. Includes Bolivia, Colombia, Ecuador, Peru, and República Boliviariana de Venezuela under the Andean Trade Preference Act. m. Twenty Caribbean countries covered by the Caribbean Basin Economic Recovery Act, as well as The Bahamas and Nicaragua. n. Includes 80 developing countries or territories under the GSP scheme but excludes 29 other developing economies. o. Based on the UN list of 48 least-developed countries but excludes 10 countries, including Senegal. Source: WTO files. 110 Tariff Peaks and Preferences granted to LDCs is quite small; their exports face an States. Exports from other developing countries average tariff of 29 percent. would fall by some US$1.1 billion, which represents LDC exports to the world that are affected by 33 percent of the total increase in LDC exports but Japanese tariff peaks include sugar (HS 17), raw only 0.05 percent of developing country exports. hides and skins (HS 41), and footwear (HS 64). Of Thus, diversion is significant but does not add up to these three products, sugar is the only one for much, given the small share of LDCs in world trade. which almost no preference is granted. The 5 per- The distribution of changes in export revenue cent LDC preference margin for sugar brings the across products and countries will vary across mar- tariff faced by LDC exporters to 66 percent. Full kets (Figure 12.1). In the case of the European Union, duty-free access is granted for raw hides and skins, 65 percent of the increase in LDC export revenue is while for footwear an 80 percent preference margin concentrated in sugars and confectionery (HS 17), is granted to LDCs, bringing the tariff down to less with the primary beneficiaries being Malawi, Zam- than 8 percent. bia, and Mozambique. The EBA initiative, however, will be applied to LDC exports of sugar only in 2009. In Japan, as well, most of the increase (90 percent) is Effects of Eliminating Tariff Peaks in the in sugars and confectionery. For Canada and the Quad United States most of the increase in exports occurs Tariff peaks are important for all developing coun- in apparel and clothing (HS 61 and 62) and, to a tries, but, as explained above, they are relatively much smaller degree, in footwear, with Bangladesh more restrictive for LDCs. Unilateral initiatives to expected to be the main beneficiary, given its large grant duty- and quota-free access have focused pri- export potential in these sectors. marily on LDCs. Such preferential access will be beneficial to recipients but comes at the cost of Conclusion greater discrimination against non-LDC developing countries. That is, there is likely to be trade diver- The gains from preferential access are conditional sion (see Chapter 55, by Hoekman and Schiff, in on the ability to redirect and expand exports, which this volume). Studies of the impact of granting full, requires the establishment of strong business rela- unrestricted access to LDC exports in Quad markets tionships and a good reputation as a supplier in new suggest that the increase in export revenue could be markets. The benefits of preferential access are also as large as US$2.5 billion, or 11 percent (see Table heavily dependent on the extent to which other 12.3). Most of the increased export revenue for policies that affect market access constrain exports LDCs would be earned in Canada and the United from LDCs. Rules of origin and the threat of contin- Table 12.3 Effects of Granting Duty- and Quota-Free Access to Quad Markets to LDC Exporters (millions of U.S. dollars) European United Quad Canada Union Japan States economies Change in LDCs' exports 1,602) 185) 496) 1,107) 2,497) (7.20) (0.83) (2.23) (4.97) (11.22) Change in GSP beneficiaries' exports ­558) ­100) ­292) ­387) ­929) (­0.03) (­0.01) (­0.02) (­0.04) (­0.05) Change in all developing country exports 1,013) 72) 204) 654) 1,362) (0.03) (0.00) (0.01) (0.02) (0.04) Change in imports by Quad economies 15) 2) 3) 108) 117) (0.01) (0.00) (0.00) (0.01) (0.01) Change in LDC welfare 1,159) 122) 332) 915) 1,694) (0.67) (0.07) (0.19) (0.53) (0.99) Note: Figures in parentheses are percentages of values at the base year (1996­98 averages). Source: Authors' calculations. 111 S E L E C T E D T R A D E P O L I C I E S A F F E C T I N G M E R C H A N D I S E T R A D E gent protection--antidumping, countervailing the MFN tariff. Sapir (1997) has shown that in 1994 duties, and safeguard actions--are examples of only half of total European imports that could policies that can be so used as to greatly reduce the potentially benefit from the GSP entered under this value of duty-free access. Examples abound of pro- preferential regime. The other half paid the MFN tectionist lobbying in Quad economies aimed at duty as a result of the combined effect of rules of tightening GSP rules of origin to restrict the benefi- origin and tariff quotas. The threat of instruments ciaries' ability to significantly expand exports (see of contingent protection can also reduce the incen- Bovard 1991 for some U.S. examples). Rules of ori- tive to undertake investments to benefit from duty- gin that require high levels of local value added can free access. Noteworthy in this regard are the imply that developing countries are forced to pay European EBA safeguard provisions (see Box 12.1). 12.1 Projected Changes in LDC Export Revenues as a Result of Duty-Free Access, by Product and Importer Canada European Union Footwear 1% Headgear 1% Flour; malt and starch 2% (Bgd 55%, Cam 15%, Cpv 9%) (Bgd 96%, Npl 2%, Mdg 1%) (Npl 61%, Bdg 14%, Myr 9%) Residues and food waste 5% Others 5% Other textile Others 2% (Myr 35%, Cgo 17%, Npl 16%) articles 2% (Bgd 80%, Mwi 7%, Meat and Npl 6%) edible meat 10% (Sdn 52%, Mdg 22%, Vut 18%) Cereals 13% (Myr 39%, Mdg 14%, Sdn 13%) Apparel and Apparel and clothing, clothing, knitted 39% not knitted 55% Sugars and (Bgd 65%, Cam 11%, (Bgd 79%, Cam 5%, Myr 4%) confectionery 65% Hti 7%) (Mwi 27%, Zmb 19%, Moz 15%) Japan United States Oil seed and Flour; malt Footwear 2% Meat and misc. grain 1% (Sdn 92%, Myr 2%, and starch 1% (Bgd 55%, Cam 20%, edible meat 1% Afg 2%) (Npl 60%, Bdg 14%, Oil seed and Cpv 9%) Others 0% (Mdg 33%, Sdn 31%, Eth 8%) Vut 28%) misc. grain 2% Others 3% (Gmb 50%, Sdn 38%, Cereals 4% Mwi 3%) (Myr 37%, Moz 25%, Mdg 24%) Tobacco 30% (Mwi 76%, Tza 13%, Uga 3%) Apparel and Sugars and Apparel and clothing, clothing, confectionery 90% not knitted 30% knitted 36% (Mwi 22%, Zmb 16%, Moz 15%) (Bgd 84%, Myr 5%, (Bgd 71%, Cam 9%, Cam 3%) Hti 7%) Note: Afg, Afghanistan; Bgd, Bangladesh; Cam, Cambodia; Cgo, Congo, Dem. Rep.; Cpv, Cape Verde; Eth, Ethiopia; Gmb, The Gambia; Hti, Haiti; Mdg, Madagascar; Moz, Mozambique; Mwi, Malawi; Myr, Myanmar; Npl, Nepal; Sdn, Sudan; Tza, Tanzania; Uga, Uganda; Vut, Vanuatu; Zmb, Zambia. Source: Authors' calculations, based on Hoekman, Ng, and Olarreaga 2001. 112 Tariff Peaks and Preferences Notes 1 Under preferential treatment we include both unilateral schemes such as GSP, Lomé, or LDC preferences and those granted under bilateral agreements, such as NAFTA, Canada- Chile, and the Euro-Med agreements. 2 The WTO and UNCTAD define tariff peaks as all tariff lines above 15 percent (at the HS six-digit level). 3 "Potentially" because tariff preferences granted to developing countries through bilateral or unilateral schemes will bring down the tariff faced by these exporters. 4 The EU was the first customs territory to grant GSP preferences to developing countries, in 1971. For a detailed description of the EU GSP, see Kennan and Stevens (1997); Hallaert (2000). 5 Data on the average MFN import duties on tariff peak prod- ucts at the HS two-digit level, and preference margins granted by the Quad to different groups of developing countries, are provided in Appendix A, Tables A6­A9, in this Handbook. 113 13 L U I S J O R G E G A R A Y S . R A F A E L C O R N E J O Rules of Origin occurred most recently. "Signifi- cant" or "substantial" is defined as sufficient to give the product and Trade its essential character. Rules of origin aim at prevent- Preferences ing what is technically known as trade deflection. This may arise when goods from third coun- tries confront different tariffs in FTA member countries, creating an incentive to bring merchan- dise into the FTA through the member country with the lowest T tariffs and then ship it as a duty- he application of trade prefer- free item to countries in the FTA with higher tariffs. ences, whether unilateral (such as The same incentive is created by GSP regimes for the Generalized System of Preferences, or GSP) or firms located in nonbeneficiary countries. Requir- granted as the result of free trade agreements ing a minimum level of substantial transformation (FTAs), requires guidelines that enable the origin of aims to prevent such practices by limiting the appli- goods to be defined so as to ensure that preferences cability of trade preferences to those goods that sat- benefit only those products originating in the bene- isfy rules of origin. ficiary countries. Preferential trade agreements Origin regimes can result in inefficient produc- therefore include origin regimes that stipulate the tion and discrimination (by favoring the companies provisions and procedures for determining country that are best able to adapt to and satisfy the require- of origin. ments); an unequal distribution of benefits among Commercial exchanges involve goods that are factors of production, activities, and countries; and wholly obtained or produced in the exporting administrative and transactions costs.1 Stringent nation or that contain components from third rules of origin can severely restrict the sourcing of countries. For the second type of merchandise, it is inputs from outside an FTA, thereby leading to necessary to define the conditions, types, and investment diversion--decisions by multinational amounts of imported components that these goods firms to locate production facilities within the can contain and still be considered as originating region. If the region is not large and dynamic, this inside the country or region to which preferences may negatively affect firms' efficiency and competi- have been granted. The general approach taken in tiveness (Barfield 1996; Winters 1997). The opera- most jurisdictions is that the origin of a product is tional and administrative costs of certifying and determined by the location where the last substan- verifying origin are potentially large and can tial transformation took place; that is, the country increase efficiency losses. Net operating costs can be in which significant manufacturing or processing expected to rise with increased administrative com- 114 Rules of Origin and Trade Preferences plexity, lack of transparency, multiple qualification verification and administration equal about 3 per- criteria, and the proliferation of "rules of origin cent of product prices (Garay and Quintero 1997).2 families." In Europe the costs of collecting, manag- The complexity of rules of origin regimes is illus- ing, and storing the information needed for origin trated in Box 13.1. B O X 1 3 . 1 T H E U . S . T R A D E A N D D E V E L O P M E N T A C T O F 2 0 0 0 Gomi Senadhira for SSA countries therefore provides greater security for potential investors. The act also The U.S. Trade and Development Act of 2000 eliminates the competitive-need limitations in contains two important sections providing for the GSP program for AGOA beneficiaries. Final- preferential access to the U.S. market. Title I of the ly, the AGOA countries receive significant bene- law consists of the African Growth and Opportuni- fits for apparel exports. ty Act (AGOA), which extends significant trade benefits to Sub-Saharan African countries. Title II Apparel Provisions of the AGOA contains similar preferences for Caribbean coun- Under the apparel provisions, SSA countries tries; these are not discussed here. would get duty- and quota-free access for: The AGOA recognizes that trade and invest- ment can be powerful tools for promoting sus- · Apparel assembled in SSA from U.S. fabric, tainable economic growth, and it provides a formed from U.S. yarn cut in the United number of market access concessions to the States. countries of Sub-Saharan Africa (SSA). To be eligi- · Apparel cut and assembled in the SSA from ble to receive benefits, an SSA country has to be U.S. fabric formed from U.S. yarn and stitched designated a beneficiary country by the president with U.S. thread. of the United States. Necessary conditions for this · Apparel made from African regional fabric are that the SSA country has established (or is (fabric formed in one or more SSA countries making progress toward) a market-based econo- from U.S. or SSA yarn), subject to a tariff rate my, the rule of law and political pluralism, eco- quota set at 1.5 percent of total U.S. apparel nomic policies to reduce poverty, a system to imports in the 12 months preceding October combat corruption and bribery, protection of 1, 2000, to be increased over the next seven internationally recognized workers' rights, and years by equal increments to a level of 3.5 per- elimination of barriers to trade and investment. cent by October 1, 2007. Beneficiary countries may not engage in activities · Apparel made in designated "lesser devel- that undermine U.S. national security or grossly oped" SSA countries with a 1998 per capita violate human rights. income below US$1,500, from fabric of any Under the AGOA, beneficiary countries origin, subject to the same tariff rate quota. receive preferential access for 1,835 tariff line This would only apply for a period of four items, in addition to the standard GSP list of years. approximately 4,600 items available to all GSP- · "Knit to shape" sweaters from third-country eligible countries. This additional list includes a yarn such as cashmere and merino wool. number of important products that were previ- · Apparel made from fabric or yarns not avail- ously excluded from GSP benefits such as able in commercial quantities in the United footwear, luggage, handbags, watches, and flat- States. ware. These benefits for AGOA-eligible countries · Products that are hand-loomed, handmade, would continue to September 30, 2008, seven or folklore articles. years longer than the present extension of GSP benefits to the rest of the world. In recent years AGOA eligibility alone does not provide a SSA U.S. GSP benefits have been renewed every two country with these benefits. To qualify for these years. The eight-year duration of GSP benefits concessions, a country must establish an effective (continued) 115 S E L E C T E D T R A D E P O L I C I E S A F F E C T I N G M E R C H A N D I S E T R A D E B O X 1 3 . 1 ( C O N T I N U E D ) (textiles) visa implementation system and Impact on Other Developing Countries enforcement mechanism to prevent illegal trans- It is difficult to predict the impact that preferential shipments. As of August 1, 2001, only six coun- access for SSA countries will have on the countries tries had been listed as eligible for benefits under excluded from these arrangements. The market AGOA apparel provisions, but in the following distortions created by prevailing quota arrange- three months five more were added. All but three ments, as well as the caps and other limits on knit (Botswana, Mauritius, and South Africa) are apparel and apparel made from regional fabric or entitled to ship apparel manufactured with fab- third-country fabrics, may limit the adverse impli- rics of any origin.* cations for other developing country exporters. However, a number of countries have indicated to Implications of the AGOA for Beneficiaries U.S. authorities that the trade preferences granted The most significant feature of the apparel provi- to SSA countries under the Trade and Develop- sions of the AGOA is the need to use U.S. fabric or ment Act of 2000 would weaken countries' ability yarn in manufacturing apparel. Given the rela- to compete in the U.S. market. Among these tively higher cost of U.S. fabric, the less efficient countries are the beneficiaries of the Andean transport logistics in Sub-Saharan Africa, and the Trade Preferences Act (ATPA), as well as Asian limited size (or the absence) of apparel industries apparel exporters such as Bangladesh and Sri in most SSA countries, it is unlikely that many Lanka. The concerns of the ATPA countries were countries will benefit from the provisions for taken into consideration in the Andean Trade Pref- export of apparel manufactured from U.S. fabric. erences Expansion Act, introduced in the U.S. Less-developed SSA countries may find it difficult Congress in March 2001. The further expansion to attract investors to take advantage of the duty- of such arrangements would weaken the compet- and quota-free access for apparel manufactured itiveness of countries outside the arrangements. with third-country materials. Although these The smaller countries in Asia, in particular, are countries would receive an average duty advan- concerned that as they move toward the full tage in the U.S. market of around 17 percent, the implementation of the WTO Agreement on Tex- regional cap on such imports and the time limit tiles and Clothing in 2005, they are being placed (the facility expires in September 2004) reduce in a position of competitive disadvantage. the incentive to undertake major investments in the poorest countries in Sub-Saharan Africa. For the relatively more efficient producers that have * Revisions to the act, proposed under AGOA II, may the necessary infrastructure, however, short-term extend this privilege to Botswana. benefits can be significant. Lesotho has emerged The AGOA also appears to have served as a catalyst as the single largest beneficiary, followed by for general expansion of dutiable textile and apparel Madagascar and Kenya. exports from Africa. Criteria for Origin Qualification 1.A change in tariff classification. This involves a requirement that the tariff classification of the Origin regimes define a good as originating inside finished (processed) good differs from that of the an FTA when it is produced or obtained entirely foreign components or materials (from third within the member nations. If it embodies compo- countries outside the integrated area) used in the nents from non-FTA countries, whether the production process. For example, a change in the required levels of substantial transformation have tariff heading--defined at the four-digit level of been achieved has to be determined. Possible crite- the Harmonized Commodity Description and ria for doing so are: Coding System (HS) tariff classification--is often 116 Rules of Origin and Trade Preferences the basis for the preferential rules of origin system Bolivia, Costa Rica, and Colombia and República used in Latin American FTAs, as described in the Boliviariana de Venezuela and for Chile's agree- next section. A problem with the application of ments with Canada and Mexico. The CACM stands this criterion is that the HS was not designed to at an intermediate point between the two. serve as the sole instrument for determining the NAFTA-type FTAs such as the G3 agreement by origin of goods; it is meant to be used for classify- Colombia, Mexico, and República Boliviariana de ing merchandise in terms of other criteria. Venezuela and Mexico's bilateral treaties tend to be 2. A minimum value added threshold; that is, a mini- more comprehensive than the ALADI type in that mum value of national or regional content incor- they cover issues such as investment and public pro- porated in the product. This criterion suffers curement. They also contain more specific and from several shortcomings. It tends to penalize detailed origin regimes. Traditional integration the use of more efficient, cost-saving techniques schemes in Latin America have relied on rules that and is highly sensitive to changes in the factors are less selective and more uniform than those that determine countries' production costs, such found in NAFTA-type agreements, which employ a as exchange rates, interest rates, wages, and work- multiplicity of "rule families" at the tariff item level. ers' fringe benefits. It can also increase the cost of What follows compares the principal features of compliance, given the need for laborious and three regimes that are used as reference frameworks: demanding accounting, operational, and finan- ALADI, NAFTA, and the CACM. cial procedures to be carried out by customs authorities and manufacturing firms. Finally, it The ALADI Regime may sustain imbalances in the distribution of benefits among countries by favoring those with Resolution 78 of the ALADI agreement establishes more vertically integrated production (that is, the general origin regime for the ALADI member industrial nations) and penalizing those with low nations. The basic criterion for origin qualification wages (Garay and Estevadeordal 1996). is a change in the tariff classification in terms of the 3. Use of a specific technical process or of certain com- four-digit level of the HS, or, alternatively, a region- ponents in manufacturing a product. In addition to al content value equal to or greater than 50 percent the technical difficulties of keeping an updated, of the free on board (f.o.b.) cost of the merchandise. comprehensive inventory of the productive This applies to practically all tariff classifications, processes available at any given time--which are with the exception of a group of goods, specially constantly changing--this criterion is discre- negotiated by the member nations, for which specif- tionary because of the absence of classification ic origin requirements apply. These specific require- elements that objectively guarantee the equiva- ments take precedence over the general criteria and lence of different degrees of transformation in the may be more or less stringent than the general rule. production of different goods. Resolution 78 allows differential treatment for rela- tively less developed countries (Bolivia, Ecuador, and Paraguay), for whose exports a lower national or regional content is admissible. One requirement Types of Origin Regimes Used in the of Resolution 78 involves an obligatory certificate of Americas origin, using a special form and issued by a public Regional integration agreements in the Americas or private agency authorized for the purpose by the include the Latin American Integration Association member states. The lack of precision as to compli- (ALADI), the Central American Common Market ance with qualification criteria and certification and (CACM), the Andean Community, the Common administration of rules of origin has, in practice, Market of the South (MERCOSUR), the Caribbean hindered the strict observance of this provision Community (CARICOM), and the North American (Devlin, Estevadeordal, and Garay 1997). Free Trade Agreement (NAFTA), as well as other Although the main elements of the origin regimes agreements signed in recent years. ALADI has of MERCOSUR and the Andean Community are served as a model for MERCOSUR, the Andean similar to those of Resolution 78, there are some Community, and CARICOM. NAFTA has been noteworthy differences. For some goods, the MER- used as a model for Mexico's agreements with COSUR regime demands a 60 percent level of 117 S E L E C T E D T R A D E P O L I C I E S A F F E C T I N G M E R C H A N D I S E T R A D E added value and, in addition, a change in tariff 2.It applies changes of tariff classifications in a heading. When substantial transformation cannot much more versatile fashion than do the other be measured by a shift in tariff classification, the regimes. Classification shifts are not unique for all MERCOSUR regime states that the price of third- tariff classifications but are defined according to country inputs, inclusive of cost, insurance, and merchandise type broken down by chapter, head- freight (c.i.f.), shall not exceed 40 percent of the ing, and subheading and, in some cases, even by f.o.b. cost of the merchandise. Furthermore, MER- tariff item (the HS eight-digit level of disaggrega- COSUR Decision 16/97 sets specific origin require- tion). The different levels of tariff liberalization ments for a list of goods from the chemical, iron are used both to define the required changes in and steel, data processing, and communications classification and to limit their scope by providing sectors, and these requirements take precedence the option of excluding certain tariff levels from over the general criteria. Although the MERCOSUR the main requirements. Slightly more than 40 regime contains no provisions for differential treat- percent of the existing tariff items use a movable ment, agreements with Bolivia and Chile do provide classification shift for determining their origin, for differential treatment in that they set less strin- and a number of these goods also have more than gent requirements for goods from Paraguay and one alternate qualification rule. Bolivia. 3.It uses the regional content criterion for around a The Andean Community has an origin regime third of all items, either alone or, more frequently, similar to that of Resolution 78, and it also admits in combination with one of the other criteria. It special requirements in exceptional cases. In addi- establishes a minimum regional content value of tion, it grants Bolivia and Ecuador preferential 50 or 60 percent, depending on the method, and treatment. The Andean Community used some spe- calculations use the net cost or transaction value cial requirements in the 1970s as part of its import- method. substitution and industrial sector planning 4.It includes concepts not used in earlier regimes, strategies. The Andean Community's origin regime, such as the de minimis clause, accumulation, and established by Decisions 416 and 417 of July 1997, the introduction of self-certification by exporting introduced important provisions regarding origin companies.3 administration. These stipulated in detail the func- tions and obligations of the member countries' The NAFTA approach exhibits much greater competent government authorities in this area and selectivity, specificity, and detail than the general specified procedures for requesting the General Sec- regimes of ALADI or unilateral trade preferences retariat's intervention and guidelines for its deci- such as the GSP. This NAFTA level of detail can be sions. They also detailed the sanctions applicable to seen in the official Mexican bulletin General Rules certification agencies and officers for issuing for the Application of the Customs Provisions of the improper origin certificates and specified the North American Free Trade Agreement, in which the requirements to be met by nongovernmental agen- rules of origin run to almost 100 pages. cies empowered to certify the origin of merchan- dise. Finally, they regulated the criteria and The Central American Common Market Regime procedures for setting specific origin requirements. (CACM) The CACM regime is a combination of the ALADI The NAFTA Regime and NAFTA systems. The main criterion is tariff NAFTA, launched in January 1994, gave rise to a classification change, but it is applied more flexibly new type of regime for origin rules, with the follow- than under ALADI Resolution 78. Instead of being ing elements, among others: applied uniformly at the HS four-digit level, it is measured in terms of changes in chapter, heading, 1.It is a system of specific rules at the tariff-item and subheading. In a number of cases the CACM level that are arrived at by combining some or regime allows exceptions to be made to the primary even all of the three qualification criteria change in tariff heading that is specified. Only with described above; frequently, there is more than regard to some specific goods does it set additional one rule for determining a good's origin. specific criteria, such as regional content and tech- 118 Rules of Origin and Trade Preferences nical requirements. To date, these have rarely been Alternation applied. Use is made of concepts such as the de min- imis clause; there is no provision for differential The regimes also differ in their application of the treatment for less-developed countries. qualification criteria at the level of individual The CACM regime also introduces a series of goods. Alternation is to be understood as the appli- rules and procedures to ensure correct administra- cation of more than one rule in classifying the ori- tion of and due compliance with the rules of origin. gin of a given good. In ALADI, MERCOSUR, the The use of tariff shifts as the basic criterion, but CACM, and the Andean Community alternation is applied differently across the full range of tariff clas- uniform across all tariff classifications, with the sifications, appears to be an attempt to combine additional feature that each rule is based exclusively administrative simplicity with greater detail and on a single qualification criterion: for example, the selectivity in the rules of origin applied to different first criterion is based on a change in tariff heading types of goods. and the alternate one on a specific regional content value. In contrast, NAFTA, the G3, and the Mexican and Chilean bilateral agreements frequently offer a Differences among Systems variety of alternate rules for determining a good's The principal differences among origin regimes origin, without each rule necessarily being based on have to do with whether they follow uniform or dif- a single qualification criterion. ferentiated application of the rules, apply multiple The set of alternate rules applicable at the individ- criteria, and use value-added tests. ual item level is defined as a "rules of origin family," which, at least in principle, should stipulate equiva- lent demands in terms of substantial transforma- Diversity tion. In practice, however, the levels of stringency The three types of criteria used to determine origin within a family differ as a result of the different can be employed uniformly or selectively. Thus, the requirements of the criteria used to determine ori- tariff classification change criterion is applied uni- gin. If there are goods for which the implied degree formly in the ALADI regime at the HS four-digit of transformation varies between the alternate level, regardless of the type of merchandise. In con- applicable rules, de facto inconsistencies and trast, under NAFTA and G3 the required tariff inequalities can arise among different types of com- change varies according to the good in question, panies in the FTA and its member countries. Similar and in different cases a change in chapter, heading, consequences tend to arise when different rules of subheading, or even tariff item may be required. origin families are applied to goods that, in terms of their production techniques or economic nature, are strictly similar, or when a single rules of origin Multiplicity family is used to qualify goods produced through Although the regimes in force in the Americas different processes. include more than one criterion for classifying ori- gin, they differ in the relative weights they assign to Calculation Method each. The origin regimes in MERCOSUR, the CACM, the Andean Community, and ALADI are The method used for calculating regional content basically defined in terms of the tariff classification value varies among regimes. ALADI, MERCOSUR, change criterion or, alternatively, a given level of and the Andean Community require the f.o.b. or regional content; in some exceptional cases a com- c.i.f. transaction value of the merchandise to be bination of criteria is used for specific lists of goods. used in calculating its regional or national content. In contrast, the NAFTA and G3 regimes and some These values are well known, clear, and published, of Mexico's bilateral agreements are based on a mul- and they require neither the exporter nor the cus- tiplicity of criteria, which prevents any one criterion toms authorities to keep special records or employ from being singled out as the guiding principle for additional controls. NAFTA and some of Mexico's determining origin. In part, this multiplicity reflects bilateral agreements use two alternate methods for the high degree of detail and selectivity contained in calculating regional content: net cost and transac- "new-generation" agreements tion value. Estimating the value of regional content 119 S E L E C T E D T R A D E P O L I C I E S A F F E C T I N G M E R C H A N D I S E T R A D E using the net cost method requires detailed records compliance with origin requirements would be less of and information on merchandise promotion and sensitive to evolution in variables external to pro- sales costs. The CACM regime stands midway duction processes themselves. This is also the between these two groups in that it uses two meth- approach that is being pursued in defining origin ods to determine regional content: transaction for nonpreferential trade (see Chapter 14, by Inama, value, defined in accordance with the WTO's Cus- in this volume). toms Valuation Code, and normal price, calculated The chief U.S. negotiator for rules of origin in the from the f.o.b. price of the exported goods and the NAFTA has recommended (a) eliminating the re- c.i.f. price of third-country components. gional content value requirement because of associ- The "new-generation" agreements contain novel ated demands for information storage, processing, concepts aimed at, among other goals, increasing and auditing, which makes it "Byzantine in its com- the flexibility of the tariff classification change crite- plexity"; (b) using simple rules of origin based on rion by introducing de minimis clauses; facilitating tariff classification changes, in particular avoiding the regional integration of production processes by changes at a level of detail beyond the HS six-digit allowing the accumulation of regional components level; and (c) creating "sectoral customs unions" to in calculating regional content values; and stream- allow the elimination of rules of origin in the corre- lining the origin certification process by enabling sponding sectors.4 exporting companies to issue their own certificates. In general, emphasis should be placed on choos- They also specify verification, control, and sanction ing principles aimed at (a) specifying the goal procedures and activities with greater detail and sought by the origin regime; (b) keeping the num- precision --aspects that an origin regime must ber of criteria for determining origin as low as pos- address and that were not dealt with adequately in sible; (c) ensuring consistency between alternate some "first-generation" agreements. Although, it rules of origin and the levels of productive transfor- should be noted, some of these stipulations or inno- mation demanded; (d) maximizing the simplicity vations can increase the cost of administrating the and transparency of procedures for overseeing com- rules of origin for both the public and private sec- pliance with the rules; (e) duly assessing the advan- tors, they do guarantee adequate rigor in the appli- tages of adopting alternate transparent policy cation of the regime. measures, other than restrictive rules of origin, such as prolonging the period over which the market is extended or reducing differentials between the The FTAA and Origin Regimes national tariffs imposed on third countries; and (f) At the Summit of the Americas, held in Miami in ensuring, to the extent possible, consistency with December 1994, it was agreed to begin working the origin regime to be adopted by the WTO (Garay toward the creation of the Free Trade Area of the and Estevadeordal 1996). A key challenge con- Americas (FTAA), with negotiations due to con- fronting negotiators is the multiplicity in the origin clude in 2005. Twelve working groups were set up to regimes that are currently applied in the hemi- analyze common problems associated with an inte- sphere, giving rise to the question of which is most gration project of this size. One of these groups was appropriate and how to progress toward greater charged with studying customs procedures and harmonization. rules of origin. The country representatives in the working group Conclusion agreed that, in principle, changes in tariff classifica- tion should be the basic criterion for determining Given the prevalence of preferential trade arrange- origin, supplemented, as appropriate, by regional ments, in particular free trade agreements, the ques- content value, and allowing for exceptions. Using a tion of rules of origin is of particular policy relatively consistent regime for changes in tariff relevance. In light of the many economic impacts classification across all tariff items, while allowing and the problems in predicting the restrictiveness of for exceptions to be made according to the level of rules of origin, it is essential that clear-cut principles transformation demanded by a specific good's pro- and criteria for determining the origin of goods be duction process, would substantially facilitate the adopted in order to ensure their transparent and administration of rules of origin and ensure that objective application. Rules of origin will tend to 120 Rules of Origin and Trade Preferences vary among FTAs depending on the underlying between FTA members' third-country tariffs are "sensitivity" to intraregional competition and on minimal or when their tariff levels are low. Finally, member countries' strategic goals. efforts should be made to harmonize external tariffs The proliferation of FTAs and GSP regimes has on a sectoral basis in areas where the nature of pro- generated a problem of multiple rules of origin, duction processes and the internationalization of which entail costs of origin administration for both production make administrating rules of origin governments and individual manufacturing and particularly complex. exporting companies and which give rise to ineffi- ciencies in resource allocation and specialization Notes patterns. Efforts to establish basic principles for greater harmonization of the rules applied by FTAs 1 For a more detailed treatment of these issues, see Garay and and those to be agreed on by the WTO should Estevadeordal (1996). therefore be pursued. Although this will be a com- 2 What follows focuses primarily on free trade agreements, but plex task, a number of basic, transparent principles the discussion is equally applicable to preferential trade in gen- eral. for the harmonization process can be applied. In particular, preferential rules of origin should use 3 De minimis is a clause under which a good can be classified as being of regional origin provided that the value of the raw nonpreferential (WTO/World Customs Organiza- materials which fail to meet the tariff classification change tion) rules as a reference point (see Chapter 14, by requirement does not exceed a given percentage of the good's Inama; see also Hoekman and Kostecki 2001). The value. rules should also be as consistent as possible with 4 Presentation by J. P. Simpson, U.S. Department of the Treasury; regard to the classification criterion used. Rules of partially reproduced in Inside NAFTA, vol. 4, no. 6 (March origin should not be used when the differences 1997). 121 14 S T E F A N O I N A M A Nonpreferential determining origin. In the case of preferential rules of origin, if the origin criterion is not met, Rules of Origin the preferential tariff will not be applied; there is no need to fall and the WTO back on alternative methods. In order to administer trade policy Harmonization measures in the case of nonpref- erential origin rules, if the pri- Program mary origin criterion is not met, there must be an alternative method for determining the origin of the good. Thus, other R criteria are needed to define ori- ules of origin have long been con- gin when the primary rule has not been met; cus- sidered a technical customs issue toms administrations have to be able to determine having little bearing on trade policy. Determina- where goods come from. Such ancillary rules to tions of origin may, however, have far-reaching determine origin in cases where the primary rule is implications, with linkages to domestic disciplines not met are commonly referred to as "residual regulating the marketing of products to final con- rules." sumers, the geographic denomination of goods, and the definition of domestic industries. The impact of The WTO Agreement on Rules of Origin rules of origin as a "secondary trade policy instru- ment" can only be fully grasped when they are con- The WTO Agreement on Rules of Origin calls for a sidered in association with the primary policy Harmonization Work Program (HWP) to create a instruments that they support, such as tariffs, con- common set of nonpreferential rules of origin. Pref- tingency protection measures, trade preferences, erential rules of origin are covered by a common and enforcement of health and safety standards. declaration, but they are not subject to the harmo- Rules of origin are often associated with prefer- nization program. Article 1 of the agreement states ential trade regimes, in that satisfaction of origin that nonpreferential rules of origin are to be utilized criteria is a precondition for the application of a to determine the origin of goods for the following preferential tariff. (Preferential trade regimes are purposes: discussed in Chapter 13 in this volume.) Nonpref- erential rules of origin apply to trade flows that do · MFN tariffs and national treatment not benefit from tariff or other trade preferences. · Quantitative restrictions One of the main differences between nonpreferen- · Antidumping and countervailing duties tial and preferential rules of origin is that the for- · Safeguard measures mer must provide for an exhaustive method for · Origin marking requirements 122 Nonpreferential Rules of Origin and the WTO Harmonization Program · Any discriminatory quantitative restrictions and arose regarding how to maintain the integrity of tariff quotas certain trade policy measures or regimes pertaining · Government procurement to a particular product or product area. A recent · Trade statistics. submission by Japan on the relationship between harmonized rules of origin and labeling require- A principal objective of the Agreement on Rules ments for foods identified the problems that may of Origin is to harmonize nonpreferential rules of arise from the implementation of the HWP, as well origin (Art. 9) and ensure that they are applied as other non-WTO agreements such as the Codex equally for all purposes (Art. 3). The agreement General Standard for the Labeling of Prepackaged embodies a built-in agenda for achieving this. Foods. Provisions contained in a non-WTO agree- Although the work program was to be completed ment may exceed what is required for rules of origin within three years of the entry into force of the purposes, since the objectives are different. In the WTO (that is, in mid-1998), this did not happen-- case of the Codex, the aim is consumer safety-- some say, because of the complexity of the issues. something that goes beyond customs administra- The HWP is carried out by a technical committee tion purposes. that works under the auspices of the World Cus- A possible solution to this impasse suggested by toms Organization (WCO) and the WTO Commit- the United States would be to agree that the provi- tee on Rules of Origin. As of mid-2001, negotiations sion on applying rules of origin "equally for all such were primarily conducted in the latter committee, purposes" does not necessarily mean members have as most of the technical aspects of the rules had "to use rules of origin for all such purposes." For become well understood. Article 9, paragraph 2(c) some members, this à la carte approach would facil- states that the technical committee is to develop itate agreement on the HWP. For other members, harmonized definitions of: however, such flexibility may greatly diminish the value of the HWP exercise, as it would impair the "(i) wholly obtained products and minimal oper- legal certainty and predictability that the agreement ations or processes; was designed to provide in this area. Developing "(ii) substantial transformation--based on the countries, and especially least-developed countries, product concerned undergoing a change in may find some attraction in a flexible interpreta- tariff classification subsequent to being tion, as they would then not be obliged to imple- processed (transformed) in the country of ment the results of the HWP. (The HWP might turn export, and out to be excessively complicated for their import "(iii) supplementary criteria, upon completion of requirements and burdensome for their customs the work under subparagraph (ii), in cases administrations to apply.) Exports originating in where substantial transformation cannot be developing countries, however, would remain sub- determined on the basis of a change in tariff ject to the disciplines of the HWP when shipped to heading alone." those WTO members that decide to apply the agree- ment. An alternative procedure suggested in sub- missions to the WTO is to conduct an examination The Issue of "Equally for All Purposes" of the possible implications of the HWP for other Articles 1 and 3(a) of the WTO Agreement specify WTO agreements through communication with all that on implementation of the HWP, WTO mem- the other WTO bodies responsible for the matters bers shall "apply rules of origin equally for all pur- outlined in Article 1. This approach, although in poses as set out in Article 1." Uniformity contradicts principle attractive, would further limit the possi- prior practices of some WTO members, and rules of bility of concluding the HWP work any time soon. origin are a novelty for most developing country members. To date, only 34 WTO members have Impact of Alternative Nonpreferential Rules notified nonpreferential rules of origin. of Origin The possible implications of common rules of origin for the implementation of other WTO agree- The application of origin rules may have unexpect- ments such as that on antidumping have been a fac- ed and unintended consequences for developing tor impeding consensus on the HWP. Concerns country exporters. In many instances, especially in 123 S E L E C T E D T R A D E P O L I C I E S A F F E C T I N G M E R C H A N D I S E T R A D E the agriculture and processed foodstuffs sectors, originating in the region. Thus, domestic producers origin may be attributed to another country as a of dried and seasoned meat could use cheaper result of relatively simple processing. This can have imported meat while retaining origin and labeling as effects on the application of tariff quotas or of sani- high-quality regional products. tary and phytosanitary measures. For example, if a There is a close link between rules of origin and rule of origin regarding manufacture of shoes from geographical indications. (The latter are also cov- shoe parts is based on where assembly operations ered by Articles 22.1 and 23 of the Agreement on are carried out, and if assembly is carried out in Trade-Related Aspects of Intellectual Property many different countries, this will imply the pro- Rights, or TRIPS.) If the HWP applied equally to duction of "originating shoes" across many coun- origin under TRIPS, there might be implications for tries. Conversely, if origin rules state that origin products that are not produced in a specific region depends on specific manufacturing operations such but do meet the origin requirements according to as the making of shoe uppers, from an "origin point the HWP, and vice versa. For instance, some mem- of view" the production of shoes may become more bers have argued that making wine from imported concentrated, possibly facilitating the invocation of grape must (unfermented juice) should be consid- contingent protection measures such as antidump- ered an origin-conferring operation. This proposal, ing. Depending on production and industrial if adopted, could have implications for the protec- strategies, industries will have different incentives in tion of geographical indications under TRIPS Arti- lobbying for alternative rules of origin. cle 23, since some wine producers could argue that To give another example, if a country is a big pro- they are producing originating Bordeaux because ducer and exporter of cotton fabrics, which is com- they are fulfilling the origin requirement laid down monly a "sensitive product" subject to quotas under in the HWP. That is, making wine from imported the ATC, it may have an interest in ensuring that grape must (from France, in this case) would be ori- printing and dyeing are origin-conferring opera- gin conferring. Not surprisingly, traditional wine tions. In that case, all the cotton fabrics exported to producers have opposed this view, arguing that the third countries for printing and dying will change production of wine is origin conferring only if the origin status once they are shipped out of the coun- wine is made from grapes grown and harvested in try. The fabric-producing country's exports will the same country where the wine is made. A com- become less concentrated and specialized, possibly promise proposal has been put forward that consid- reducing the threat that exports of cotton fabrics ers the production of wine an origin-conferring will trigger contingent protection. Conversely, if operation only if the whole process, from grapes to printing and dyeing are not considered to be origin wine, is performed in the same country. This would conferring, even if these operations are carried out allow grapes (but not grape must) to be imported. in third countries this will have no effect on the ori- In some cases a country may have an interest in gin determination of the fabric. "retaining" origin even if the exported product is In some other cases, countries may be interested in processed in a third country before being sold to a "obtaining" origin even if the amount of working final consumer. For example, Colombia argued in and processing carried out is minimal. This can the technical committee that the processes of decaf- occur with agroprocessing and foodstuffs. For feination and roasting were not origin-conferring instance, in discussions in the technical committee, operations. The United States, the European Union one delegation argued that the drying and seasoning (EU), and Japan took the opposite view. If roasting of imported meat was an origin-conferring opera- and decaffeinating are considered to be origin con- tion. The domestic industry involved sold a dried ferring, most of the Colombian coffee roasted or meat product in the domestic market that usually decaffeinated in the EU and the United States could fetched high prices, given consumer perceptions that be marketed as EU and U.S. products. This could this product had a distinctive character. Traditional- severely diminish the image value and marketing ly, the meat used also originated in a particular potential of Colombian coffee as a quality product region. Local manufacturers, however, had begun to with a distinct character and taste.1 use imported meat. If the processing were to confer To sum up, a major issue for countries with origin, the dried and seasoned meat obtained from respect to harmonization of origin rules is to decide imported fresh meat could legitimately be sold as whether to "lose," "retain," or "obtain" origin. The 124 Nonpreferential Rules of Origin and the WTO Harmonization Program difficulty is that this must be done by product or by ary, residual rules should be available to determine categories of product; the best rule for each country origin. Given the existence of multistage, multi- may depend on considerations of industrial strategy country manufacturing operations, failure to pro- or structure at the national and global levels. vide exhaustive residual rules would leave a The relationship between the compilation of loophole in the predictability of the harmonized set trade statistics and harmonized rules of origin is of nonpreferential rules of origin. another issue. Applying the same origin rules for The basic question confronted by the technical both statistical and customs purposes is almost committee is how to determine the sequence of the unprecedented in world trade. In most cases, application of residual rules and their implementa- import statistics are classified according to the tion; that is, how to specify what happens when the country of origin as indicated in the invoice. This is, goods cannot be subject to the primary rule of ori- for the most part, the country of exportation and gin and the residual rules come into play. Two basic not necessarily the origin of the goods for customs approaches have been discussed. One approach, purposes. Moving toward greater consistency supported by the United States, is that if the pri- between customs origin rules and collection of mary rule is not met for a country, then the primary trade statistics could have significant implications rule should be applied to countries farther up the for the measured magnitude of trade flows and production chain to ascertain whether the rule has trade balances. been met in any of them. Only when the primary There is also a strong linkage between customs rule has not been met in any "preceding" country rules of origin and marks of origin that are intend- would the use of residual rules be warranted. This ed to inform consumers of a product's country of can be called the tracing-back option. A second origin. The issue of the relationship between marks approach, supported by the EU, would limit the uti- of origin (how a finished product is to be labeled lization of the primary rule to the country where before being marketed to final consumers) and ori- the last production process has taken place. Thus, if gin for customs purposes is addressed in Article IX the primary rule is not met in the country where the of GATT 1994. A change of the country of origin last production process took place, residual rules will also imply a change in the mark of origin. This should be utilized. can have important consequences for consumer The issue, then, is to assess the potential implica- choice, especially where brand names or goods of a tions of these alternative secondary rules. Under the certain quality are commonly identified with cer- tracing-back proposal, the customs administration tain countries. Environmental or humanitarian would have to trace back, on the basis of the avail- concerns may influence consumer choice toward able documentation, the origin through the preced- products from countries that are recognized as ing countries. In some cases this procedure may be respecting human rights, labor laws, or environ- difficult, as it requires that origin certificates be pro- mental treaties. Although the globalization of pro- duced for the different manufacturing stages the duction has rendered outdated the notion that a finished product has undergone. Commercial con- product is wholly produced and obtained in a par- siderations may also be an impediment to the trac- ticular country, consumers may still identify prod- ing-back method. For developing country ucts of a certain quality with specific countries or exporters, producers, and administrations, the geographic regions. application of this rule demands a certain degree of customs cooperation. Moreover, the provision of relevant information and documentation may Secondary or Residual Rules of Origin require an extensive knowledge of the rules and As noted above, nonpreferential rules of origin are awareness of the possible implications on the part aimed at assigning origin to all goods imported into of exporters, producers, and customs administra- a country. Thus, there must be an origin determina- tions. tion in all cases, as the customs authorities must be Under the EU approach, origin determination able to ascertain the origin of the goods in order to relies to a greater extent on the ability of the cus- administer trade and other policy instruments. If toms administration to determine origin at the time the primary origin criterion (change in tariff head- of importation. If the primary rule is not satisfied, ing or processing requirements) is not met, second- the customs official will immediately have to resort 125 S E L E C T E D T R A D E P O L I C I E S A F F E C T I N G M E R C H A N D I S E T R A D E to general residual rules, that is, value-added crite- will always have to determine origin, apply the rules, ria. This approach also has significant implications, and enforce them. This may be one reason why the as it seems to empower the customs authorities to elaboration of the harmonized rules has involved a make a final origin determination at the time of level of technical detail and sophistication that is importation. almost unrivaled by other WTO agreements. A basic In November 1999 a partial agreement was issue is that the rules of origin that are being negoti- reached on this issue. The tracing-back approach ated are tailored to the industrial and technological was rejected for the most part but was retained in processes used in industrial countries and do not the residual rules and, to some extent, in the list of necessarily mirror the needs, abilities, and resources rules on wholly obtained products in Appendix II of developing countries and their customs adminis- (that is, as an alternative to a value-added criterion). trations. The benefits of transparency and pre- No final agreement has been obtained on the con- dictability of harmonized rules are certainly positive tent and sequencing of the application of these for the multilateral trading system. When translated residual rules. into a WTO commitment, however, they may become an additional burden for administrations that are not adequately equipped. Rules of Origin and Anticircumvention of Moreover, the WTO Agreement is silent on sever- Antidumping Actions al issues related to implementation, such as certifi- Once the HWP has been completed, the rules of cates of origin. For instance, it is not clear whether, origin may help resolve the issue of third-country following the implementation of the HWP, WTO anticircumvention actions in antidumping cases. members may request certificates of origin for each These measures, which became controversial in the import transaction, nor is it clear who should be early 1990s, involve actions against imports of prod- issuing and certifying such certificates. The latter ucts subject to antidumping duties from countries issue has already attracted considerable attention not originally subject to such actions. Such anticir- following a U.S. request for information on aspects cumvention actions were taken on the basis of of the German policy pertaining to certificates of claims that the firms previously found to be dump- origin. Apparently, release of certain goods in Ger- ing had shifted to production facilities located in many and other EU member states was made sub- third countries. As was pointed out by the Republic ject to the presentation of a certificate of origin of Korea, the WTO Agreement on Antidumping is issued in the United States or to other related for- not clear-cut on the issue of rules of origin since, in malities. The United States had no mechanism for Articles 2.2 and 2.5, it refers to both "exporting certifying determinations made by its local cham- country" and "origin country" (see G/RO/W/65). bers of commerce and did not give legal recognition Some countries have argued that the issue could to these certificates or sanction their issuance. It be addressed through the use of harmonized resid- therefore raised a number of questions concerning ual rules of origin in cases of alleged third-country the consistency of this practice with the agreement circumvention, coupled with Rule 2(a) of the Har- and with the needs of a modern economy.3 monized System.2 In the absence of an agreement on the issue of third-country circumvention, a substan- Conclusions tial number of WTO members, including not only the United States and the EU but also Latin Ameri- It is impossible to determine the best rule of origin can developing countries, have unilaterally adopted or the best proposal without being product-, coun- anticircumvention provisions. Thus, nonharmo- try-, and industry-specific. The fact that the harmo- nized, nonpreferential rules of origin continue to be nization process involves some 10,000 specific used to enforce antidumping duties and, conse- products, each involving a certain industrial quently, to combat third-country circumvention. process, further complicates matters. That said, some general guidelines can be drawn from efforts to date to agree on rules of origin. Implementation Issues The most important inputs into the formulation Business life evolves at a faster pace than multilateral of a national position on a specific rule are provided trade negotiations, yet customs and trade officials by domestic producers or by importers and 126 Nonpreferential Rules of Origin and the WTO Harmonization Program exporters. The input of domestic producers is criti- the institutional memory and domestic bodies of cal, as they are the only ones with detailed informa- laws prevailing in the major trading nations. Most tion on the use of imported inputs, the structure of of the implementation burden can therefore be the production chain, their own cost structure, and expected to fall on the developing countries. The the implications of alternative origin rules for their difficulties in meeting the requirements of preferen- competitors. Without this input, trade negotiators tial rules of origin, especially where the issuance of and customs experts cannot arrive at a sound nego- certificates of origin is concerned, should be duly tiating position. taken into account. The HWP should not result in a The question of the impact of "retaining,""obtain- set of unnecessary administrative formalities that ing," or "losing" origin of a product should be care- conflict with recent initiatives on trade facilitation. fully evaluated at the country and subregional levels to better define the possible implications as regards Notes other WTO agreements and the implications for ori- gin marking, statistics, and so on. 1 In this specific case, it appears that Café de Colombia is a pri- Early consideration should be given to the imple- vate trademark owned by a private federation. mentation aspects of the harmonized rules of origin 2 The first part of Rule 2 (a) extends the scope of any heading and especially to the requirement to apply these referring to a particular article to cover not only the complete rules equally for all purposes. This latter require- article but also that article incomplete or unfinished, provided that, as presented, it has the essential character of the com- ment may have decisive and different implications, plete or finished article. The second part of Rule 2(a) provides depending on the country and product involved that complete or finished articles presented unassembled or and on the direction of trade (that is, import or disassembled are to be classified under the same heading as export trade flows). As has been emphasized here, the assembled article. When goods are so presented, it is usu- applying a harmonized set of rules is a sophisticated ally for reasons such as the requirements or convenience of and technical affair requiring a highly trained packing, handling, or transport. administration and an informed private sector. The 3 The matter was discussed at the Committee on Rules of Origin rules that will apply will largely be "inherited" from on May 18, 2001. 127 15 V I N O D R E G E Customs value of the imported goods).2 The incidence of such duties depends on the systems that cus- Valuation and toms administrations adopt for determining the dutiable value Customs Reform of imports. The benefits that accrue to trade as a result of binding of tariffs would be con- siderably reduced if customs officials had significant discre- tion not to use the actual invoice price as the basis for determining P rocedures for determining the dutiable value. The rules that are applied for valua- dutiable value of imported goods tion of goods are of crucial importance in ensuring and for their clearance have been the sub- that the incidence of duties levied is not higher than ject of international negotiations since the early the bound tariff for the good concerned. 1920s. This chapter briefly describes the provisions of GATT Article VII provides that "the value for cus- the WTO Agreement on Customs Valuation, the gen- toms purposes of imported merchandise should be esis of the agreement in the Tokyo Round, and its based on `actual value.'" Article VII allows countries evolution during the Uruguay Round. It then dis- substantial flexibility in defining the "actual value" cusses why many developing countries have been of imports, thus permitting GATT contracting par- reluctant to apply its rules. A key prerequisite for ties to use widely differing valuation practices. In securing universal application of the rules of the 1950, in an effort to achieve greater harmonization agreement and for helping countries that are already of valuation practices, 13 European governments applying the rules is the provision of assistance for developed the Brussels Definition of Value (BDV), modernizing and reforming customs procedures. I under which the price of imported merchandise is argue that while it may be desirable to provide assis- to be determined on the basis of the price of the tance for applying the standards of the revised World merchandise or the price that the "merchandise Customs Organization (WCO) Kyoto Convention on would fetch" if sold on the open market under fully good procedures and practices in customs clearance competitive conditions for export to the country of and control, it may not be possible for some develop- importation. The concept implies that there is a ing countries to accept all these standards.1 "notional price" which can be determined by cus- toms on the basis of the available information, tak- ing into account the conditions and other GATT Provisions and the Negotiating History circumstances relating to the specific transaction of the Agreement on Customs Valuation being valued. Most countries levy customs duties on an ad va- By the beginning of 1970, over 100 countries were lorem basis (for example, a 10 percent duty on the applying the BDV, but Australia, Canada, New 128 Customs Valuation and Customs Reform Zealand, and the United States steadfastly refused to cases in which imported goods were undervalued to join in. These four countries used a "positive" con- reduce the incidence of duties. Underinvoicing, they cept (in contrast to the BDV's notional concept), argued, was more prevalent in developing countries reflecting a desire to restrict customs authorities' because the average level of tariffs in those countries discretion in determining value. These positive was higher, and thus the use of transaction values norms laid down standards that were based on the would lead to a significant loss of revenue. These prices actually agreed on in sale either for export or pleas by developing countries were ignored by the for domestic consumption. industrial nations. The protracted negotiations that In the preparatory phase of the Tokyo Round, the ensued resulted only in agreement on provisions European Union (EU) was the main demandeur for that, among other things, would delay the applica- improvements in the GATT rules on valuation. Ini- tion of the rules by five years for developing country tial negotiations resulted in a tentative ad referen- signatories. dum agreement on Draft Principles and Draft Under the 1948 GATT, countries could decide Interpretative Notes for the uniform application whether or not to join agreements that were negoti- and interpretation of GATT Article VII. The draft ated to supplement or clarify the original GATT texts were based largely on the BDV and its explana- rules. Most developing countries chose not to join tory notes. Developing countries participating in the valuation agreement. As noted in earlier chap- these negotiations fully supported the drafts. These ters in this volume, the situation changed with the texts would have required Australia, Canada, New Marrakech Agreement that established the WTO. Zealand, and the United States to change their sys- The Agreement on Customs Valuation became tems so as to bring them into conformity with the binding on all developing country members. The principles and rules in the draft--something they WTO agreement, however, is different from the were far from willing to do. The situation changed Tokyo Round code in that a provision entitled dramatically at a November 1977 meeting, when EU "Decision Regarding Cases Where Customs Have negotiators announced that the EU had agreed to Reasons to Doubt Truth or Accuracy of the make a fundamental change in its valuation systems Declared Value" was added. This decision shifted by opting "for a positive approach" instead of the the burden of proof: in cases where customs author- "notional approach of the BDV." They argued that ities have reasonable doubts as to the truth or accu- the new proposals were based on what they racy of the transaction value declared by the "believed to be a good feature of the United States importer, the decision allows customs to ask valuation system."3 importers to provide explanations, documents, or The draft agreement provided that in almost all other evidence to establish that "the declared value cases customs value should be determined on the represents the total amount actually paid or payable basis of "price paid or payable" for the imported for the imported goods." That is, it is not up to cus- goods in the particular transaction. This meant that toms to prove that the invoice is inaccurate. If the customs should, as a rule, accept the "invoice price" customs authorities are not satisfied and consider in the transaction being valued. An "extremely limit- that "reasonable doubts" regarding the truth or ed number of cases" were listed for which customs accuracy of the declared value remain, they may use could reject the transaction value. In these cases value an alternative valuation basis as laid out in the was to be determined by using five specified meth- agreement. ods, which were to be applied in the hierarchical order in which they were listed (see the next section). Main Provisions of the Agreement on To developing countries, the change in the EU Customs Valuation negotiating position came as a total surprise. Many had recently changed over to the BDV system, and The basic purposes of the Customs Valuation agree- all of those actively participating in the negotiations ment are to require countries to adopt a valuation had supported the initial EU position that the inter- system that is "fair, neutral and uniform" and to pre- national valuation system should be based on an vent the use of arbitrary or fictitious values. Toward improved BDV.4 They noted that requiring customs this end, the agreement requires countries to deter- to accept the transaction value reflected in invoices mine the customs value of imported goods on the submitted by importers would impede detection of basis of the price paid or payable for export to the 129 S E L E C T E D T R A D E P O L I C I E S A F F E C T I N G M E R C H A N D I S E T R A D E country of importation (for example, the invoice The agreement prohibits the determination of price), adjusted, where appropriate, to include cer- customs value on the basis of the selling price in the tain payments made by buyers, such as cost of pack- domestic market of the goods produced in the aging and containers, assists, royalties, and license importing country; the price of goods in the fees. Buying commissions may not be included in domestic market of the country of exportation; the the transaction value, discounts obtained by sole price of goods for export to countries other than the agents and concessionaires must be accepted, and no country of importation (third-country prices); add-ins or exclusions other that those provided for minimum values; or arbitrary or fictitious values. in the agreement may be made to the invoice price. Because, however, a large number of developing Customs may reject the transaction value by fol- countries use "minimum values" to determine cus- lowing the procedures laid down in the WTO deci- toms duties, particularly for products the prices of sion on shifting the burden of proof when they have which fluctuate widely, the agreement allows devel- doubts about the truth or accuracy of the transac- oping countries to make a reservation enabling tion value declared by the importer. In all such cases them to maintain these values for a reasonable peri- the agreement limits the discretion available to cus- od of time on a limited and transitional basis. toms in deciding on the dutiable value, requiring The agreement requires customs authorities to that the value be determined by applying the fol- accept transaction values not only in cases of lowing five methods, in the hierarchical (sequential) "arm's-length transactions" but also in instances order in which they are listed: where transactions are between "related parties." Such relationships may result, for example, from · Value of identical goods sold for export to the partnership, control by one company of the other, same country of importation or transactions among parent companies and their · Value of similar goods sold for export to the same subsidiaries or affiliates. In all these cases the agree- country of importation ment urges customs authorities not to reject cus- · Deductive value calculated on the basis of the unit toms value unless they consider that the price at which identical or similar imported relationship has influenced the value declared by goods are sold in the domestic market, less appli- the importer. It further provides that customs must cable deductions for costs incurred within the accept the declared value if the importer demon- country of import strates, on the basis of prices charged in transac- · Constructed value computed on the basis of cost tions between "unrelated parties" for identical or of production similar goods or on the basis of deductive or com- · Finally, the "fallback" method described below. puted value arrived at in accordance with the provi- sions of the agreement, that the value declared In the case of identical or similar goods, consign- reflects the correct value of the goods. ments should have been imported at or about the same time as the goods being valued. In cases where Problems and Issues Relating to the value is to be determined on the basis of deductive Implementation of the Agreement value, the unit price of goods sold in the domestic market within a period of 90 days of importation The Agreement on Customs Valuation allows devel- has to be taken into account. oping countries to delay application of the agree- Where value cannot be determined under any of ment for a transition period of five years after the the first four approaches, other reasonable methods entry into force of the WTO. For developing coun- ("fallback") that rely on information available in the tries that were not members of the Tokyo Round country of importation may be used. These should code, this period expired on January 1, 2000. For be consistent with GATT Article VII and with the countries that have acceded to the WTO since 1995, provisions of the Customs Valuation agreement. For the five-year period is counted from the date of instance, rather than determining the value of accession. The agreement provides that the Com- goods by comparison with identical or similar mittee on Customs Valuation may, on request, agree goods imported from the same country of exporta- to grant an extension of the transition period. As of tion, customs could consider the value of goods early 2001, all industrial countries were applying the imported from other countries. agreement. About 23 developing and transition 130 Customs Valuation and Customs Reform economies that acceded to the WTO after 1995 are cal assistance in this area must be broadened and in the process of changing over to the transaction- reoriented to create confidence among assistance- based valuation system, and another 11 countries receiving countries that the adoption of the system have been granted an extension or have submitted will result in full collection of revenue due and will requests for one. For the remaining 25 or so coun- facilitate rapid clearance of imported goods. tries, many of which are in the least-developed group, no definite information is available on their Implementation Issues and Possible Solutions plans to change over from their existing valuation systems to the WTO system, although the transition As noted above, the agreement lays down five stan- period has expired. dards on the basis of which customs value may be There are three major reasons for the difficulties determined when customs authorities decide not to developing countries confront in implementing the accept the transaction value. The first two methods Customs Valuation agreement. First, many coun- call for determination of value on the basis of the tries do not feel any urgency about changing over, as transaction value of identical or similar goods they have no feeling of ownership toward the agree- imported at about the same time. For a number of ment. Most countries were not directly involved in developing countries, particularly least-developed the decisions requiring them to abide by the agree- countries, small economies, and other low-income ment's obligations. As argued by Finger and Schuler countries, imports are spaced over time because of (2000), the agreement was "imposed on them in an the relatively small domestic demand, and there is imperial way with little concern for what it will cost usually a significant gap between import consign- [them to change over from BDV to the new system], ments. Moreover, import trade is often dominated how it will be done or if it will support their devel- by a few importers that determine the countries and opment efforts."5 firms from which they want to import. Conse- Second, and more important, not only do devel- quently, there are greater possibilities for collusive oping countries not see any immediate advantages; deals or informal understandings on prices between they are apprehensive that the changeover may importers and the exporting firms. result in loss of revenue. Customs revenue still gen- In many such cases value cannot be based on erates a relatively high proportion of total revenue prices for identical or similar goods, and the cus- in many developing countries; for some least-devel- toms authorities will have to turn to the other two oped countries it could be 30 percent or higher.6 methods: deductive value (based on price for sale in The possibilities of raising additional revenue the domestic market), and constructed value (based though other direct or indirect taxes to compensate on cost of production). The first is costly for cus- for the loss are nonexistent or limited. toms and is time consuming; from importers' point Third, customs officials are apprehensive that the of view, it could result in delays in customs clear- application of the WTO standards may create prac- ance. The second method requires cooperation by tical problems because of the differences in trading the producer in the exporting country and for this environments and the absence in developing coun- reason will ordinarily be used in transactions tries of computer systems and databases that can be between related parties where the price is deter- used for price comparison purposes. mined on the basis of transfer pricing. These fears are not unfounded. Proposals to make In most cases, therefore, value must be determined the agreement more responsive to the needs of on the basis of the last method--the "fallback" developing countries have been made in the context method. Although no definitive information is avail- of more general discussions in the WTO on the able, it appears that developing countries applying problems of implementing WTO agreements. the agreement use the fallback method to determine To persuade developing countries that have not customs value for a relatively high proportion of yet started to apply the valuation agreement to transactions. Some of these countries have suggested change over to the system, a two-pronged approach that in order to facilitate determination of dutiable at the international level can be considered. First, value under this method, the "residual restrictions" proposals to improve and further clarify the agree- on the type of prices that could be used should be ment should be examined and the necessary removed. In particular, they have proposed that the changes adopted. Second, and more crucial, techni- prohibition on determining value on the basis of the 131 S E L E C T E D T R A D E P O L I C I E S A F F E C T I N G M E R C H A N D I S E T R A D E price of the goods in the domestic market of the available on prices of identical or similar goods exporting country or of the price of goods for export imported from other exporting countries. Although to the country other than the country of importa- the agreement prohibits use of prices charged by the tion be removed (WTO, WT/GC/M/56). exporting firm in third-country markets, the Agree- It is important to note in this context that under ment on Preshipment Inspection permits the use of the fallback method a country may use information such prices (see Box 15.1).7 B O X 1 5 . 1 T H E W T O A G R E E M E N T O N P R E S H I P M E N T I N S P E C T I O N Over 40 countries, many of them in the least- panies should not "arbitrarily impose the lowest developed group, use the services of preship- price upon the shipment." ment inspection (PSI) companies to assist their These provisions, which permit PSI companies customs authorities in detecting undervaluation to base the value they recommend on prices or overvaluation of imported goods and other charged for exports to third countries, conflict customs malpractices such as misclassification of with the provisions of the Agreement on Customs goods with the aim of reducing the incidence of Valuation, which stipulates that customs value customs duties. Such companies physically should not be based on prices charged to coun- inspect the goods to be imported in the export- tries other than the country of importation. The ing country to ensure that they conform to the reasons for this anomalous situation are related to contract terms. They then give their opinion as to the negotiating history of the PSI agreement. At whether the price charged by the exporter the time the negotiations were being held, all reflects the correct value of the goods. countries employing PSI services were using as Exporting enterprises have claimed that in their the basis for their valuation systems the Brussels "clean report of findings," PSI companies often Definition of Value (BDV), which does not pre- recommend prices that are arbitrary. This led to clude use of third-country export prices in deter- the negotiation in the Uruguay Round of an mining customs value. PSI-using countries Agreement on Preshipment Inspection that, argued that the ability of the companies to detect among other provisions, lays down rules which under- or overvaluation would be considerably the governments of PSI-using countries must reduced if they were prevented from using third- require the companies to follow in carrying out country export prices. physical inspections and verifying prices. The Some industrial countries argued against the rules state that in order to ascertain the true value use of such prices on the grounds that this would of goods used for price comparison, PSI firms not be consistent with the rules of the Customs should use prices of identical or similar goods Valuation agreement. In response, the PSI-using offered for export from the same country at countries maintained that since they had not about the same time, or prices charged to differ- acceded to the agreement and did not have ent export markets (third-country export prices). immediate plans to do so, it would not be appro- Where, for price comparison purposes, prices priate to incorporate its provisions into the PSI charged for exports to third-country markets are agreement. used, the rules recognize that exporting firms The compromise solution reached was to add a often charge different prices in different markets, footnote to the provisions on verification of prices reflecting demand and growth potential, as well in the PSI agreement.* The note clarifies that PSI- as factors such as per capita income. The PSI using countries are to be bound by the obliga- agreement stipulates that when third-country tions imposed by the Agreement on Customs prices are used for price comparison purposes, Valuation when using opinions on prices given by the factors responsible for variations in the prices PSI companies for determining customs value. charged to importers in different countries The aim of the clarification is to ensure that cus- should be taken into account, and the PSI com- toms administrations in countries having 132 Customs Valuation and Customs Reform B O X 1 5 . 1 ( C O N T I N U E D ) recourse to PSI services use the prices recom- declared reflect the true value of the goods. Such mended by the PSI companies only as test values importers have a right to expect customs to give or advisory opinions in checking the truth or them the opportunity to produce documentary accuracy of the importer's declared value. Cus- and other evidence to justify their declared price. If, toms could use such recommended prices as test after examination, customs still perceives that the values even when the recommended prices are price declared by the importer involves either arrived at on the basis of the prices charged by under- or overvaluation, it cannot, under the provi- exporters to third-country markets. sions of the Agreement on Customs Valuation, Customs, however, cannot automatically deter- determine the value on the basis of the PSI-recom- mine dutiable value for levying customs duties on mended price. It will have to determine it by fol- the basis of prices recommended by a PSI compa- lowing the methods laid down in the agreement ny. An examination has to be carried out in each for the determination of value when the transaction case. If, on the basis of the examination and a value declared by the importer is not acceptable. comparison of the price declared by the importer and the one recommended by the PSI company, * The footnote reads as follows: "The obligations customs finds that the latter reflects the correct of the Member with respect to the services of the price, and the importer does not contest the find- pre-shipment inspection entities in connection ing, the value can be determined on the basis of with customs valuation shall be the obligations that price. which they have accepted in GATT 1994 and There will always be importers who will contest other Multilateral Trade Agreements included in the PSI-recommended prices that are acceptable to Annex I of the WTO Agreement." The latter customs and maintain that the prices they have include the Agreement on Customs Valuation. Information Issues and Cooperation among tion. PSI companies are also understood to be Customs Authorities preparing to offer such data to countries. Both sources of price data would be available for a fee. The It is increasingly recognized that efficient and effec- ability of customs to benefit from such services tive operation of the WTO valuation system is feasi- therefore depends on how willing and able their gov- ble only if steps are taken to build computerized ernments are to bear the cost of obtaining the data. price databases that enable customs to detect cases An alternative suggested by some WTO delegations of under- or overvaluation and other customs is a formal arrangement under the agreement for frauds. Most industrial countries have computer- mutual cooperation and assistance among customs ized systems that flag invoices showing prices that authorities. This would enable developing country differ from base values derived from previous trans- customs administrations to seek and obtain informa- actions and automatically refer these to the valua- tion on prices from their counterparts in exporting tion directorate for investigation on the basis of countries in cases where this is considered necessary other available data. The customs administrations to determine whether goods are under- or overval- in developing countries must develop such data- ued.8 These ideas are not new; the WCO Nairobi bases and keep them up to date to ensure that cases Convention aims at encouraging cooperation among of undervaluation are detected and the full revenue customs administrations to facilitate action against due is collected. custom fraud. The convention, however, has only Information technology companies are develop- been ratified by a limited number of countries, and ing computerized international centralized data- most industrial countries have been reluctant to bases that could help provide developing country accept these obligations. Some industrial countries valuation units with data on prices that could be argue that the obligation under the convention used as test values in transactions under investiga- applies only in cases where importing countries 133 S E L E C T E D T R A D E P O L I C I E S A F F E C T I N G M E R C H A N D I S E T R A D E requesting information allege that fraud has been reform of customs procedures. Building up the committed, as the industrial country governments comprehensive databases that are required for are otherwise prohibited by law from furnishing ascertaining whether goods are undervalued is not information on prices (WTO, WT/GC/M/56). It has possible without computerization. The adoption of also been argued that because of budgetary con- the ASYCUDA customs software developed by straints on customs departments in industrial coun- UNCTAD would also assist these countries in tries, it would be difficult to provide the required streamlining and reducing customs forms and pro- price information if requests started pouring in from cedures. a large number of countries. Technical assistance programs should recognize The foregoing should not be construed to imply that valuation is only one of the many functions of that countries which have not yet taken steps to customs. The emphasis placed on this function in apply the agreement should wait until solutions are the past in the technical assistance provided by found to these practical problems. All countries are international organizations such as the WTO and obliged to implement the WTO valuation system, the WCO and by the countries providing assistance and the transition period that applied to developing on a bilateral basis was directed toward enabling countries has expired. However, the difficulties dis- countries to fulfill their WTO obligations, not nec- cussed above do point to the need for provision of essarily toward general modernization and reform effective and adequate technical assistance. of customs clearance methods and procedures. Some programs failed even to recognize that with- out computerization and well-developed databases Technical Assistance on prices it would be difficult for customs adminis- Technical assistance to help developing countries trations of countries where undervaluation is wide- improve their customs procedures is provided by ly prevalent to apply the agreement's rules. international organizations and countries on a This is not the case with the assistance provided bilateral basis. Among the international organiza- by institutions such as the World Bank and the IMF, tions that are most active in the customs area are the which have a wider mandate. Such projects, howev- WCO, UNCTAD, the IMF, and the World Bank. er, can involve high costs. The estimated cost of a UNCTAD and the WTO, as well as organizations project in Tanzania, for example, was US$8 million such as the Asia-Pacific Economic Cooperation to US$10 million over three years (Finger and (APEC) and the Commonwealth Secretariat, have Schuler 2000). Covering all least-developed coun- programs that focus specifically on assisting coun- tries, low-income countries, and small states under tries in adopting the WTO agreement. Initially, the comprehensive programs of this kind would involve emphasis of these programs was almost entirely on significant outlays. explaining the rules, supported by case studies and A related question that needs to be addressed is simulations, and on providing assistance in reform- whether programs aimed at assisting countries to ing national legislation. Although such programs apply WTO rules should also provide support for were found useful in improving customs officials' these countries to apply the standards and recom- understanding of the rules of the system, they were mendations of the 1999 WCO Kyoto Convention. far from effective in allaying apprehensions that the The convention is expected to provide a blueprint Uruguay Round decision on shifting the burden of for procedures that customs administrations could proof would not be sufficient to enable customs to adopt for customs control and facilitation of clear- deal effectively with undervaluation or other fraud- ance (see Box 15.2). The WCO is arranging semi- ulent practices. Some of the new programs there- nars and courses to familiarize customs officials fore emphasized on-the-job training at customs with the provisions of the convention and to facili- ports, under which senior customs officials tate its acceptance. Although many developing obtained practical training in how to handle such countries participated in the negotiations to revise cases effectively. the convention, opinion at the national level among In addition to assistance provided under these both customs and trade associations appears to be WTO-focused programs, technical assistance for divided on the extent to which it is possible for least-developed and small economies must provide developing countries with entrenched customs cor- support for computerization, modernization, and ruption to apply all the procedural standards the 134 Customs Valuation and Customs Reform convention lays down. For instance, it is generally dards such as those requiring customs officials to believed that it may be possible for all developing ensure that they carry out physical inspections on countries to adopt and apply standards relating to or verify the prices of consignments only as war- the UN layout key for documentation, payment of ranted by properly conducted risk assessment, or to duties and taxes, cooperation with trade associa- allow "authorized persons" with a record of compli- tions, and the establishment of procedures for ance with customs requirements to clear goods by appealing decisions taken by customs administra- providing minimum information and to pay duties tions. The same cannot be said of some other stan- on the basis of self-assessment. B O X 1 5 . 2 T H E I N T E R N AT I O N A L C O N V E N T I O N O N T H E S I M P L I F I C AT I O N A N D H A R M O N I Z AT I O N O F C U S T O M S P R O C E D U R E S : 1 9 9 9 R E V I S I O N O F T H E K Y O T O C O N V E N T I O N In 1999 the WCO completed a full revision of its and the use of information technology, which 1973 Kyoto Convention. The goal of the changes provides the key to keeping procedures simple was to provide customs administrations with a while ensuring adequate customs control. modern set of uniform principles for simple, No reservations may be entered against the effective, and predictable customs procedures standards and transitional standards set forth in that also achieve customs control. The revised the General Annex. In recognition, however, of convention is intended to be the blueprint for the fact that many countries may not be able to standard and facilitative customs procedures in apply a number of the standards immediately, the 21st century. This revision was necessary the revised convention provides a transition peri- because of the radical changes in trade, trans- od during which present and new contracting port, and administrative techniques that have parties are to make any necessary changes in taken place since the original adoption of the their national legislation. Contracting parties convention. have up to three years to implement the stan- dards and five years to implement the transitional General Annex standards. The General Annex contains the core procedures and practices for clearance of goods that are Specific Annexes common to all customs procedures. Acceptance The revised convention has 10 Specific Annexes of the annex is an obligatory condition for acces- containing a total of 25 chapters and dealing sion and implementation by contracting parties. with the following customs procedures: The annex, which contains 10 chapters, covers areas relating to: · Annex A. Formalities prior to the lodgment of the goods declaration; temporary storage of · Clearance of goods goods · Payment of duties and taxes · Annex B. Clearance for home use; reimporta- · Customs cooperation tion in the same state; relief from import · Information to be supplied by customs, and duties and taxes appeals in all customs matters · Annex C. Outright exportation · Areas of concern to both customs administra- · Annex D. Customs warehouse; free zones tions and the trading community. · Annex E. Customs transit; transshipment; cab- otage The annex also deals with customs control, · Annex F. Inward processing; outward process- including risk management, audit-based controls, ing; drawback; processing of goods for home administrative assistance between customs use administrations and from external organizations, · Annex G. Temporary admission (continued) 135 S E L E C T E D T R A D E P O L I C I E S A F F E C T I N G M E R C H A N D I S E T R A D E B O X 1 5 . 2 ( C O N T I N U E D ) · Annex H. Customs offences sions of the General Annex and the Specific · Annex I. Travelers; postal traffic; means of Annexes on the request of contracting parties. transport for commercial use; stores; relief consignments Acceptance of the Protocol · Annex J. Rules of origin; documentary evi- The revised convention will be put into force by a dence of origin; control of documentary evi- protocol of amendment. Forty of the current con- dence of origin. tracting parties have to accede to the protocol for it to enter into force. As of July 2000, five con- Management Committee tracting parties had acceded to the protocol, and A management committee, which is required to nine had signed it, subject to ratification. Many meet at least once every year, will administer the other contracting parties are carrying out the revised convention. This will ensure that the con- necessary internal consultations and legislative vention's provisions are kept up to date and, if processes for their accession to the protocol. The necessary, are revised at appropriate times to WCO secretariat is conducting technical assis- meet the needs of customs and trade. The man- tance missions and regional seminars to promote agement committee is also empowered to extend the revised convention and to assist contracting the time periods for implementation of the provi- parties with accession. The procedures and practices mentioned above These considerations suggest that the decision as to are increasingly being adopted by industrial coun- which aspects of the Kyoto Convention a country tries. A number of middle-income developing coun- will adopt must be left to its own discretion, taking tries that have been able to reform and modernize into account the trading environment and realities their customs and reduce or contain corruption have prevailing in the country. also adopted or are considering adoption of the Discussions in the WTO on possible work on trade Kyoto norms for physical inspection and verification facilitation have included suggestions by industrial of prices on the basis of risk assessment techniques. countries that new WTO rules in this area include a Not all countries, however, should be required to do binding obligation on WTO member countries to this. As Finger and Schuler note, "Where tariffs are accept and adopt the standards laid out in the revised high, and where accounting experience and access to Kyoto Convention. Many developing countries have electronic information is limited, shifting to a risk- argued that the limited progress achieved to date in based valuation system that depends on in-depth computerization and modernization of their cus- examination of a sample (15 to 20 percent) of ship- toms procedures makes it difficult for them to accept ments might increase rather than reduce the number any binding obligations. The proponents of a trade of shipments on which importers attempt to under- facilitation proposal have therefore revised it to state invoice. Traders might view the change as giving that the aim of WTO negotiations should be to clari- them a better not a worse chance to get away with fy and elaborate the GATT articles which lay down under-invoicing" (Finger and Schuler 2000: 13). rules relating to trade facilitation, taking into account Similarly, in countries where there are widespread the relevant standards in the revised Kyoto Conven- doubts in the public mind about the integrity of cus- tion and in the instruments adopted by other inter- toms officials, governments may find it difficult to national organizations.9 adopt the practice of designating authorized compa- nies. The general public, as well as importers who Conclusion have not been able to secure authorized status, would always look with suspicion at any such deci- Two lessons can be drawn from the experience with sions and allege that the firms had secured author- past negotiations in the customs administration ized status by corrupt means or political contacts. and valuation area. The first is confirmation of the 136 Customs Valuation and Customs Reform basic contention of public choice analysts that in Notes multilateral trade negotiations on rule-making, The author is thankful to Janet Chakarian-Renouf and Markus Jelito rational actors are not always fully informed about of the WTO secretariat for factual verification of the text and con- the economic and noneconomic factors that make it structive suggestions and to Vandana Aggarwal, Habib Ahmed, S. necessary for other countries to adopt administra- Palayathan, Simone Rudder, C. Satapathy, and K. J. Weerasinghe tive rules and practices different from their own. for helpful comments and discussions. Overcoming such "rational ignorance" about condi- 1 The next section draws on Rege (1999). tions in other countries is possible, but it involves 2 Alternatives are specific duties (for example, US$1 per kilo- costs. Industrial country negotiators are willing and gram or per liter) and mixed duties combining ad valorem and able to incur such costs by arranging discussions on specific rates (for example, 10 percent on the value of the a bilateral basis and, where necessary, by financing good plus US$2 per kilogram). joint studies with a view to improving their under- 3 GATT, MTN/NTM/W/126, November 1977. The proposal standing of the procedures followed in counterpart reflected a secret understanding with the United States. countries with similar bargaining positions. They 4 The new EU proposals made irrelevant papers prepared by the do not generally make such efforts if the countries GATT secretariat and UNCTAD under technical assistance pro- expressing concerns about the proposed rules are grams explaining the provisions in the text agreed on an ad those with less political and economic clout and referendum basis and the advantages that would flow from its weaker bargaining positions and are on the periph- acceptance. The Customs Cooperation Council (CCC, now the ery of negotiations. This puts the onus on develop- World Customs Organization), which had so far relied on the ing country negotiators to educate the main players EU to develop an international valuation system based on the BDV, felt that it had been let down with no warning. During regarding the conditions prevailing in developing the subsequent two years, the CCC was concerned as to countries that make it difficult for them to imple- whether it would have any role to play in the customs valua- ment the proposed rules. (For more detailed discus- tion field. In the end, these concerns were met by establishing sion, see Rege 1999.) A major challenge for (under the Customs Valuation agreement) a technical commit- developing country negotiators in a future round of tee in the CCC responsible for the examination of the techni- cal matters relating to customs valuation. negotiations is to improve the knowledge and understanding not only of bureaucrats and negotia- 5 Finger and Schuler (2000: 523) also observe that "the customs tors but also of the general public in industrial valuation agreement . . . provides neither appropriate diagno- sis nor appropriate remedy. It addresses only a small part of countries regarding their situation. Developing least developed countries' problems with customs administra- countries must be able to explain why it may be pre- tion and . . . provides no remedy over other parts. Over the mature for them to adopt procedural rules and small part of the problem it covers, it provides an inappropri- practices that are applied effectively and efficiently ate remedy, one incompatible with the resources they have at by industrial country administrations with access to their disposal." highly computerized systems and other modern 6 See also Kubota (2000), who notes that the average share of information technology tools and with effective import duties in total revenues was 25.9 percent for low- mechanisms in place for controlling corruption. income countries, 15.5 percent for low-middle-income coun- A second important lesson is that unless the rules tries, 16.1 percent for high-middle-income countries, and 0.5 percent for the OECD countries. that are adopted are responsive to the needs of the countries that must implement them, persuading 7 The rationale for these rules is based on the premise that these countries to apply negotiated rules by provid- determination of value by the customs administration would be facilitated and costs for importers would be reduced if cus- ing technical assistance only results in misdirection toms authorities based decisions on the information available of financial resources given by the donor countries in the importing country. Since, however, preshipment inspec- and of the time and manpower resources of the tion companies verify prices using information in the exporting assistance-receiving countries. In order to build countries, the Agreement on Preshipment Inspection (PSI) confidence and improve the overall credibility of requires that the companies rely only on information available the WTO system, it is necessary for WTO members in the country of export. This explains why the Customs Valua- tion agreement prohibits use of prices charged by exporters to to give high priority to resolving the problems and third-country markets for valuation purposes, while the PSI issues that have arisen in the implementation of agreement permits it. Developments in information communi- WTO agreements. cation technology have eliminated some of the difficulties that 137 S E L E C T E D T R A D E P O L I C I E S A F F E C T I N G M E R C H A N D I S E T R A D E previously confronted customs administrations in obtaining information from exporting countries. 8 The establishment of such an arrangement would also help countries using preshipment inspection companies secure such information and reduce their dependence on the companies. 9 The relevant GATT articles include Article X, Publication and Administration of Regulations; Article VII, Fees and Formalities connected with Importation and Exportation; and Article V, Freedom of Transit. 138 16 B R I A N R A N K I N S TA P L E S Trade tion initiatives. The complexity of international trade is increas- ing, as well. Whereas previously Facilitation: the primary objective of inter- national corporations was to Improving the identify the most cost-effective location for producing a fin- Invisible ished product, today the intense pressures of globalization have Infrastructure forced multinationals to identify the most cost-effective locations for the production of the sub- assemblies that go into a final N product. These commercial pat- ational customs authorities are hav- terns have led to an explosion in regional free trade ing to process ever-higher volumes agreements around the world. Such agreements of trade with the same or a declining number of usually feature complex and confusing rules of ori- employees, at the same time as traders are demand- gin, placing considerable additional administrative ing faster clearance for their goods and increased burdens on both the public and private sectors. administrative efficiency in all programs related to In addition to the greater volumes and complexity international imports. The Internet and e-com- of trade, there is the issue of trade velocity. Product merce are transforming shipments that once would life cycles are now measured in months, not years, have been transported in a single container and and modern supply chain management techniques cleared on a single entry into dozens of individual- have increased the use of "just-in-time" manufactur- ized shipments, each requiring separate customs ing, global production sharing, and outsourcing. documents and clearance procedures. This is a Trade now moves, and must move, at higher speeds major challenge for customs authorities and for than ever before. In this environment, businesses express couriers with customers who expect goods simply cannot afford to have imported or exported to be cleared immediately. goods tied up for weeks or even days because of Not only are trade volumes growing steadily; unnecessary or antiquated trade formalities. The because of foreign direct investment, there is also interaction of all of these issues and factors has led to dramatic growth in trade transactions between heightened awareness of the importance of trade related parties. Imports and exports increasingly facilitation in attracting trade and investment. take place between the same corporate entity. The result is to heighten the visibility of unnecessary Trade Facilitation: The Plumbing transactions costs to transnational corporations that are under intense competitive pressure to What exactly is trade facilitation? Although initia- reduce these costs, primarily through trade facilita- tives such as improvement of transport infrastruc- 139 S E L E C T E D T R A D E P O L I C I E S A F F E C T I N G M E R C H A N D I S E T R A D E ture, trade liberalization, and trade promotion do, Chapter 14, by Inama). These agreements deter- in a sense, facilitate trade, they do not constitute mine the tariff classification of an imported prod- what is known today as trade facilitation. Rather, uct, its country of origin, and its value for duty trade facilitation involves reducing all the transac- purposes. The HS is a legal and logical international tions costs associated with the enforcement, regula- product nomenclature developed through the Brus- tion, and administration of trade policies. Trade sels-based WCO and introduced by international facilitation, by its nature, is technical and detailed. It convention on January 1, 1988. Contracting parties has been referred to as the "plumbing" of interna- have committed to apply the HS uniformly at the tional trade. six-digit level, which covers 1,241 headings. The objective of trade facilitation is to reduce the Although the HS is primarily designed for tariff cost of doing business for all parties by eliminating classification purposes, it is also used extensively to unnecessary administrative burdens associated with determine the goods subject to import and export bringing goods and services across borders. The controls, freight tariffs, the application of (or means of achieving this objective are the modern- exemption from) value-added tax regimes, trade ization and automation of import procedures to statistics, and origin. The HS provides a common match established international standards. The "trade language" for all public and private actors in meaning and implications of trade facilitation have the international trade arena, and the concept of changed since as recently as a few decades ago. Trade trade facilitation is dependent on a complete under- facilitation in previous generations revolved around standing and mastery of the HS. attempts to harmonize different regimes or to Other WTO agreements that have an effect on the establish an element of mutual recognition between trade facilitation agenda include: different customs and related policy regimes. Today, the WTO and various World Customs Organization · The Agreement on Import Licensing Procedures, (WCO) conventions set forth a common set of which is designed to ensure that the process of international standards or customs good practices administering import-licensing systems is for all countries. The current challenge is, much nondiscriminatory and neutral and does not more than in the past, one of implementation and restrict trade of convergence in procedures and customs opera- · The agreements on Technical Barriers to Trade tions based on these international norms. From a and on Sanitary and Phytosanitary (SPS) Mea- trade facilitation viewpoint, asymmetrical customs sures, which are discussed in Chapter 41, by regimes create uncertainty and, therefore, costs for Wilson international traders. · GATT Articles V (on freedom of transit), VIII It should be emphasized that use of the term (calling for the simplification of fees and formali- "trade facilitation" does not imply abandonment of ties related to the importation and exportation of efforts to improve trade compliance. On the con- goods), and X (requiring the timely and compre- trary, trade facilitation, using modern risk analysis hensive publication of all laws, guidelines, and techniques, allows compliant importers access to decisions that may affect imports or exports and improved, automated import procedures and gives the establishment of judicial or administrative tri- customs authorities the opportunity to concentrate bunals to review customs administration and resources on noncompliant traders. decisions) · The Agreement on Preshipment Inspection (PSI). International Agreements and Programs Trade facilitation became the subject of WTO dis- The World Trade Organization. A number of inter- cussions in 1997, following the 1996 (Singapore) national agreements have direct implications for ministerial meeting. At the Doha ministerial meet- trade facilitation. In the WTO these fundamental ing in 2001, it was decided to launch negotiations in tools of international trade regulation include the 2003, subject to consensus on the modalities of Harmonized Commodity Description and Coding negotiation. Any such negotiations will have to System (HS), the Customs Valuation agreement draw on the work of numerous other specialized (discussed in Chapter 15, by Rege, in this volume), bodies, the most important of which are discussed and the Agreement on Rules of Origin (discussed in briefly below. 140 Trade Facilitation: Improving the Invisible Infrastructure The World Customs Organization. Of all the interna- porary Admission (1993), which deals with tempo- tional organizations, the WCO's activities and man- rary admission of goods, means of transport, and date are the most closely aligned with the issue of animals. trade facilitation. The WCO, founded in 1953, has a United Nations Agencies. The United Nations membership of 142 economies. Its objective is to Conference on Trade and Development (UNCTAD) increase the efficiency and effectiveness of customs is actively involved in trade facilitation and encour- administration around the world by reviewing the ages the input and participation of developing technical aspects of customs programs and sharing economies in trade facilitation initiatives. UNC- the results of these studies cooperatively with cus- TAD's Automated System for Customs Data and toms administrations. A major international con- Management (ASYCUDA), a customs software pro- vention designed to promote the standardization gram, is used in more than 70 developing countries. and simplification of customs procedures world- ASYCUDA simplifies and automates customs func- wide is the International Convention on the Simpli- tions with a view toward increasing revenue collec- fication and Harmonization of Customs Procedures tion, speeding clearance of cargo, and improving (the Kyoto Convention), which contains general data collection and dissemination. Much of UNC- provisions and special annexes dealing with cus- TAD's trade facilitation activity has involved the toms procedures. The convention, originally estab- transport sector. Transport initiatives include port lished in 1973, underwent a major revision in 1999, development; development of an electronic trans- resulting in improved provisions for automation, port management tool, the Advance Cargo Infor- electronic commerce, postentry or audit-based mation System (ACIS); and the concept of national reviews, and risk management techniques. The trade and transport facilitation committees that revised Kyoto Convention is the most comprehen- bring together all transport stakeholders within a sive existing instrument for promoting internation- country to create and promote policies which al trade facilitation (see Chapter 15, by Rege). Many enhance the efficiency of trade facilitation. in the private sector would like the WTO to eventu- In October 1994 the UN hosted a ministerial-level ally incorporate the Kyoto Convention, or at least International Symposium on Trade Efficiency, held related principles, into its structure, thereby making in Columbus, Ohio. The focus of the symposium such provisions binding and enforceable. was customs procedures and other microeconomic The WCO maintains a customs reform and mod- features that prevent full realization of the potential ernization program (CRM)--an evolving technical trading benefits negotiated in the WTO. At the sym- assistance product that supports customs reform posium, a set of detailed recommendations, referred through training in diagnostic study and in customs to as the Columbus Declaration, was adopted. These needs analysis. The program helps domestic cus- recommendations have become critical guidelines toms authorities implement the required changes in the pursuit of trade facilitation. The symposium that have been identified and evaluate their impact also established UNCTAD's Trade Point Global Net- on trade facilitation and customs compliance. work, a program that aims to create approximately Another major instrument is the 1993 Declaration 180 "trade points" in 109 countries. These trade Concerning Integrity in Customs (the Arusha Dec- points will be electronically linked to national cen- laration), which addresses the issue of corruption ters for trade facilitation and will act as providers of within customs administrations. The Arusha Decla- trade-related information and data. ration is indirectly linked to the CRM and the Another relevant UN body is the Center for Facil- revised Kyoto Convention in that it promotes stan- itation of Procedures and Practices for Administra- dardized customs procedures, electronic commerce, tion, Commerce, and Transportation (CEFACT- and improved relations between customs brokers UN/ECE). Since 1960, this organization has pur- and customs. sued the harmonization and automation of customs The WCO is responsible for literally dozens of procedures and information requirements and it additional programs, guidelines (such as the issued the internationally recognized UN/ECE Trade Express Consignment Guidelines), resolutions, Facilitation Recommendations. CEFACT is perhaps norms, recommendations, and conventions best known for its work on electronic data inter- (including the HS). A particularly important instru- change (EDI), a form of electronic commerce that ment is the Istanbul Customs Convention on Tem- uses a structured exchange of data between two par- 141 S E L E C T E D T R A D E P O L I C I E S A F F E C T I N G M E R C H A N D I S E T R A D E ties, and for the development of the UN Electronic customs operations, the free flow of trade and of Data Interchange for Administration, Commerce, enforcement information, and a common and Transport (EDIFACT). EDI and EDIFACT have approach, wherever possible, to customs valuation. become important instruments for reducing cus- Trade facilitation has also figured on the agenda of toms paperwork and exchanging trade-related the Group of Seven countries; an example is an information between parties that typically handle effort in 1997 to standardize and simplify customs international trade transactions (for example, insur- procedures. ance firms, customs, freight forwarders, and customs Trade facilitation is a major focus of the Free brokers). In addition, the United Nations Commis- Trade Area of the Americas (FTAA), as well. The sion on International Trade Law (UNCITRAL) has FTAA has developed a series of recommendations, developed significant conventions on the interna- known as business facilitation measures, that incor- tional sale of goods, the carriage of goods by sea, and porate trade facilitation principles such as electron- arbitration rules, as well as the UNCITRAL Model ic compatibility and risk analysis. Law on Electronic Commerce. Regional Integration Initiatives. A number of The Private Sector regional efforts have been undertaken to facilitate trade. Two major trade facilitation initiatives under Most of the nongovernmental entities that actively the North American Free Trade Agreement support trade facilitation programs are, not surpris- (NAFTA) are the Canada-U.S. Shared Border ingly, international transport organizations. They Accord and the Heads of Customs Conference include the International Express Carriers Confer- (HCC). The Shared Border Accord, signed by Cana- ence, the International Air Transportation Associa- dian and U.S. customs and immigration agencies, tion, the International Chamber of Shipping, the creates a set of common objectives for a joint International Road Transport Union, the Interna- approach toward trade facilitation and trade com- tional Federation of Freight Forwarders Associa- pliance. The HCC holds regular trilateral meetings tions, and the International Federation of Customs of Canadian, Mexican, and U.S. authorities to Brokers Associations. The International Chamber of review common customs issues, including enforce- Commerce (ICC), based in Paris, is an important ment cooperation and ways to improve the process- nongovernmental actor in trade facilitation and has ing of the cross-border movement of goods. For pursued customs simplification and harmonization example, the HCC endorsed the North American since the early 1920s. It promotes harmonized busi- Trade Automation Prototype, which uses EDIFACT ness practices through a variety of instruments, syntax and is designed to facilitate trade by stan- including the Commission on International Com- dardizing data elements and electronic customs mercial Practices, the Standing Committee on procedures. Extortion and Bribery, and the ICC Incoterms, With the launching of the Osaka Action Agenda which are standard trade terms and definitions for in 1996, members of the Asia-Pacific Economic use in international contracts. The ICC, in conjunc- Cooperation (APEC) committed to standardizing tion with the WCO, also administers the ATA Car- customs requirements throughout the region. In net System for the temporary entry of goods. The 1997 APEC trade ministers agreed to align national ICC has issued a set of 60 international customs norms with international standards and to recog- guidelines relating to a wide variety of trade facilita- nize each other's national standards. Recognizing tion matters. These include the reduction of paper- that simplification and harmonization of customs work and the increased use of electronic commerce; procedures can make a major contribution to trade the introduction of risk assessment techniques (pre- facilitation, ministers noted the importance of find- clearance and postclearance audits); and profes- ing technological solutions to expedite clearance of sional training for customs employees. frequent travelers. An example is the project for an The ICC was instrumental in having the issue of APEC business travel card. trade facilitation introduced at the 1996 Singapore The European Union (EU) has concluded cus- ministerial conference of the WTO. There, ministers toms cooperation and mutual assistance agree- directed the WTO Council for Trade and Goods to ments with several countries. These agreements draw on the work of other relevant international cover the simplification and computerization of organizations in the area of simplification of trade 142 Trade Facilitation: Improving the Invisible Infrastructure procedures to assess the scope for WTO rules in this modernization. It has also called for harmonization area. The ICC has encouraged the WTO to concen- of nonpreferential rules of origin, greater reliance trate on customs modernization as an essential on preshipment inspection (discussed in Box 16.1), complement to WTO rules on customs valuation and political support for making the revised Kyoto and to establish a WTO working group on customs Convention a binding multilateral agreement. B O X 1 6 . 1 I N C R E A S I N G T H E E F F E C T I V E N E S S O F P R E S H I P M E N T I N S P E C T I O N S E R V I C E S Preshipment inspection (PSI) refers to the verifica- want the services because they reduce available tion of unit prices and to the examination and rents. Unless governments consistently use a rec- reporting of the quantity and quality of exports onciliation system and act on the information before they are shipped to the importing country. generated, PSI will not contribute much to rev- PSI can help control over- or underinvoicing of enue collection. To ensure sustainable revenue imports, misclassification of imports, undercollec- collection, customs modernization and institu- tion of taxes on imports, and misappropriation of tional reform are also needed, and the strategy donor funds and can assist with monitoring of should specify how PSI services will be phased origin, compliance with national regulations and out over time (Low 1995). tariff exemption schemes, trade facilitation, and consumer protection. PSI services are provided Trade facilitation. Opponents of PSI often argue by private companies in the exporting country. that PSI hinders trade by creating additional cost- Thus, PSI can be thought of as a temporary quasi- ly steps for traders that may duplicate control privatization or contracting out of selected cus- functions imposed by customs administrations. toms functions to meet specific objectives. It The Indonesian Chamber of Commerce, howev- should not be viewed as a substitute for an effec- er, has argued that PSI facilitated trade by speed- tive program of customs modernization and insti- ing up the customs clearance process. tutional reform, which is the proper route to Overall, experience suggests that if PSI is to long-term gains in efficiency and growth. make a positive contribution, several conditions Because institution building takes time, PSI can are essential: play a useful interim role in three main areas: · Transparent procurement rules for the pre- Disbursement verification. One motivation for PSI shipment inspection contract is to monitor the use of donor funds. Where gov- · Preshipment inspection values and classifica- ernments have poor statistical capacity, PSI can tions that are an integral part of import docu- provide useful evidence. ments · Good use of provided services (with reconcilia- Revenue collection. Probably the most important tion by the ministry of finance, at a minimum) reason that governments use PSI is to deal with · Arbitration provisions to settle disputes swiftly inefficient or corrupt customs administrations. without holding up goods (Revenue collection shortfalls of up to 50 percent · Enhanced competition for service provision are reported to have occurred in some countries.) and fee setting. The effectiveness of PSI in this regard depends on how well it is implemented. Although reported Competition among service providers (split revenue savings generally exceed PSI fees, which contracts) can reduce fees but could increase are about 1 percent of the value of inspected transactions costs for users. Serial competition goods, case studies suggest that the information (bidding for a time-bound monopoly franchise, provided by PSI companies has often been disre- either at the country level or within a certain garded; customs administrations often do not area) can avoid conflicts of interest by eliminating (continued) 143 S E L E C T E D T R A D E P O L I C I E S A F F E C T I N G M E R C H A N D I S E T R A D E B O X 1 6 . 1 ( C O N T I N U E D ) the ability of importers to threaten to switch to may be a basic conflict of interest in appointing a another service provider if they are not assessed PSI company as the primary external agent sufficiently low duty payments. For serial compe- responsible for customs modernization. If the tition, the challenge is to design and supervise a company is successful in modernizing customs bidding process that awards contracts often administration, its success closes off more prof- enough to render the market contestable. With itable flows from PSI work. Governments should either contract option, termination clauses in therefore consider separating PSI support from case of poor performance are desirable. It is also customs modernization assistance. This does not desirable to explore the commercial feasibility of preclude use of the customs modernization skills allowing fuller price competition, where pricing of PSI companies, but it does mean that a distinct reflects actual services rendered, rather than the agent, perhaps another PSI company not current ad valorem practice. For instance, bulk involved in the country's PSI work, should be goods such as wheat or petroleum could be responsible for modernization. Such unbundling exempted or should face lower fees. can also facilitate different fee structures for dis- Competition to strengthen customs adminis- tinct lines of business, making it easier for prices tration should be an element of PSI contracts. A to reflect services rendered. number of countries have been using PSI for more than 10 years without seeing concomitant Source: Prepared by the volume editors, based on Dutz customs modernization. This suggests that there (2001). Policy and Practical Implications time expense and its benefits are long-term--are also a consideration. Possibly the single most signif- It has been estimated that costs stemming from icant challenge is that trade facilitation itself is customs and related import formalities are on the predicated on trade reform in general and customs order of 2 to 5 percent of the value of merchandise reform in particular. For example, computerization trade. Trade facilitation, defined as pertaining to is a critical element in the trade facilitation process, the standardization and modernization of customs but computerizing outdated procedures and techniques, is therefore a potentially very cost- requirements is not very useful. In 1995 UNCTAD effective way of reducing the costs of trading. The performed an in-depth study of its widely used main concerns of traders include excessive docu- ASYCUDA automated system and concluded that it mentation requirements; lack of automation and "cannot be successfully implemented without first limited use of information technology; lack of undertaking a major reform of customs procedures. transparency; unclear and unspecified import and This may include elimination of unnecessary pro- export requirements; inadequate procedures, espe- cessing steps, simplification and elimination of cer- cially a lack of audit-based controls and risk assess- tain forms, streamlining of the document ment techniques; and lack of cooperation among processing flow within the office, and adoption of customs and other government agencies. Other international codes." concerns are the need to implement improved Recognizing that obstacles and challenges will be management techniques and to reorganize man- encountered on the path toward comprehensive agement structures to manage the changes required trade facilitation is not to say that the difficulties are by trade facilitation. insurmountable. In the first place, given that trade For facilitation to be effective, it has to be execut- facilitation is primarily concerned with the stan- ed properly and to be closely coordinated with other dardization and modernization of customs proce- reform priorities, such as tax administration. The dures, the process of reform can draw on costs of trade facilitation--even though it is a one- internationally based standards, such as the new 144 Trade Facilitation: Improving the Invisible Infrastructure Kyoto Convention, that are largely based on multi- · A great number of trade facilitation initiatives are lateral obligations and agreements. In contrast to taking place around the world, and this can some- other policy reform matters--in taxation, for exam- times lead to confusion. The Kyoto Convention, ple--there is a common base and a common lan- including its draft revisions, provides the most guage for all facilitation initiatives. In addition, the comprehensive and convenient expression of WCO's Customs Reform and Modernization Pro- trade facilitation objectives, mechanisms, and gram provides many of the tools necessary for per- best practices. forming an audit and needs analysis, which are · Trade facilitation should be perceived not solely essential for effective customs reform. as a risk to customs revenues but primarily as a The goal of trade facilitation is clear: streamlined way of reducing the cost of operating customs and modernized customs procedures. The process regimes while at the same time attracting of achieving this objective can, however, be rather importers and investment. (See Box 16.2 for an complicated. The following summary provides a example from Jamaica.) personal perspective of policy measures and best · The current world trading system is characterized practices: by much potential and much uncertainty. Around the world, regional groupings are actively review- · Protection of customs revenue is paramount. ing and considering dozens of free trade agree- Facilitation measures such as the expedited ments and preferential trading practices. At the release of goods imported by parties with "clean" same time, a new WTO round is beginning. As customs records have to be counterbalanced by interesting and exciting as these projects may be, compliance mechanisms (that is, penalties). Pro- their completion is far from guaranteed. Trade tection of customs revenue implies that the issue facilitation and customs reform may represent a of customs integrity has to be addressed directly. better return on investment. B O X 1 6 . 2 I M P R O V I N G E X P O R T C L E A R A N C E I N J A M A I C A The experience of Jamaica illustrates that customs In 1993 the government decided to take a top- reform may have no implications for revenue col- down approach to the issue of export clearance. lection, while facilitating exports. In the early Rather than try to analyze individual steps and 1990s the Jamaican customs administration was procedures and seek agreement with customs and operationally inefficient; administrative practices other authorities to eliminate or simplify them, were poor; and corruption was widespread. Ille- targets, determined by the top management of gal narcotics trade had led to a regime of strict the Ministry of Finance and Customs, were estab- cargo control. Although the restrictions were well lished for export clearance. The emphasis was on intentioned, exporters complained that the collecting statistics on national trade movements regime simply added costs and augmented and on ensuring that no goods were exported delays. that were on the list of controlled items. Outside Clearance of export consignments required 23 technical assistance was then sought to draft an steps. It took two or three days for a typical ship- action plan to achieve these two objectives with ment to complete all clearance procedures. Par- minimum costs and delays. The advantage of this ticularly damaging was the uncertainty created approach was that it encouraged policymakers to by the procedures; a holdup of a day could result identify policy objectives and then ask whether in an exporter's missing a scheduled vessel. As and how customs had to be involved in their real- such vessels tend to visit Jamaica relatively infre- ization. The alternative bottom-up approach quently, the resulting delay could span several assumes implicitly that procedures are valuable weeks, leading to the possible cancellation of the and must be continued. letter of credit and, in the case of time-sensitive By 1995, the reforms that were instituted as a goods such as apparel, to loss of contracts. result of this rethinking had reduced clearance (continued) 145 S E L E C T E D T R A D E P O L I C I E S A F F E C T I N G M E R C H A N D I S E T R A D E B O X 1 6 . 2 ( C O N T I N U E D ) time for export consignments to 10­20 minutes. inspects only 10 to 15 percent of shipments, This dramatic improvement was achieved using clearly specified risk criteria. In addition, through three changes: lack of statistical information is no longer a jus- tification for holding up a consignment. · The implementation of a single-point clear- · The introduction of a binding, comprehensive ance mechanism, allowing exporters to go manual of procedures setting out all customs straight to the dockside with their documents rights and responsibilities in export clearance. instead of having to go to customs offices sep- This manual is published, so that exporters arately and then having to match documenta- and their agents know what the rules of the tion to cargo at a later stage. game are. · The introduction of selective inspection based on risk assessment instead of discretionary Source: Prepared by the volume editors, based on World physical inspection of consignments. Customs Bank, "Ethiopia Export Development Strategy" (1997). · As mentioned earlier, it makes no sense to simply administrations accept passively civil service transform outdated and unnecessary customs rules, including controls on organization struc- procedures from a paper format to an electronic ture, job classification, and salary levels, instead of format. That is, trade facilitation requires cus- striving for control of their own organizations. toms reform, and reform implies modernizing · Lack of understanding of the need for coordination and streamlining customs programs to interna- and cooperation between tax and customs adminis- tional standards, which are primarily found in the trations. Kyoto Convention. · High levels of corruption that plague many admin- istrations, causing loss of tax revenue and eco- nomic costs. Common Problems and Rules of Thumb In June 1994, in a speech before the Customs Coop- Tanzi went on to outline solutions and ways for- eration Council (later renamed the WCO), Vito ward. Tanzi, director of the IMF's Fiscal Affairs Depart- Make customs administrations technology based. ment, identified the following major problems in The eventual goal is paperless processing systems trade facilitation and customs reform: that include electronic reporting of import and export transactions through electronic data inter- · Out-of-date customs procedures that have not kept change; selective checking based on risk assessment pace with developments in transport and technology. techniques supported by extensive computerized · Inadequate legislation that makes it difficult to databases; periodic declaration and payment; and introduce the changes required to support new increased application of postrelease controls. All of ways of doing business, and administrations that these procedures are already operational in many often use excuses related to legislation to delay or countries. Countries with less developed infrastruc- fail to adopt new systems or procedures instead of tures will encounter difficulties. Yet almost every working to change the legislation. country has industries such as airlines and banking · A belief that computerization is the answer to all in which sophisticated technology is already opera- problems, with little thought to understanding the tional. It is, therefore, not unrealistic to expect such role of computers, the need to simplify proce- technology to be applied in the customs environ- dures, and the use of information produced by ment. computerization to control operations effectively. Rely more on postrelease audits. Experience shows · Inadequate attention to the organization and that reliance on postrelease audits yields better staffing needs of a modern administration; many results than traditional controls. An IMF technical 146 Trade Facilitation: Improving the Invisible Infrastructure assistance mission found that an audit office with a For the administration to function well, all its com- staff of 22 had issued assessments totaling US$70 ponents must be in order: its operational proce- million over a five-year period, whereas there were dures; its organizational structure and management virtually no results from more than 350,000 physi- systems, including information systems, superviso- cal inspections of containers, employing hundreds ry systems, and internal control; its human and of staff, over the same period. Little thought had financial resources; and its legislative basis. been given in this case to reallocating staff from the Comprehensively redefine the operational role and unproductive physical inspection activities to the procedures of customs. It is time for many cus- postrelease audit. Of course, countries with a well- toms administrations to rethink the way they are developed administration and a more sophisticated doing business. New control strategies need to be trading community will be able to move more introduced that result in minimal interference with quickly to a postrelease audit system of control than trade yet ensure proper enforcement of fiscal and countries in which the administration has inade- trade laws. Experience has shown that importers quate human resources or where bookkeeping stan- are more willing to pay what is due if procedures dards in the trading community are low. are efficient and customs has a service-oriented Forge a closer working relationship with the tax attitude. department and exchange information and data on Adopt innovative and flexible management sys- the foreign trade activities of importers and tems. This involves decentralization of responsibili- exporters. These measures can help improve rev- ties and decisionmaking and more autonomy and enue assessment. The tax department needs to accountability for administrators in the field. Head- know the amounts of value-added tax (VAT) paid quarters should concentrate on central manage- on imports, and it needs to know that export goods ment functions, including administrative policy, have actually left the country. In some cases there strategic planning, review of the operational sys- are benefits to joint audits by the customs adminis- tems, analysis of performance, and internal audit. tration and the tax department, particularly for Strive for autonomy in the management of value-added taxation. resources. Decisions related to human, physical, and Promote service orientation and good relations with financial resources should be the responsibility of the trading community to improve compliance. the administration. Autonomy must, of course, be These goals can be achieved through clear, transpar- combined with greater accountability through per- ent procedures; regular meetings between customs formance evaluation. officials, importers, brokers, freight forwarders, and Privatize functions that can be effectively performed port and airport authorities; joint training sessions at lower cost by the private sector. Laboratory ser- and seminars; establishment of services offices; and vices, receipt of duty and tax payments, and devel- dissemination of information. Often, customs opment and operation of computer systems are all administrations in developing countries do not activities that could and perhaps should be carried provide sufficient information to the trading com- out by the private sector. Warehouses should not munity. belong to or be operated by customs, although this Inculcate professionalism and a high level of still happens in some countries. integrity. Both can be more easily developed Invest in human resources. Traditional approaches through increased autonomy of the customs to recruiting and training will have to change. administration, meaning the ability to control Methods that rely mainly on recruitment at lower budgets and implement changes, as well as through levels and on learning on the job need to be altered accountability for performance and the require- if the administration is to keep pace with develop- ment to seek out and remove corrupt officials. ments. If customs administrations are going to rely in the future on technology- and audit-based sys- tems, different skills will be required. Getting There Establish firm management control, in particular as An efficient customs administration must be flexi- it relates to integrity. Integrity in an organization ble and able to respond quickly to the needs of gov- requires a clear, well-articulated code of conduct, ernment. It is not enough to introduce sophisticated willingness to take disciplinary action, and effective technology; this alone will not guarantee success. internal control systems. 147 S E L E C T E D T R A D E P O L I C I E S A F F E C T I N G M E R C H A N D I S E T R A D E Elements of the Trade Facilitation Toolbox sense customs is vital to the flow of international trade in goods, which totals US$6 trillion each year. Trade facilitation is an objective; comprehensive Pursuing the objectives of trade facilitation on the reform of customs and related import requirements national level will make the domestic public and are the means of achieving this objective. Customs private sectors more efficient and will also play an reform does not take place in isolation but in a larg- important role in securing and attracting foreign er context that includes other considerations, such investment. Clearly, for many countries, achieve- as transport policies, and other actors, such as ment of trade facilitation objectives will be a long- importers and carriers. Recently, in an explicit term process requiring substantial technical and recognition of all these factors, the director general financial assistance. In order to determine what role of the International Express Carriers Conference the WTO could usefully play to attain national facil- published an audit methodology (Raven 2000) con- itation objectives, countries must start by defining taining detailed questionnaires intended to supple- these objectives and determining where reforms are ment personal interviews during a trade facilitation needed. audit. Such an audit should be the first step taken by countries concerned with reducing trade costs.1 Note Conclusion 1 Individual questionnaires focus on forwarders/agents/customs brokers/multimodal transport operators; exporters; importers; Every year, the private sector spends considerable shipping lines; road carriers; airlines; express operators; ports; sums of money to design and develop seamless sup- airports; border-crossing points; customs; commercial banks; ply chains for intercompany transactions and for exchange control/central banks; preshipment inspection agen- cies; chambers of commerce; and departments of trade/exter- transactions with suppliers and customers. National nal trade. Further information and tools, including the audit, import and export requirements are a major obsta- can be obtained from the Trade and Transport Facilitation cle to achieving the seamless supply chain. In this Website, . 148 17 M A R I PA N G E S T U Industrial Policy not limited to the manufactur- ing sector; it also encompasses the processing of agricultural and Developing and mining products, as well as services industries, both of Countries which sectors add value to man- ufactures. In practice, industrial policy often has multiple objectives, including short-term employ- ment, increased output, more even income distribution, more equal regional distribution of I economic activity, and enhanced ndustrial development is an inte- technological capacity. There are often also noneco- gral part of any economy's growth nomic objectives, including national pride and strategy. Most countries pursue some kind of indus- prestige and the perceived need to promote "strate- trial policy, although their objectives and approach- gic" domestic industries. These objectives are fur- es may differ radically and may change over time. ther confused to the extent that many developing Given the changing global and domestic environ- countries are concerned about foreign ownership ment, developing countries need to reassess the and how it can affect domestic capabilities.2 It is options open to them for conducting an effective, important to pursue an industrial policy that has WTO-consistent industrial policy. limited and clearly defined objectives, as there may not be sufficient policy instruments to meet multi- ple objectives. Moreover, different objectives may be Objectives and Scope of Industrial Policy inconsistent with each other. The economic literature and the lessons from implementing industrial policy emphasize that an Justifying Industrial Policy effective industrial policy or strategy needs well- defined objectives, justification, and scope. The The economic case for government intervention World Bank has provided a working definition of designed to achieve long-run productivity improve- industrial policy as "government efforts to alter ments rests on the need to correct alleged market industrial structure to promote productivity based failures stemming from externalities, missing mar- growth" (World Bank 1992).1 This definition is use- kets, or other failures, while taking into account ful because it focuses on the objective of economy- potential side effects on other sectors in the econo- wide factor productivity growth rather than on my. The traditional economic argument for provid- merely changing the structure of industrial outputs ing government assistance to certain industries is to or dealing with certain sectors. Industrial policy is protect infant industries.3 Import protection in the 149 S E L E C T E D T R A D E P O L I C I E S A F F E C T I N G M E R C H A N D I S E T R A D E form of a tariff or a subsidy based on the output of best policy requires the unrealistic requirement of firms (the two instruments have an equivalent effect perfect knowledge of all aspects of the economy so on output of a particular industry) is justified on that the net effect of the intervention can be known. the basis of some dynamic externality such as learn- Third-best interventions made in ignorance of the ing-by-doing or on-the-job training that reduces true values of some behavioral parameters may lead costs. Under this rationale, only learning processes to further distortions and reduce welfare. external to the firm should be assisted, since the Other justifications for industrial policy rest on firm cannot obtain rents or profits from such train- the rationale of technology development. The ing and will thus not invest in it. appropriate policy under that rationale would call There are important qualifications to the infant for technology-based intervention, not an output- industry argument. First, the reductions in cost over based intervention such as subsidies, or for assis- time should compensate for the higher costs during tance for technology development and policies to the period of assistance. Second, the provision is not encourage foreign direct investment (FDI), which is for blanket assistance to all firms in an industry; the an important vehicle for the transfer of technology. existence of an externality and provision of the (See Chapter 19, by Bora, and Chapter 34, by Saggi, assistance should be linked to performance by the in this volume.) recipient (for example, to increased efficiency or cost reduction) and the assistance should be phased Instruments of Industrial Policy out over time. Third, the appropriate instrument for realizing the positive externality from the expansion In practice, countries have used a wide range of of domestic industry may not be a tariff or a sub- instruments in the name of industrial policy. These sidy, both of which are output based. A more appro- can be categorized as external, product, and factor priate policy is a subsidy related to the process, job, market interventions. or product that creates knowledge or learning. External market interventions involve protecting The appropriateness of policy instruments follows domestic industries from imports, using instruments a more general theme in the literature on govern- such as import tariffs,quotas,licensing,and local con- ment intervention (see Bhagwati 1971; Corden tent programs, as well as export promotion measures 1974). Each externality or market failure calls for a to assist industries to catch up and break into new tax subsidy whose base is the variable that generates markets. Common export promotion instruments the externality or market failure, and the tax subsidy are export subsidies, export promotion zones, and rate will be the rate that has the optimal effect. Any subsidized credit (sometimes tied to export targets). tax subsidy other than the optimal tax subsidy causes Product market interventions to promote competi- what Corden (1974) called by-product effects, which tion in domestic markets include competition poli- impose undesired costs elsewhere in the economy. cy (to ensure fair competition between domestic The economywide effects of intervention in one players as well as for foreign players) and domestic industry also need to be borne in mind. One way to market entry regulations. do this is to focus on the effective rate of protection, Factor market interventions include policies such as which takes account of the impact of a tariff on performance requirements and restrictions on FDI both inputs and outputs. For instance, a tariff on an designed to influence the operations of foreign affili- input will cause the effective protection of the ates so that the host country realizes a net benefit downstream user to decline (see Box 17.1). from FDI (UNCTAD 1999a). Factor market interven- Another economic argument for government tions in the capital market and the financial sector are intervention--what is known as the second-best aimed at correcting financial market imperfections, argument for tariffs or subsidies for some goods-- promoting infant industries, and protecting or phas- has to do with the presence of "unremovable" dis- ing out declining industries. These measures include tortions in the form of tariffs or other import setting up development finance institutions, provid- protection.4 In practice, intervention to correct ing direct capital subsidies to selected industrial enter- such distortions poses several problems. First, it is prises, furnishing capital subsidies and capital not clear why the preferable (first-best) policy of assistance to declining or mature industries and pro- removing the distortion cannot be implemented. viding priority access to credit (often at subsidized Second, the correct determination of the second- rates) by requiring financial institutions to lend to 150 Industrial Policy and Developing Countries B O X 1 7 . 1 N O M I N A L A N D E F F E C T I V E R AT E S O F P R O T E C T I O N The nominal rate of protection (NRP) can be trade prices is the same, except that in this case defined as tariffs do not exist (the value of t is 1). For example, suppose 1 ton of steel is worth NRP = (P ­ P*)/P* US$1,000 on the world market. To produce it, a where P is the domestic tariff-inclusive price of a factory has to buy 1 ton of iron ore at a world good and P* is the free trade price. As the latter price of US$600. Assume, for simplicity, that noth- cannot be observed in practice, most empirical ing more is needed for steel production. Under studies take the world price as a measure of P*. these circumstances, the value added per ton of The effective rate of protection (ERP) can be steel in the factory will be US$400. If a 20 percent defined as the proportional increase in value nominal tariff rate is imposed on steel imports and added per unit of a good produced in a country there is no tariff on iron ore, the effective rate of in relation to value added under free trade (no protection in those circumstances will be protection). The magnitude of the ERP depends (1,200 ­ 600)/400 = 1.5, or 50 percent. not only on the nominal tariff on the final prod- uct concerned but also on the tariffs applied to The ERP in this example is more than double the the inputs used and the importance of those 20 percent NRP on steel. If no tariff is imposed on inputs in the value of the final product. A simple steel but a nominal tariff of 33 percent is imposed formula for calculating the ERP is on imports of iron ore, the ERP would be ERP = (V ­ V*)/V* [1,000 ­ (600 + 200)]/400 = 0.5, or ­50 percent. where V is the domestic value added per unit of This example illustrates that an NRP of zero the final good (including the tariffs on that good does not necessarily imply that trade is undistort- and on its inputs) and V* is value added under ed. As another example, assume that cocoa free trade. Value added per unit, in turn, is beans account for 95 percent of the production defined as the gross value of output minus the cost of cocoa butter. The imposition of a 5 per- cost of inputs used in production: V = tfPf ­ tiPiX, cent nominal tariff rate on cocoa butter would where tf and ti equal 1 plus the tariffs on the final then imply an effective rate of protection for the good and on inputs, respectively; Pf and Pi are the cocoa butter industry of 100 percent. prices; and X is the amount of input used to pro- duce a unit of the final good. Value added at free Source: Hoekman and Kostecki (2001). particular sectors or types of companies. Intervention in the external environment. For the sake of con- in the labor market may have efficiency and equity creteness, this section focuses on examples from objectives. The former have to do with human East Asia, but much of what is said applies to all resource development through education and train- developing countries. ing; the latter include minimum wage requirements Industrial policy in East Asia has evolved over the and social safety net schemes. past three decades (Table 17.1) as import substitu- Box 17.2 gives examples of the types of industrial tion has given way to export orientation and, subse- policy instruments used in the Republic of Korea and quently, to development of a knowledge-based Japan in the early phase of their industrialization. infrastructure. Shifts in policy approaches and instruments have been influenced by internal fac- tors such as the size of the market, the need to adjust Evolution of Industrial Policy to adverse shocks, the ineffectiveness of import- The approach to industrial development and the substitution industrialization strategies, and the range of instruments used have evolved over time as need to attract FDI for technology and to gain mar- a result of changes in development paradigms and ket access. Policy has also been influenced by exter- 151 S E L E C T E D T R A D E P O L I C I E S A F F E C T I N G M E R C H A N D I S E T R A D E B O X 1 7 . 2 . I N S T R U M E N T S O F I N D U S T R I A L A N D E X P O R T P R O M O T I O N P O L I C I E S E M P L O Y E D I N T H E R E P U B L I C O F K O R E A A N D J A PA N Export Promotion and Import Restrictions Product and Factor Market Interventions · Import restrictions, both general and specific · Lax enforcement of competition policy, · Preferences for particular sectors and, in some including the extensive use of cartels cases, particular firms in export promotion · Government creation and promotion of con- · Export targets for specific firms as conditions glomerates (Korea) for the provision of subsidies (Korea) · Tax concessions to corporations to increase · Interest rate subsidies and credit and foreign investment exchange availability for favored firms that · Promotion of a close, long-term relationship meet export targets between finance and industry (which was crit- · General export promotion through JETRO ical to the implementation of the industrial (Japan) and KOTRA (Korea) policy) · Provision of infrastructure, including human · Labor repression to ensure labor peace in a capital, in support of exports period of structural change (Korea) · Tax relief on imported inputs and for research · Establishment of state industries to enhance and development expenditures industrial development (Korea) · Permission to favored conglomerates to import · Extensive administrative guidance. capital goods and foreign technology and to raise cheaper finance on international markets. Source: Adapted from Singh (1996). nal factors such as increased competition, techno- restructuring. The main challenge for many East logical change, pressures from major trading part- Asian developing countries is to maintain tradition- ners to become signatories to GATT codes, al comparative advantage (based on unskilled and multilateral rules negotiated under the WTO, and low-cost labor and on resources) while building up the financial crisis that erupted in 1997. new sources of comparative advantage and "gradu- Rapid changes in transport, communications, ating" to higher value added production embodying production, marketing, and distribution technolo- more technology and human capital. gies, as well as management processes, have acceler- To address the structural issues faced by East ated globalization and reduced traditional Asian countries, a number of policy responses are comparative advantage in mass production. The appropriate, many of which have been adopted by production of components and parts, and final pro- governments. One is to maintain competition in the duction itself, are increasingly outsourced or spread domestic market. Imports can influence productivi- among different locations on the basis of cost and ty through embodied technology and can be an market considerations. The decentralization of pro- effective way to assimilate new techniques and duction and production processes that occurred in knowledge (see Chapter 34, by Saggi, in this vol- the 1990s in East Asia provides a strong example of ume). Other factors conducive to exports are the how the region has been integrated through trade promotion of FDI and participation in internation- and investment linkages. Technological develop- al production networks; appropriate macroeco- ments will continue to be an important source of nomic policy; efficient infrastructure and pressure for continuous industrial restructuring. supporting services; and policies to enhance human The decline in export growth before the crisis, in capital and technological capability, such as research part stemming from structural factors such as and development, education, and creation of indus- declining competitiveness and low productivity trial clusters. East Asian countries have, in general, growth, also points to the need for industrial shifted their focus on such complementary policies 152 Industrial Policy and Developing Countries Table 17.1 Evolution of Industrial Policies in East Asia, 1950s­1990s Economy 1950s 1960s 1970s 1980s 1990s 1965­76 1977­78 1980s 1990s China Defense/industry Plant Coastline Infrastructure (heavy industriali- importation liberalization High zation) (light industries) technology 1950­ 1979­ 1990s Hong Kong EO (laissez-faire, education, infrastructure, institutional Improved Upgraded (China) support) institutional support for support for technology industry 1967­73 1974­85 1986­ Indonesia Stabilization Strong IS Liberalization Beginning IS EO 1950­58 1959­ 1967­ Mid-1980s Japan IS EO Liberalization Deregulation International- ization 1961­72 1973­79 1980­ 1990s Korea, Rep. EO EO Liberalization Deregulation of IS (heavy (trade, invest- since mid-1980s Industry) ment, finance) (innovation oriented) 1950­70 1971­85 1986­ Malaysia Moderate IS Continued IS Liberalization Added EO EO 1950­ 1980s 1990s Philippines IS Continued IS Liberalization Continued (political liberalization instability) (strengthened political stability) 1950s 1960s­1980s 1990s Singapore IS (while EO Strategic still part independence of Malaya) (high tech- nology and services) Regionaliza- tion 1953­57 1958­80 1986­ Taiwan IS EO Liberalization (China) 1961­71 1971­86 1986­ Thailand IS IS (capital goods, EO beginning in 1981) Technology- intensive industries Some EO Note: IS, import substitution; EO, export orientation. Source: Masuyama, Vanderbrink, and Chia (1997): table 1.1. 153 S E L E C T E D T R A D E P O L I C I E S A F F E C T I N G M E R C H A N D I S E T R A D E and now endeavor to provide an environment con- countries with incomes per capita above $US1,000 ducive to the development of competitive industries and lays out rules for the use of countervailing and to enhance the economy's flexibility in measures to offset injury to domestic industries responding to changes. Table 17.2 provides a sum- caused by foreign production subsidies.8 The SCM mary of the policies pursued by East Asian agreement covers financial contributions made by economies just before the 1997 financial crisis. The or at the direction of a government that provide a response to that crisis reinforced the need for a benefit to a specific enterprise, industry, or region.9 more comprehensive approach to policy reform, to Subsidies that are conditional on exports are pro- some extent mandated by IMF reform packages. hibited, as are subsidies that encourage the use of Countries not supported by IMF programs have domestic rather than imported inputs. Taking also been induced to pursue similar reforms. 5 action against subsidies requires a determination that subsidies exist and have a negative effect on the trade of another member. This is done by showing Multilateral Rules Regarding Use of that there is harm to another member in the form of Industrial Policy Instruments injury, serious prejudice, or impairment and nullifi- An important question concerns the extent to cation of benefits.10 The SCM agreement and its which WTO provisions constrain the policy mea- provisions relating to developing countries are dis- sures that members can use to protect domestic cussed in greater detail in Chapter 18, by English suppliers and promote exports and transfer of tech- and De Wulf, in this volume. nology. This section contains brief summaries of The SCM agreement has important implications the major WTO rules regarding industrial policy for industrial policy. Take, for example, the case of instruments. Many of these are discussed in greater Korea, which has been notorious for its use of tar- depth in other chapters of this book. geted subsidies. Before 1995, Korea offered 26 dif- ferent types of subsidies, totaling about 2.5 trillion won per year. In 1995 it reduced the number to one Tariffs, Antidumping, and Safeguards subsidy to small and medium-size enterprises, only Most developing countries have undertaken tariff 15.2 billion won in amount (WTO 1996b). In con- reduction programs in the past two decades. They trast to the voluntary Subsidies Code negotiated have also undertaken to bind many of their tariffs, during the Tokyo Round, all countries are bound by although frequently at relatively high tariff rates the WTO agreement, and the SCM agreement that provide considerable scope for raising applied extends to subnational governments. It should be tariffs (see Chapter 54, by Francois and Martin, in noted, however, that the disciplines on subsidies this volume). Although the average level of tariff constrain primarily export subsidies; constraints on protection has declined, there continue to be peak production subsidies are weak. For developing tariffs in "sensitive" industries in both industrial countries, the SCM agreement is a two-edged and developing countries, and the dispersion of sword: it contains a number of loopholes that allow protection remains substantial in many countries.6 them to continue to use subsidies to promote indus- Import protection can also be imposed through trial policy objectives, but these also apply to indus- antidumping or safeguard measures, which are trial countries. Thus, developing countries have no often used by industrial countries to protect declin- prospect of using subsidies to gain a competitive ing industries. The WTO Antidumping Agreement advantage vis-à-vis industrial countries. imposes disciplines on the use of antidumping by countries and contains a number of provisions The Agreement on Trade-Related Investment Mea- aimed at reducing the extent to which antidumping sures can be used against developing countries that are trying to develop their exports.7 Under the Trade Related Investment Measures (TRIMs) agreement, a number of investment per- formance­related measures that have an effect on Export Promotion and Export Subsidies trade were to be notified and eliminated by January The Agreement on Subsidies and Countervailing 2001 (January 2003 for least-developed countries). Measures (SCM) prohibits export subsidies by The trade-related performance requirements that 154 esY -- -- -- No -- -- esY esY -- -- esY seY esY esY seY seY (continued) Japan Kong seY o o N No -- No N No seY seY -- -- -- -- -- seY -- esY (China) Hong No seY esY o N esY seY Fall esY -- seY -- -- -- esY esY seY o o N N Indonesia esY esY seY seY o N No No seY seY -- seY seY esY o N seY No seY seY Singapore of seY o o N No No N No No esY esY seY esY esY seY seY esY esY esY esY Korea, Rep. esY seY No esY seY -- -- seY -- seY seY No esY seY seY seY seY Partly Thailand seY esY esY seY esY o N allF -- esY -- esY esY o No N esY esY esY No Philippines Asia in No seY seY esY -- -- -- seY seY -- seY seY No seY seY seY seY No Bangladesh Exports esY esY esY -- esY esY -- esY esY esY esY esY esY esY esY esY esY esY Malaysia Promoting for India seY seY seY -- Fall Fall Fall seY -- seY seY seY -- seY seY seY seY seY Measures policy and production targeting y y regulations depreciation Policies R&D affecting incentives rates subsidy marketing assistance development industr protection training subsidies/facilities for measure measures controls assistance 17.2 and of subsidy Measures Strategic/domestic Export Import Price Investment Credit concessions Holiday/exemptions Reduced Accelerated arrangements ableT ypeT 1. Industrial General Specific/industr Support Manpower Investment Deregulation axT Production Input Assistance Pricing Regional Adjustment 155 o N esY esY o o o N No N N seY esY No esY seY seY No esY esY Japan (China) seY seY seY seY o o No N N No seY No seY A esY esY o No N Hong MF Kong seY esY esY esY esY o o N N No esY No esY seY seY No seY esY ndonesiaI seY seY seY o N seY o o N N No seY No seY esY No No No seY Singapore of seY esY esY o N esY o o N N No esY No esY seY No No No No Korea, Rep. seY seY seY seY seY seY seY No seY No seY No esY two others others Thailand On A, A, MF MF seY esY esY esY esY No esY No esY esY esY o others others inimal N No A, Philippines M A,F MF M seY seY seY o N seY A esY seY No seY seY seY seY seY esY AF M MF Bangladesh seY esY esY o N esY o N No No esY esY esY esY esY o o others N N Malaysia A,F M India seY seY seY seY seY seY No No seY seY seY seY seY seY seY A MF inputs exports zones and II. exports management prices imported subsidies on formance affecting Arrangement. requirements restraints annex y (continued) affecting finance insurance quality per cash cartels promotion export licensing prohibitions taxes/levies quotas (1996): measure drawback/tax Multifibre 17.2 incentives guarantees requirements organizations of Export Export Export Export-processing Export Export Export Export measures oluntarV A, available. Singh exemptions MF Measures Duty Registration Export Export Export Minimum Export ce: ableT ypeT Not 2. Export Other -- Note: Sour 156 Industrial Policy and Developing Countries must be removed include local content and trade- industrial policy. This has happened to some extent balancing requirements, both of which are and have with bindings in tourism, but not in other sectors. been important policy instruments of industrial From an industrial policy viewpoint, liberaliza- policy. Some developing countries have yet to notify tion in the area of infrastructure is extremely and phase out these measures, and many are seeking important for developing countries. Given that the to extend the length of the transition periods. Many disciplines on subsidies and performance require- developing countries are also strongly resisting any ments are forcing developing countries to think of extension of the list of prohibited TRIMs in future more neutral ways to develop export capacity (Laird negotiations. (See Chapter 19, by Bora, for a more 1997), improvements in infrastructure--in particu- comprehensive discussion.) lar, in telecommunications, financial, and transport services--constitute an important mechanism for improving competitiveness. Agreement on Trade-Related Aspects of Intellectual Property Rights (TRIPS) Infant Industry Protection Perhaps more than any other WTO agreement, implementation of the TRIPS agreement involves GATT Article XVIII, Sections A and C, allows substantial changes in national legislation and members that are in "early stages of development" strengthening of enforcement institutions. The to use trade barriers to protect domestic industry. required strengthening of protection of intellectual Section B of Article XVIII affords developing coun- property rights (IPRs) has implications for indus- tries flexibility in imposing trade measures to pro- trial policy. In the case of domestic firms, it implies tect their balance of payments. Before the Uruguay both a need to and greater incentives to innovate Round, little use had been made of Section C, on and compete dynamically; reverse engineering and infant industries, in part because the use of this imitation have become less feasible. For foreign provision requires the payment of compensation firms it means that, where permitted, market access and in part because developing countries had few through a commercial presence may become more tariff bindings. Instead, many countries used Sec- attractive as IPR protection improves. Given that tion B, which does not require compensation and developing countries do not in general have a com- which provides leeway for selective intervention. parative advantage in innovation, attracting FDI as The WTO has tighter balance of payments disci- a means of transferring and diffusing technology is plines that constrain the scope and duration of this important for them. Thus, TRIPS implies a greater exception (Singh 1996: 166). As tariff bindings need to improve FDI policies. An important provi- expand, developing countries may have to rely sion for developing countries from an industrial increasingly on Article XVIII, along with safe- policy perspective is TRIPS Article 66.2, which guards and domestic subsidy programs, to protect requires industrial countries to support technology domestic industries. transfer to least-developed countries. So far, little is known about the extent to which this provision has Special and Differential Treatment been implemented (UNCTAD 1999a). The WTO has numerous "special and differential" (S&D) treatment provisions in favor of developing General Agreement on Trade in Services countries, and there is scope for S&D treatment in The General Agreement on Trade in Services the application of industrial policy under each of (GATS) allows sectoral commitments (bindings) to the agreements mentioned. In addition to transition be made for the four modes of supplying services: periods allowing for delay in implementation, some cross-border, consumption abroad, commercial agreements (such as those on SCM, safeguards, and presence, and movement of natural persons. antidumping) include exemptions and less strin- Through the inclusion of commercial presence as a gent disciplines for developing countries. There is mode of supply, rules on foreign investment in ser- mounting concern on the part of developing coun- vices have now become part of the multilateral trad- tries, however, that these provisions do not suffi- ing system. Members can therefore use foreign ciently promote their interests and are not being investment liberalization commitments as tools of implemented (UNCTAD 1999a). 157 S E L E C T E D T R A D E P O L I C I E S A F F E C T I N G M E R C H A N D I S E T R A D E Implications of WTO Rules for Industrial ronment that is conducive to investment and inno- Policy vation. Moreover, because development of the man- ufacturing and resource-processing sectors depends Some common features of the agreements that dis- on the existence of an efficient services industry, cipline the use of government policy to promote regulatory reform and liberalization must span the particular industries deserve to be highlighted in services sector. order to understand the impact on industrial policy Multilateral rules are developing in line with the instruments under the present agreement, as well as shift toward the use of more generic policies for in the future. promoting industrial development. Since export First, the agreements take a trade, not a balance of subsidies can no longer be used to promote exports, payments, approach to disciplining policies. Since policy should move in the direction of reducing fis- nondiscrimination is the cornerstone of the WTO cal and procedural constraints on exports (Laird system, any nonborder policy that affects trade in 1997), trade facilitation, and implementation of goods and services (that is, by resulting in discrimi- non-sector-specific (generic) policies to make the nation) is subject to discipline or requires an country more competitive. The effect of the WTO exemption. rules is not to eliminate the role of government but Second, the rules are ownership neutral. Apart to shift its emphasis toward the supply side. Policies from the GATS and TRIPS, in which a national related to infrastructure, human capital formation, treatment standard is applied to covered invest- innovation, and diffusion of technology are now ments, disciplines on policies such as subsidies and critical for export competitiveness. These policies local content protection do not distinguish between need to be complemented by stable exchange rates foreign affiliates and domestic enterprises. What is that do not penalize or favor exports and by a com- important is the "trade effect" of the instrument. petition policy that promotes rivalry among pro- This means that countries seeking to apply a partic- ducers that contest the domestic market, including ular policy to foreign-owned firms must find a pro- foreign companies. These are generic prodevelop- vision in an agreement that allows the use of the ment policies that are not confined to--and do not policy; they can then apply it to a foreign firm as favor--particular industries or producers. long as there is no "trade effect." The appropriate response to fears of anticompeti- Third, policies for promoting industries tive behavior by foreign companies is not to impose (designed to stimulate investment or export performance requirements and restrictions but to growth) are restricted to generic instead of specific put in place an effective national competition law to policy instruments. This has the effect of leveling ensure fair competition. Many countries have the playing field for international trade by not begun to introduce or are preparing to introduce allowing countries to develop specific industries competition laws. The crucial issue here, however, is through specific policy instruments. having the right institutions and mechanisms, able Finally, the approach to S&D treatment in the to implement the law objectively and to conduct the WTO has typically been limited to transitional necessary investigations. Given capacity constraints, arrangements, complemented by de minimis provi- this process will take time in many developing sions (see Chapter 49, by Oyejide, in this volume). countries. It is therefore important to focus on lib- eralization of trade in services, domestic regulatory Conclusion reform, FDI, and other factor markets and to ensure that policies in these areas are subject to WTO rules Shifts in development paradigms, technologies, and and disciplines. One advantage of the WTO rules in multilateral rules imply that an effective and WTO- this context is that they are neutral between foreign consistent industrial policy for developing countries and domestic producers, helping to ensure that in the 21st century must be comprehensive, rather domestic and foreign producers are able to compete than target specific sectors. Recognition of the on equal terms. importance of complementary policies for ensuring There is still considerable scope for using indus- competitiveness has shifted the policy focus toward trial policy instruments such as tariffs (within enhancing the efficiency of infrastructure, improv- bound rates); subsidies for regional development, ing human capital formation, and creating an envi- R&D, and the environment; and export promotion 158 Industrial Policy and Developing Countries measures such as credit and insurance schemes at second-best set of tax subsidies. Discussion of other second- subsidized rates, concessional tax and duty provi- best instruments such as local content can be found in Rodrik (1987); Greenaway (1992); Chao and Yu (1993); Richardson sions, and export-processing zones. Developing (1993); Morrissey and Rai (1995); and Moran (1998). For a countries intending to use such policies (or seeking discussion of export subsidies, see Harris and Schmitt (1999). to extend transition periods to allow the use of 5 In the case of the crisis-affected countries that were under an other, WTO-inconsistent policies) need to assess the IMF program--Indonesia, Korea, and Thailand--the reforms extent to which policies favoring particular produc- undertaken have been comprehensive in terms of liberalization ers are in their national interest. At the same time, of market access in goods and services, and for FDI and com- the appropriate transition period for changing to a petition policy. Implementation is still at issue, but the steps more generic policy stance needs to be based on a taken have been dramatic. realistic assessment that reflects the country's devel- 6 It is useful to distinguish between sunset and infant industries. opment strategy and the need to build up institu- The former are industries that are declining; the latter are tions, capacity, and capability. Finally, the pursuit of industries that are expanding and, owing to market failures, industrial policies needs to be subjected to the crite- require protection from competition. ria identified at the beginning of this chapter: clear- 7 See Chapter 22, by Finger, who notes that these instruments ly defined objectives; a determination that the are increasingly being used by developing countries; see also policy instrument is the most appropriate one for Laird (1997). meeting the objective; and implementation that 8 The agreement applies only to nonagricultural products; the responds to clear criteria and is transparent, prefer- WTO Agreement on Agriculture contains separate, and more ably with clear performance and exit requirements. comprehensive, disciplines on agricultural subsidies. 9 The agreement contains a list of types of measures that would be considered to be financial contributions: grants, loans, equi- Notes ty infusions, loan guarantees, fiscal incentives, and the provi- This chapter draws on Bora, Lloyd, and Pangestu (2000). sion of goods and services. Since a government is defined to include any public body within the territory of a member, sub- 1 Martin and Mitra (2001) show that the productivity growth national governments, public bodies, and state-owned compa- rate in agriculture is higher both on average and for groups of nies are covered. The definition of a benefit has not been fully countries at different stages of development. resolved in cases where indirect financial contributions are involved. 2 For a discussion of how foreign ownership matters in the con- text of development, see UNCTAD (1999c). 10 Injury is defined as harm to a domestic industry caused by sub- sidized imports into the territory of the complaining member. 3 See Kemp (1964) for the first careful statement of the infant Serious prejudice is defined as adverse effects in the market of industry argument; also see Baldwin (1969). the subsidizing member or in a third market. Nullification of 4 See Lipsey and Lancaster (1956) for the first theoretical exposi- benefits can arise when improved market access resulting from tion; see Lloyd (1974) and Hatta (1977) on the nature of the a bound tariff reduction is undercut by the subsidy. 159 18 P H I L I P E N G L I S H L U C D E W U L F Export Trade Promotion Organizations Development The creation of the Internation- al Trade Centre (ITC) in the Policies and mid-1960s led to the establish- ment of export promotion or trade promotion organizations Institutions (TPOs) in many countries. These were to be "focal point" institutions to assist exporters in penetrating foreign markets. The TPOs have largely been state organs that provide com- M mercial intelligence, market any countries, at all levels of devel- research, services to foreign buyers, group promo- opment, have made use of policies tions, and advice on shipping, transport, and pack- designed to promote exports. Sometimes these aging. Some TPOs also administer incentive policies are intended to offset distortions created schemes, train exporters, provide export licenses, by other policies, such as an overvalued exchange and engage in investment promotion. rate. They may also be motivated by market fail- A consensus has emerged that, except in a few ures--for example, asymmetric information, cases (Australia, Finland, Ireland, New Zealand, and which means that potential exporters do not know Singapore), TPOs have not lived up to expectations. about market opportunities and cannot obtain The experience with TPOs suggests that there are access to finance. Almost all countries maintain seven characteristics or elements that determine policies to ensure that exporters are not subject to whether they are likely to be effective. double taxation; thus, duties and taxes collected on inputs embodied in exported goods are gener- 1.An overall incentive framework that is favorable to ally rebated. This chapter deals with aspects of exports. TPOs can only function well if the overall export development policies, including trade pro- incentive framework is not stacked against motion organizations, matching grants, duty exports. They can overcome some antiexport bias drawback and temporary admission schemes, of the incentive system, but there are limits, and export-processing zones, and trade finance. The the fact that many TPOs operate in an environ- intention is not to be exhaustive but to discuss ment characterized by a strong antiexport bias in cross-country experience, identify good practices part explains their failure. The antiexport bias fre- in these areas, and assess the WTO-compatibility quently stems from an overvalued exchange rate, of such policies. a tariff structure that provides high nominal and 160 Export Development Policies and Institutions effective protection, nontariff barriers resulting ship" of the problem and its solution to the from dysfunctional customs practices and poorly exporter community. Thus, trust is important. designed quality control mechanisms, the absence Although--since country circumstances differ-- of trade finance, costly infrastructure services there is no single model for such a partnership, (roads and ports), and excessive bureaucratic many recommend that TPO boards have a major- control of trade procedures. Special mechanisms, ity of recognized exporters and be headed by a such as export-processing zones, duty drawbacks, well-respected business leader of acknowledged and temporary admissions, can be devised to integrity. For example, in the Philippines the pri- shelter exporters from the worst effects of import vate sector is represented on the Export Develop- protection, and export-financing facilities can ment Board by a private accredited export help compensate for the lack of a well-function- organization, which rotates every three years and ing financial system. Even if these mechanisms are which has to coordinate the private sector posi- made to work well, however, they cannot substi- tion and defend it before the board. In Finland tute for the establishment of a truly export- the export promotion strategy is a joint venture friendly incentive system. between the government and the private sector 2. Autonomy of operations. The TPO must be able to and is designed and executed on the basis of a influence policy, mobilize the resources and ser- consensus among government, industry, and vices needed to support an export drive, and labor. These examples contrast with the reality in deliver these services when and where required. most developing countries, where very few TPO This argues for a flexible and autonomous institu- heads have significant export experience. tion that operates with top-level political support, 4. A balance between offshore and onshore objectives. and maintains close formal and informal links Export promotion means promotion of competi- with public and private sector actors. Such a TPO tiveness and goes to the heart of how business is can react quickly to new requests and changing done. Competitiveness has to do with pricing; circumstances without having to obtain time- quality standards; the ability to interface with new consuming clearances and permissions. An business modes that emphasize timeliness of autonomous TPO is more likely to enjoy the con- delivery, outsourcing, and business-to-business fidence and mutual trust that is required to relations; and availability of supportive infrastruc- engender and sustain a supportive relationship ture services (telecommunications, finance) and with the business community and overcome the of quality domestic inputs. The traditional TPO suspicion or conflict that often dominates rela- emphasis on offshore activities--information tions between the state and the private sector. In gathering, market research, trade representation, reality, most TPOs operate under the trade min- fairs, and the like--ignores much of this agenda. If istry, which is often badly positioned to deal with TPOs were to focus on being attentive to supply the main issues that affect external competitive- conditions, they would address firm-specific sup- ness and has little power to mobilize the necessary ply bottlenecks faced by potential exporters resources. As public agencies, most TPOs are through well-targeted enterprise support. A well- handicapped in influencing exporters or arguing designed matching grant program could help against public policy that hurts them. reduce production costs and enhance output qual- 3. A demand-driven strategy. The private sector ity (Box 18.1). Such a scheme may well be run by should play a dominant role in defining, imple- other entities, but the TPO can be an advocate for menting, and monitoring the TPO strategy. it. Helping enterprises obtain International Orga- Although the government must set the ground nization for Standardization (ISO) certification is rules of the export "game," it is the private sector another example. In general, instances abound in that does the exporting. (The export performance which consultant advice can lower production of public enterprises has often been dismal.) costs and wastage. Such enterprise services can be Export promotion is not an analytical issue; diag- supplied by both private and public providers, but nosis of the problems is usually relatively straight- the two should compete on equal terms, and the forward. Bringing the solution home is more emergence of private sector services providers complex and requires the transfer of the "owner- should be encouraged. Enterprise support should 161 S E L E C T E D T R A D E P O L I C I E S A F F E C T I N G M E R C H A N D I S E T R A D E B O X 1 8 . 1 M AT C H I N G G R A N T S C H E M E S T O P R O M O T E E X P O R T S Disillusionment with the performance of trade examine these questions was conducted for a promotion organizations has led to experimenta- scheme in Mauritius, which happened to be one tion with other techniques of export develop- of the most successful as measured by its apparent ment. Among these are matching grants: effect on exports (Biggs 1999). The study con- projects proposed by individual firms receive cluded that nearly half of the firms assisted would grants that have to be matched with the firm's have carried out their projects anyway and that capital. The justification for these schemes is gen- the existence of externalities could not be proved. erally that there exist exporting firms that would For example, little new demand was generated for like to increase their exports and nonexporters local suppliers who might then have been in a that would begin to export were it not for lack of better position to serve other exporters. The eval- crucial information and services--for example, uation called for better targeting of beneficiaries. information about export markets, production This leads to a second issue: the selection techniques, packaging and delivery require- process. Some analysts argue that a targeting ments, and product standards. It is also usually approach will introduce bureaucracy and asserted that these firms underestimate the bene- increase the scope for discretion, slowing dis- fits of successful exporting, or overestimate the bursements and undermining the momentum risks, and therefore are unwilling to undertake necessary to build a market for services suppliers. the necessary effort and investment. Hence, there They defend the first come, first served approach, is a case for reducing their exposure to risk by which is the one almost always applied in prac- supplementing the investment they are willing to tice. Judging, however, by the rate of disburse- make through a grant. The case for a grant ment across schemes, the momentum effect has (rather than a loan or an equity infusion) is based not been good. on the premise that external benefits will accrue A third question concerns the cost-effectiveness to other firms and to the economy in general as a of the schemes, including the effect of firm size. result of the grant-receiving firms' export success. The administration and monitoring of grants are Such benefits operate through demonstration easiest when grants are large and few in number. effects, increased awareness of and interest in the The task becomes almost impossible when there country on the part of foreign buyers, and trans- is a very large number of small grants. This cre- fer of knowledge and experience acquired by the ates a dilemma because it is likely that larger firms innovating firm through labor turnover. Another applying for larger grants are least in need of increasingly important objective is to spur the them and are most likely to undertake the project development of specialized services providers in any case. This was confirmed in the Mauritius that can be beneficial to all sectors of the econo- evaluation. my. Some matching grants are therefore made Finally, it is difficult to insulate the grant available to these services suppliers, as well as to process from local lobbies and political pressures. potential exporters. There are plenty of anecdotes among practition- Various questions have been raised concerning ers concerning misuse of grants, especially of for- matching grant schemes. First and most impor- eign travel that was, in practice, only marginally tant, it is not clear whether they have actually devoted to such purposes as contacting foreign increased exports and generated external bene- buyers and exploring other export opportunities. fits. Generally, exports have increased significant- In the future, more attention needs to be paid ly, but this does not justify either the program or to the economic justification for such schemes, in the subsidy element; there needs to be evidence particular to ensure that there are not bigger pol- of additionality (the firm would not have exported icy or institutional obstacles that impede new as much without the grant) and of positive exter- exporters. When grant schemes are introduced, nalities (other firms have benefited indirectly). governments need to have a clearer understand- One of the few in-depth evaluations that tried to ing of their rationale, as well as an appreciation of 162 Export Development Policies and Institutions B O X 1 8 . 1 ( C O N T I N U E D ) the need for autonomous and streamlined man- that once the project has begun, its implementa- agement. The focus should be on small and tion will not be slowed. medium-size enterprises, and more resources should go to services suppliers. The targeting Source: Prepared by the volume editors, based on issue should be addressed in the design stage so Phillips (2001). be well targeted with respect to producers, com- Fee-based services may also lead to underprovi- modities, and markets. In the short term, existing sion of public goods (externalities) that such serv- exporters should be targeted, while selective sup- ices may generate: improvement of the country's port for potential exports may constitute a good image abroad, overall quality enhancement of medium-term target. Many TPOs have wasted industry, strengthening of the foreign exchange resources on firms with little or no export poten- reserve position, and so on. The funding prob- tial. In many successful exporting countries, small lems of some TPOs have come about because of and medium-size firms have proved to be power- dwindling budget allocations, requirements to ful innovators and exporters, and such firms may transfer fees raised to the treasury, and piecemeal benefit the most from well-targeted support. and badly structured donor financing. 5. Quality staffing. Staffing is crucial for the success 7. Evaluation of the results. The effectiveness and of a TPO. A good TPO must be able to pay salaries efficiency of TPO activities must be periodically that are similar to those paid by the private sector evaluated so that policymakers can learn from to talented staff with business experience. In most experience, refine strategies, and avoid self-per- cases TPO staff operate under civil service rules petuating activities. It should be kept in mind, that make discipline and accountability difficult however, that the process of evaluating these ser- and all too often imply unattractive pay and low vices is not an exact science. The impact of TPO motivation. Civil service staffing practices bring activities may well be felt after some delay, and bureaucracy into the TPO, with the result that exports are affected by many variables, only some staff often do not have the requisite commercial of which are under the control of the TPO. experience to interact efficiently and credibly Nonetheless, the export performance of the econ- with the private sector. A partial solution to this omy as a whole, as well as that of the enterprises problem could be to give TPOs greater autonomy which have benefited from the services of the in setting recruitment and salary standards and to TPO, can be documented and can be supplement- draw on the expertise of external consultants. ed by client surveys and reviews of the business 6. Adequate funding. A sustainable TPO should have plans of the enterprises. adequate revenues, derived mainly from domestic sources. Donor support can play a useful role in Duty Drawback and Temporary Admission starting up the TPO, demonstrating the returns to be gained from good TPO work, and bringing A number of economies that have experienced rapid best practice to bear, but such support should be growth in trade and GDP did so in the context of temporary and should be followed up with suffi- trade regimes characterized by significant import cient domestic resources. Much is to be gained by controls on the domestic market. The Republic of charging fees for services rendered because fees Korea, Taiwan (China), and Japan (in the early act as a rationing mechanism and ensure that the stages) are the main examples. The key to under- services provided are valued by the recipient. standing these experiences is to look at all the factors There are, however, clear limits to levying fees. that affect competitiveness and the incentives to pro- Some potential exporters will not have the neces- ducers to export or not. Protection creates incentives sary resources or will not fully appreciate the to sell to the domestic market; that is, it creates a bias services offered until they succeed in exporting. against exports. Protection of intermediate products 163 S E L E C T E D T R A D E P O L I C I E S A F F E C T I N G M E R C H A N D I S E T R A D E and services seriously handicaps export industries inputs and components, including machinery, at because it raises their costs to levels that are higher world prices through duty drawback or temporary than those of their potential competitors in world admission schemes. markets. As Shatz and Tarr discuss in Chapter 3 in this volume, the effect of protection on the real Benefits and Shortcomings of the Schemes exchange rate also discourages export industries. The East Asian countries managed elaborate systems Duty drawback involves repayment of duties paid that offset the bias against exports. A key element of on imported inputs that are used in the production these systems was to allow exporting firms to import of exports (Box 18.2). A problem with duty draw- B O X 1 8 . 2 D U T Y D R AW B A C K F O R D I R E C T E X P O R T E R S : T H E C H I L E A N C A S E Refunds of duties and indirect taxes on imported subsidy element. The maximum rate of subsidy is inputs used in export production can be made in around 6 percent, corresponding to the 10 per- two ways. Individual drawback systems refund cent drawback rate (which applies if exports are taxes actually paid. Fixed drawback schemes refund less than US$10 million for the entire tariff line). taxes on the basis of an estimate of the duties and The magnitude of the subsidy depends on the indirect taxes that enter into the cost of production extent to which imported inputs are used. of exports. Processing of rebate claims is generally This simplified drawback offers moderate and based on the ratio of inputs to exported outputs-- self-extinguishing subsidies for new export prod- known as input-output coefficients. These may be ucts. It is especially valuable for small exporters, self-declared by exporters or predetermined by the who may find it costly to access the regular draw- authorities, who use standard coefficients uniformly back scheme. It has emerged as an important for all exporters. The latter is more appropriate in export incentive: in 1994 the state paid a total of countries with weak legal regimes and weak US$150 million under the simplified system, administrative capacity. In many countries a major compared with just US$26 million on the regular source of delay in granting rebates is that authori- drawback. Although no careful econometric ties apply ad hoc checks, question the coefficients studies have been done on the impact of the sim- claimed by exporters, or do not (cannot) apply plified drawback on the emergence of new pretabulated standard input-output coefficients. exports, after its introduction the number of The Chilean experience with drawback provides exported manufacturing products and the values an example that could be replicated in other exported grew rapidly. Given the construction of countries. Since the mid-1980s, two duty draw- the scheme, many of the exports were "new" back systems have been in use: a regular draw- and were thus likely to be associated with exter- back, in force since 1988, under which duties on nalities related to information gathering. Over imported inputs used by exporters are rebated ex time, as exports grow, such externalities disap- post, and the so-called simplified drawback, intro- pear. This makes the automatic elimination of the duced in 1985 for nontraditional exports. The subsidy, once exports of the item exceed US$20 simplified plan applies to exports of goods that million, an attractive feature of the scheme. have not yet reached the level of US$20 million Although Chile will have to eliminate this subsidy for a given tariff line. For such exports, exporters by 2002 to comply with WTO rules (see Box 18.3, receive a cash subsidy of 3, 5, or 10 percent below), countries with a per capita income less (depending on the total value of exports for the than US$1,000 would be allowed to implement tariff line concerned) on their export value in lieu similar policies under Annex VII of the WTO Agree- of a regular drawback. Although the scheme has ment on Subsidies and Countervailing Measures. been justified on the grounds that it makes life easier for small exporters, it does in fact contain a Source: Agosin (2001). 164 Export Development Policies and Institutions back schemes is that their administration can be ners. Indirect tax rebate and drawback schemes are costly and can lead to cumbersome procedures and not considered export subsidies if they do not result delays when tariffs are high. The empirical evidence in rebates in excess of what was actually levied on suggests that in countries without well-functioning inputs consumed in the production of the exported public administrations, duty drawback is ineffec- product (see Box 18.2).1 Normal allowance for tive. Drawbacks are very difficult to administer at waste must be made in findings regarding con- tariff rates of more than 15 or 20 percent because of sumption of inputs in the production of the export- leakage, delays in payment, and fraudulent claims ed product. Drawback or duty suspension systems (Mitra 1992). Delays are particularly detrimental to on capital goods do, however, constitute an export small and medium-size enterprises and small- subsidy if they are conditional on exporting. farmer organizations. On receipt of a complaint that an indirect tax Temporary admission (also called duty suspen- rebate or drawback scheme acts as a subsidy sion) can be more effective in allowing tariff-free through overrebate or excess drawback of charges access to intermediate inputs for exporters in these on inputs consumed in the production of an situations. Temporary entry regimes do not involve exported product, the investigating authorities of payment of duties on imported inputs; rather, they the importing country must determine whether the allow entry on a duty-free basis with a requirement government of the exporting country has in place that firms document ex post that the imported and applies a system or procedure to confirm which inputs have been used in the production of exports. inputs are consumed in the production of the The main potential problem with this approach in exported product and in what amounts. Where such low-income countries with weak administrative a system or procedure exists, its reasonableness, capacity is leakage of goods into the economy (that effectiveness, and consistency with generally accept- is, the goods are not used for export production). A ed commercial practices in the exporting country frequently employed option for controlling such must be determined. To the extent that the proce- leakage is the bonded warehouse or, on a larger dures are determined to meet this test and to be scale, an export-processing zone, as described effective, no subsidy should be presumed to exist below. These are specific territories that are con- (Hoekman 1995). trolled by customs. Imports into these territories are Where there are no monitoring systems, or where not taxed on entry, but goods are taxed if they are these systems are not applied effectively, a determi- sold on the domestic market. nation of the actual inputs involved in the produc- Programs such as duty drawback and temporary tion of the exported good must be made, including admission, if properly administered, allow exporters a "normal allowance for waste." Determination of duty-free access to imported intermediates. To avoid whether the claimed allowance for waste is "nor- antiexport bias more completely, these schemes mal" must take into account the production must be extended to indirect exporters (firms that process, the average experience of the industry in do not themselves export but that sell to exporters). the exporting country, and other appropriate tech- Administration of such mechanisms is substantially nical factors. The existence of a substitution draw- more complicated, however, as most schemes in back provision under which exporters are allowed developing countries exclude small producers and to select particular import shipments on which indirect exporters. The experience of many develop- drawback is claimed cannot of itself be considered ing countries with drawback and temporary entry to convey a subsidy. Excess drawback of import has been mixed. Notably, the institutions needed for charges is deemed to exist if governments have paid effective implementation of duty drawback systems interest on any monies refunded under their draw- have been shown to be ineffective in most Sub-Saha- back schemes, to the extent of the interest actually ran African countries (World Bank 2000a). paid or payable.2 Requirements under WTO Rules Export-Processing Zones It is important that drawback mechanisms be Export-processing zones (EPZs) are enclaves within designed in a WTO-consistent manner to avoid the which governments attempt to provide a policy imposition of countervailing duties by trading part- environment and associated infrastructure that are 165 S E L E C T E D T R A D E P O L I C I E S A F F E C T I N G M E R C H A N D I S E T R A D E conducive to investors seeking to produce for policies. Experience suggests that investors weigh export.3 In a sense, EPZs are akin to duty drawback economic and political stability, labor skill compati- and temporary admission customs regimes except bility and productivity, and other similar factors that they are limited to a certain geographic loca- carefully. Incentives such as overgenerous tax pack- tion. Many, however, go beyond these customs ages or legal investment assurances may not attract regimes by addressing infrastructure and related the right type of investors (or any investors at all). issues. EPZs are generally used to achieve three Furthermore, forgoing tax revenues may be expen- goals: promotion of investment and employment in sive, especially if major public investments are made export-oriented production; increased foreign to develop the zone. exchange earnings from nontraditional exports; and When well set up and well managed, EPZs have encouragement of foreign direct investment (FDI) led to income generation and employment creation, in countries where legal, administrative (red tape, especially opportunities in nontraditional jobs for corruption), and infrastructure-related weaknesses women in the formal sector. In Bangladesh most impede investment in exportables. An added objec- employees in EPZs are women; for example, 70 per- tive is the transfer of technology and know-how cent of the employees in the Chittagong EPZ are from the EPZs to the rest of the economy. female, a much higher ratio than the national aver- EPZs are a second-best solution compared with age (ILO 1998). EPZ employment is seen by many generalized economywide reforms, but where coun- as an important factor in reducing the proportion trywide reforms are difficult to implement, they can of female poor in the Dominican Republic, from be a useful instrument in the development arsenal of 22.6 percent in 1986 to 15.8 percent in 1993. Wages governments confronting large reform agendas. They in EPZs tend to be higher, on average, than wages in allow the public and private sectors to cooperate in the rest of the country. creating the preconditions for efficient export pro- EPZ experiences range from the success stories of duction in a small geographic region, as opposed to Mauritius and Mexico to several failed zones, as in pursuing reforms and undertaking investment on an Senegal. EPZs in Mauritius managed to create more economywide basis. One of the most successful than 90,000 jobs in 1991, or 17 percent of national examples is Mauritius, where, in the mid-1990s, EPZs employment. Mexico's maquilas employed about generated more than two-thirds of gross exports and 900,000 workers in 1997, and the sector is among employed one-sixth of the work force. Net foreign the highest generators of foreign currency (second exchange earnings as a percentage of gross vary wide- to oil in 1992). By contrast, the Senegal EPZ ly, from a high of 63 percent in the Republic of Korea employed only 600 workers in 1990 and exported and Taiwan (China) in the mid-1980s to a low of 12 just US$15 million. Most experience with EPZ percent in Jamaica. The more developed the local experiences falls between these two extremes. In the economy, the higher the net foreign exchange earn- Philippines in 1996 the 4 public and 43 private EPZs ings, since backward linkages are greater. had approved investments totaling over US$2.5 bil- Effective EPZs combine clear private property lion, employed more than 150,000 people, and rights and investment regulations, no restrictions exported US$6.5 billion worth of goods. Nonethe- on foreign exchange, tariff-free imports for export less, the high exports have not meant greater back- production, moderate levels of taxation, stream- ward linkages with the domestic economy; these lined administrative procedures, and private sector generally depend on economywide reforms. The management. Public provision of basic infrastruc- consequence has been high import dependency, low ture outside the zone--telecommunications, roads, net exports (41 percent), and low net foreign and ports--can have positive spillover effects for exchange earnings. the local and national economies by facilitating eco- Attempts to use EPZs in Africa have, except in nomic activities. However, development of EPZs, Mauritius, been much less successful than else- including provision of infrastructure and manage- where. Some argue that the basic concept is flawed. ment, should be privately handled. Blame has also been placed on Africa's lack of ade- The success of an EPZ is highly dependent on a quate infrastructure and services to support the hospitable host country economic environment. business community, on the timidity or ignorance EPZs have tended to work better when the country of investors, and on the lack of indigenous entrepre- pursued sound macroeconomic and exchange rate neurs. Important reasons for the disappointing per- 166 Export Development Policies and Institutions formance of African EPZs include government are typical features of EPZs, could, in principle, be interference and the distortions introduced in the contested, to the extent that they represent subsidies operation of free trade and capital regimes. In Sene- to companies that are required to export most, if gal excessive administrative red tape and strict labor not all, of their output. Countries currently relying laws were responsible for the failure of the EPZ. The on EPZs would do well to seek clarification on their Gambia, too, has a highly regulated labor market, compatibility with WTO rules and so preempt the which raises port loading and unloading costs possibility of future disputes. sharply. These problems are not unique to Africa; similar shortcomings have undermined the EPZs in Trade Finance Jamaica and Panama (although Panama's has suc- ceeded as a free trade zone). Export finance is one of the primary constraints Quality of infrastructure is a major determinant inhibiting exports in many low-income developing of success. When the Colon Free Zone in Panama countries. Inadequacies may result from the overall was starting up, there was a good port in Colon and weakness of the financial sector, or it may reflect a reasonable road to the airport in Panama City. The difficulties in assessing the creditworthiness of Dominican Republic developed excellent air, sea, traders or the fact that traders do not have sufficient and road transport infrastructure in support of its assets to be judged creditworthy. Small firms and EPZs. Mauritius also has excellent port and airport the poor may face special difficulties in obtaining facilities. By contrast, with a few notable exceptions, access to the trade credit they need, just as they face much of Africa's transport infrastructure is in poor difficulties in accessing other parts of the financial condition. Parastatal operating companies often sector. Although ensuring the availability of trade provide poor services; economic conditions have finance is a matter that needs to be left to the private impeded public investments; and budgetary prob- sector, governments can use a number of mecha- lems have shortchanged basic maintenance. nisms to promote access to finance, especially for The establishment of a successful EPZ program smaller firms. requires simultaneous removal of most, if not all, of Two mechanisms that are sometimes used are for- the bottlenecks in infrastructure, the customs ser- eign currency revolving funds and preshipment vice, and labor regulations. A country must master export finance guarantee schemes. The revolving the creation of a probusiness environment, the pro- funds provide finance for imported inputs needed vision of infrastructure and services, international for export production. An exporter must obtain a marketing, and investor relations--and master letter of credit from a buyer; this letter allows the them all at the same time. Watson (2000) concludes exporter's bank to access the fund's foreign that the management of the socio-political-eco- exchange to pay for the imports. The guarantee nomic process of bringing about change on a broad schemes cover exporters' manufacturing nonper- front is the key constraint on success in Africa and formance risks and are generally targeted at smaller elsewhere. The reform process requires the presence firms and new entrants into the export area that of four critical factors: vision, consensus, concerted have difficulty in satisfying banks' collateral require- action, and continuity. Watson goes on to argue that ments but have obtained export letters of credit. it is typically easier for a weak state to start coordi- Note that preshipment export finance guarantees nating its actions on a small scale through an EPZ are not export credit insurance schemes; the latter while it works on nationwide reforms. insure against nonpayment by foreign buyers. EPZs are not defined or referred to in the WTO Another, more recent mechanism used by a number agreements. To the extent that subsidies are provid- of countries is grants that are conditional on match- ed through EPZs, however, the rules of the Agree- ing contributions by enterprises (see Box 18.1). ment on Subsidies and Countervailing Measures All these mechanisms for alleviating trade finance apply. Restrictions on export subsidies could constraints can be designed to comply with WTO impinge on countries' ability to employ EPZs in rules. What matters under the WTO is whether pro- future. This is especially true for countries with vision of the subsidy is conditional on exporting. As income per capita of more than US$1,000 (see Box discussed in Box 18.3, export subsidies are prohibit- 18.3, below). Lower tax rates, special credit facilities, ed for WTO members with per capita incomes and publicly provided infrastructure, all of which above US$1,000. This dimension of WTO rules is 167 S E L E C T E D T R A D E P O L I C I E S A F F E C T I N G M E R C H A N D I S E T R A D E actively enforced by members. Subsidized export craft, in which the panel found that Brazil's financ- financing has given rise to a number of disputes in ing terms for foreign buyers of its Embraer aircraft the WTO, including cases against developing coun- were illegal export subsidies. Most cases in this area tries. An example was a case brought by Canada have, however, been brought against industrial against Brazil's export financing program for air- countries. Examples are the U.S. tax treatment of B O X 1 8 . 3 S U B S I D I E S , T H E W T O , A N D D E V E L O P I N G C O U N T R I E S The types of subsidies used by governments to tries to offset the effects of foreign subsidy pro- support economic activities include direct pay- grams are not abused. A subsidy is deemed to ments or grants, tax concessions, soft loans, and exist if there is a financial contribution by a gov- government guarantees and equity participation. ernment (or public body). This may involve an They may be firm- or industry-specific, or they actual or potential direct transfer of funds (such may be generally available. In practice, it may be as grants, loans, equity infusions, or loan guaran- difficult to determine whether a subsidy is, in tees), forgoing of government revenue (tax con- fact, specific. Subsidies that are sector-specific cessions or credits), or the provision or purchase (say, to health, education, or transport) may have of products other than general infrastructure. economywide objectives. Conversely, subsidies Government funding of a private body to carry that are economywide in scope may be effective- out a function that would normally be vested in ly industry-specific. An example is the pursuit of the government, and any form of income or price an environmental objective the attainment of support, is also covered by the definition. In all which requires taxes or subsidies that primarily these cases the measure must confer a benefit on affect specific sectors such as the chemical or the the recipient or recipients. The General Agree- automotive industry. ment on Trade in Services (GATS) contains no subsidy disciplines for services (see Chapter 32, WTO Rules regarding Subsidies by Sauvé); special rules apply to agriculture, as The WTO rules concern specific subsidies, since discussed in Ingco and others (forthcoming). economywide subsidies are presumed not to dis- Until recently, three categories of specific subsi- tort the allocation of domestic resources with dies were distinguished in the agreement: pro- regard to tradables. A subsidy is considered non- hibited, actionable, and nonactionable. There specific if eligibility for, and the amount of, the were three types of nonactionable subsidies: subsidy is determined by objective criteria. An those provided to support research, to aid disad- example would be subsidies that focus on firms of vantaged regions, and to facilitate the adaptation a particular size (micro or small and medium-sized of plants to new environmental regulations. How- enterprises). The subsidy must not be conditional ever, as there was no consensus in the Commit- on export performance or the use of domestic tee on Subsidies and Countervailing Measures to inputs, in which case it is deemed to be specific. extend the relevant provision beyond 1999, this Equally, a subsidy that is limited to an enterprise, provision--and thus the category of nonaction- industry, or enterprises within a designated geo- able subsidies--is understood to have lapsed. graphical region is considered to be specific. There Currently, therefore, specific subsidies are either are, therefore, four types of specificity within the prohibited or actionable. meaning of the WTO Agreement on Subsidies and Subsidies that are contingent--either formally or Countervailing Measures (the SCM Agreement). in their effect--on export performance or on the The WTO subsidy rules attempt to strike a bal- use of domestic over imported goods are prohibit- ance between the need to agree on minimum ed, except for some developing countries, as noted standards regarding those subsidies that may not below. An illustrative list of export subsidies be used because they distort trade and the need annexed to the WTO SCM Agreement cites the to ensure that measures used by importing coun- provision of products or services, including trans- 168 Export Development Policies and Institutions B O X 1 8 . 3 ( C O N T I N U E D ) port, for use in export production on terms more from a command to a market economy were per- favorable than for use in the production of domes- mitted to apply prohibited subsidy programs, and tically consumed goods. The list also cites export debt forgiveness was not actionable. Least devel- credits and guarantees or insurance provided at a oped countries (as defined by the UN) and certain cost that is inadequate to cover the long-term oper- other countries with a GNP per capita below ating costs and losses of the insurer, unless, where US$1,000 are exempted from the prohibition on export credits are concerned, a WTO member export subsidies.* In the case of this latter group, applies the interest rate provisions of the OECD once their GNP per capita reaches US$1,000, non- Arrangement on export credits. If a dispute settle- conforming subsidies must be eliminated within ment panel finds that WTO members are using eight years. Developing-country WTO members at export or import substitution subsidies, the remedy or above the threshold income level are subject to will be a requirement that the measures be with- a standstill requirement and must also phase out drawn, generally within a three-month period. their export subsidies by January 2003. All devel- Actionable subsidies are those that are permit- oping countries may request a further extension of ted but may, if they cause adverse effects to the this phaseout period, and certain developing interests of a WTO member, give rise to consulta- countries are eligible for an extension for particular tions, invocation of dispute settlement proce- export subsidies under procedures adopted in dures, or the imposition of countervailing duties November 2001. If an extension is granted in by the affected importing country. Adverse either case, annual consultations with the Subsi- effects include injury or threat thereof to a dies Committee must be held to determine the domestic industry, nullification or impairment of necessity of maintaining the subsidies. Developing tariff concessions, or serious prejudice to the countries that have become competitive in a prod- country's exporting interests. Serious prejudice uct--defined as having a global market share of at may arise if the subsidy reduces exports of other least 3.25 percent--must phase out any export WTO members, results in significant price under- subsidies over a two-year period. cutting, or increases the world market share of Under the GATT, developing countries were the subsidizing country in a primary product. free to use export subsidies. This is no longer the case under the WTO. The introduction of the pro- Provisions Affecting Developing Country Members hibition on export subsidies has implications for Developing countries benefit from higher de min- countries approaching the $1,000 per capita imis thresholds in countervailing duty investiga- income threshold that assist firms in penetrating tions of their products by trading partners. If the new markets through, for example, advertising subsidy is less than 2 percent of the per unit value campaigns or matching grant schemes. Such of products exported, developing countries are schemes could be regarded as export subsidies if exempt from countervailing measures (whereas the provision of the grant element is made condi- this figure is 1 percent where a product from an tional on exports. industrial country member is under investigation). An exemption also applies if the import market * A formula has been established to adjust this thresh- share of a developing country under investigation old income level, originally set in 1994, to account for is less than 4 percent, provided that the aggregate inflation. share of all developing countries under investiga- To be eligible for an extension under these proce- tion with shares less than 4 percent is below 9 per- dures, the developing country's share of world mer- cent of total imports. chandise export trade must be no greater than 0.10 Special provisions for developing and transition percent, and its gross national income for 2000 must be economies are included in Articles 27 and 29 of at or below US$20 billion. the SCM Agreement. Through December 2001, WTO members in the process of transformation Source: Editors, based on Hoekman and Kostecki (2001). 169 S E L E C T E D T R A D E P O L I C I E S A F F E C T I N G M E R C H A N D I S E T R A D E so-called foreign sales corporations (FSCs), under relevant to other low-income countries (Harrold, which U.S. firms with exports that have at least 50 Jayawickrama, and Bhattasali 1996): percent U.S. content can reduce tax burdens by 15 to 30 percent, and the preferential government · Support for on-the-job training, through payroll loans on noncommercial terms granted by Aus- tax refunds rather than subsidies, as well as for tralia. In both instances dispute panels found that public training institutions that are demand the measures violated WTO rules (Hoekman and driven Kostecki 2001). · Technical assistance to enterprises for access to technology and design skills and for development of external markets Conclusion · Export credit support mechanisms, especially Much has been made of the extent to which some preshipment finance high-performing East Asian economies used export · Development of simple duty-exemption schemes promotion policies to support their impressive · Development of industrial parks and export-pro- export drives and of the fact that some of these mea- cessing zones. sures are no longer available to latecomers. It is true that the rules of the game have evolved. Yet it is also None of these, with the possible exception of true that many governments have tried and failed to EPZs, currently present problems in the context of replicate various elements of the East Asian model of the WTO. export promotion--sometimes at considerable More fundamental to replicating the East Asian expense in terms of government revenue and misal- success are an unequivocal commitment on the part located resources. Furthermore, poor countries can of government to working with the private sector in ill afford to engage in competitive subsidization of the pursuit of joint goals and a long-term vision that their exports, which will often benefit relatively rich places export development at the heart of the nation- consumers abroad, and they will inevitably lose out al development strategy. Many countries have yet to in any such contest with richer countries. It is clearly establish these basic preconditions, without which in their interest to discipline the use of export subsi- microeconomic interventions are likely to be wasted. dies. Export industries are still often taxed implicitly, if not explicitly, by inefficient government services or Notes poorly functioning markets. There is plenty of work 1 Indirect tax rebate schemes allow for exemption, remission, or to be done on these fronts, and there is an ample deferral of prior-stage cumulative indirect taxes levied on range of instruments compatible with WTO rules inputs that are consumed in the production of the exported that developing country governments have yet to product. Drawback schemes allow for the remission or draw- master and that would go a long way toward reduc- back of import charges levied on inputs that are consumed in the production of the exported product. ing the antiexport bias in their economies. In a review of the lessons from East Asia for 2 This rule strengthens the incentive to use temporary admission African trade and industrial policy, five priorities and duty waiver mechanisms rather than drawback. were identified at the project level that are equally 3 This section draws on Madani (1999) and Watson (2000). 170 19 B I J I T B O R A Trade-Related The Agreement The fact that there is a separate Investment text called an "agreement" is a paradox. In essence, all the Measures TRIMs agreement does is to clarify the application of GATT Articles III.4, on national treat- ment, and XI.1, on quantitative restrictions. It does not even define a trade-related invest- ment measure. Instead the approach that was taken was to A include an illustrative list of lthough it is only five pages long, measures that are inconsistent with these two key the WTO Agreement on Trade- paragraphs of the GATT. The list covers both Related Investment Measures (TRIMs) has become TRIMs that are mandatory or enforceable under a central issue in the debate on the relevance to domestic law and measures for which compliance is developing countries of the multilateral trading necessary to obtain an advantage. There is no text agreements and the WTO. A combination of factors that specifically addresses issues related to granting led to the inclusion of investment in the work pro- national treatment to investors. gram of the Uruguay Round negotiations. These The agreement allowed for a notification period of included a changing perception of the role of for- 90 days, beginning January 1, 1995, for WTO mem- eign direct investment (FDI) in development and bers to notify the WTO of measures that were not in the intense debate on the linkage between GATT conformity with the agreement. After notification, a rules and foreign investment policy stemming from member was allowed a transition period, the length the U.S.-Canada dispute on Canada's application of of which depended on its level of development, to performance measures to foreign firms.1 Despite an bring its laws into conformity with the agreement. ambitious start to the negotiations, the final text Developing countries were allowed five years; least- was limited in scope and coverage. developed countries were allowed seven years. The purpose of this chapter is to assess how well The agreement is a rather modest attempt at dis- the TRIMs agreement has been implemented and ciplining policies that are targeted at foreign enter- to identify lessons or issues that may be relevant to prises, and it was the outcome of conflicting the mandated review of the agreement and to the positions about the extent to which investment future multilateral trade negotiations on invest- issues should be covered by the WTO. In the ment called for in the 2001 Doha Ministerial Decla- Uruguay Round, many developing countries resist- ration. ed the extent to which market access for foreign 171 S E L E C T E D T R A D E P O L I C I E S A F F E C T I N G M E R C H A N D I S E T R A D E firms would be covered. As a result, the negotiations was some variance in the approach taken in the focused on policies that applied to the operations of application of these policies. The automotive indus- foreign firms. Even then, the negotiations proved try was the one most frequently subject to such difficult, as there was no agreement on whether or policies, but some members applied local content not a specific policy instrument was trade distort- schemes in a general fashion across all industries. ing. Furthermore, some developing countries took The second most prominent sector was agriculture. the position that they should have access to policy Thus, in the context of notification it would seem instruments that could be used to offset any per- that the agreement worked well in that approxi- ceived negative effects associated with the opera- mately 20 percent of the WTO membership gave tions of transnational corporations (Mashayeki notification that they had adopted policies which 2000). contravened the agreement. (After all, an agreement When the TRIMs agreement is compared with which imposed disciplines but resulted in no mem- other investment provisions, or with plurilateral bers notifying that they had policies inconsistent attempts at investment rules, including nonbinding with these disciplines would not add much to the ones such as those of the Asia-Pacific Economic international trading system.) None of the coun- Cooperation (APEC) or the World Bank Guide- tries, however, were developed or least-developed lines, it falls considerably short in terms of coverage. countries. Furthermore, some developing countries Nevertheless, as a multilateral instrument, the complained that the notification period was too agreement has allowed investment issues to be dis- short and that a country which was unable to notify cussed in the context of multilateral negotiations. in the time allowed would not be able to enjoy the These discussions have continued through the benefit of the transition period.2 Working Group on Trade and Investment (created in 1996 at the WTO ministerial meeting in Singa- Disputes pore), where members have further explored the linkages between trade, FDI, and development. Given that any member can initiate dispute pro- ceedings against any other member, it makes little sense to simply add up the number of disputes. Implementation Using that approach, one would find that 16 No established template exists with which to evalu- requests for consultation were initiated, with 2 pro- ate the implementation of an agreement. Neverthe- ceeding to an actual dispute panel. The problem less, there are some sensible criteria that can be here is that one WTO member may find itself adopted. For example, is the agreement neutral with defending a particular policy against a number of respect to its application? Has it been successful in other members. This was the case of the notifica- terms of both avoiding and settling disputes? The tions against Indonesia, when Japan, the European responses to these questions will be taken up below Union (EU), and the United States each filed notifi- in the context of three issues: notifications, disputes, cations citing the same policy. Similarly, Japan and and adequacy of the transition periods. the EU filed notifications against Canada. Once this double counting is allowed for, only seven coun- tries--four developing countries and three indus- Notification trial countries--had to defend their policies. The TRIMs agreement allowed any member access An interesting aspect of the notifications for dis- to an extended transition period for bringing its pute is that in each case the complaint listed other policies into compliance with the agreement, if and policies in addition to those that were claimed to be only if these policies were notified within 90 days of inconsistent with the TRIMs agreement. This is an the commencement of the agreement. Twenty-six important point in that it reflects on the use of mul- members, all developing countries, notified a vari- tiple types of measures in the context of general ety of policies. The economic characteristics of industrial policy objectives as opposed to an isolat- these countries varied considerably. The most com- ed or targeted use of intervention. The complaints mon policy adopted by these countries was local always listed as additional areas of conflict GATT content schemes, and the second most frequently articles and also the Agreement on Subsidies and used policy was foreign exchange balancing. There Countervailing Measures (SCM). In one recent case 172 Trade-Related Investment Measures the Agreement on Trade-Related Aspects of Intellec- tract between the government and a firm included a tual Property Rights (TRIPS) was cited. 3 policy that was not in compliance with the TRIMs Two of the notifications, against Canada and agreement, but the policy's removal would have had Indonesia, could not be resolved outside the dispute legal consequences for the government. settlement mechanisms and had to proceed to panel The most often cited reason for requesting exten- reports. In both cases the complainants were indus- sions was the financial crises that some developing trial economies: Japan, the EU, and the United countries suffered in 1997­98. Argentina, Malaysia, States in the first case, and Japan and the United the Philippines, and Thailand each cited economic States in the second. The decision in each case went crisis as a primary reason for requesting an exten- against the defending party and required it to bring sion, since the reforms needed to comply would its laws into conformity with the TRIMs agreement. exacerbate the structural adjustment problems stemming from the crisis. Colombia and Pakistan each gave specific development reasons for their Adequacy of the Transition Periods requests. Colombia noted the difficulties in trans- The transition period for developing countries to forming its economic model, especially in substitut- implement the TRIMs agreement expired on ing away from illegal crops, which, it argued, would December 31, 1999. Although the agreement did require a domestic absorption or local content poli- not specify a deadline for requests for extension, by cy to ensure that farmers were able to sell their legal May 31, 2000, nine requests had been lodged with produce. Pakistan asserted that TRIMs conformity the WTO. In March 2001 Egypt formally filed a might be contrary to its development interests; request, bringing the total to 10. In general, there- opening the economy to import competition would fore, the record for implementation for the TRIMs prevent the country from exploiting domestic agreement is not too bad; approximately a third of resources optimally or promoting transfer of tech- the members that notified policies were not able to nology, employment, and domestic linkages. Pak- comply in time. This does not mean that serious istan also asked for a minimum period for implementation problems do not exist. There are extension, not a maximum. serious problems, but they differ substantially from Another reason cited for requesting an extension the type of implementation problems in the TRIPS was the interaction between preferential trade and the Customs Valuation agreements. In the agreements and multilateral obligations. Argentina, TRIMs case, the drafting of legislation to repeal in its request, specifically noted the importance of local content schemes is fairly simple and straight- negotiations within the context of the Southern forward, and there is no substantive technical need Common Market (MERCOSUR) Common Auto- in this regard. motive Policy. Mexico did not specifically mention There appear to be two issues related to transition the North American Free Trade Agreement periods. First, some members argued that they (NAFTA), but there is an inconsistency in the lacked the capacity to identify measures that were phaseout periods for TRIMs in NAFTA and in the inconsistent with the TRIMs agreement and hence WTO. were unable to meet the notification deadline. This The most interesting aspect is the length of exten- meant that for these countries, according to their sion requested, which ranged from one year (Chile) argument, the effective transition period was zero to Pakistan's (minimum) seven years. Except for years, which is clearly inadequate. The second issue Malaysia, countries asking for additional time is that in some cases the members that did notify because of adjustment problems generally request- did not appear to be prepared to meet the deadline. ed periods longer than four years. Many members conducted activities in their coun- tries related to the implications for the affected Issues Arising from the Design and industries arising from compliance, but none actu- Implementation of the TRIMs Agreement ally had implemented alternative policies. One member country (Chile) had drafted the relevant A number of issues arising from the TRIMs agree- laws in conformity with the TRIMs agreement, but ment go beyond the problem of the time allocated the laws had not been approved by parliament. In for the transition period. These issues need to be another member, Romania, a legally binding con- addressed in the context of a review or in new 173 S E L E C T E D T R A D E P O L I C I E S A F F E C T I N G M E R C H A N D I S E T R A D E negotiations. Six such issues are identified in this articles and the TRIMs agreement. In particular, section. since the TRIMs agreement is independent, this would imply that any future negotiations would have to take this relationship into account. Ownership Neutrality Although the TRIMs agreement presupposed a Voluntarism and Backsliding direct link with the GATT, there was still some con- fusion regarding whether a policy that violated A number of countries have made clear that the GATT articles automatically violated the TRIMs central issue in the TRIMs agreement is not the agreement. As indicated above, the problem is that length of time allowed for implementing the obliga- the TRIMs agreement did not introduce new lan- tions but the obligations themselves. For example, a guage concerning disciplining policies; it merely proposal by 12 countries arguing that the text of the referred to the GATT articles. This raises a question agreement should be changed so that commitments of how the TRIMs agreement actually fits into the to TRIMs are voluntary (WTO, WT/GC/W/354) WTO multilateral trade agreements (MTAs) and clearly asked for a derogation of the commitment whether it allows or prevents a measure directly tar- and, indeed, requested a kind of special and differ- geted at a foreign enterprise. Part of the confusion ential treatment that does not exist.6 Although the lies in the extent to which the TRIMs agreement is proposal is inconsistent with existing rules and with actually an instrument related to foreign invest- rulings on the application of the rules, it does raise ment. The term "investment" is used in the title, and the question of whether some developing countries there was a general presumption that investment- were adequately prepared for the negotiations.7 In related policies which affect trade were to be this context the approach that was taken was to ban addressed. This perhaps was to be the feature that outright such policies without any agreed phasing distinguished the GATT from TRIMs. as, in say, the Agreement on Textiles and Clothing. One view is that the TRIMs agreement codifies a For example, if this approach were adopted, a coun- GATT panel decision on the Canadian Foreign try that notified a local content scheme would have Investment Review Act (FIRA). This, however, is been required to notify how the scheme was to be technically wrong because the TRIMs agreement is implemented and the minimum specified local con- a stand-alone agreement and needs to be interpret- tent. Then a simple phasing-out of 20 percent over ed independent of GATT rules.4 But since TRIMs is five years would have met the transition period independent, does it have any direct relevance to deadline while allowing individual members the foreign firms, and does it go beyond GATT rules, possibility of being able to monitor the extent of especially in the context of Articles III and XI? A implementation.8 number of developing countries have asserted that In the absence of a well-defined phase-in pro- it does.5 Yet the 1998 panel report on the TRIMs gram, members that notified under the TRIMs dispute involving Indonesian policies stated conclu- agreement were obliged only to bring their laws into sively and clearly,"We note that the use of the broad conformity with the agreement. Indeed, during term `investment measures' indicates that the meetings of the Committee on Trade and Invest- TRIMs Agreement is not limited to measures taken ment, a number of questions were put to WTO specifically in regard to foreign investment. Con- members that notified about their implementation trary to India's argument we find that nothing in the programs, and these members rightly replied that TRIMs Agreement suggests that the nationality of they were under no obligation to respond in detail. the ownership of enterprises subject to a particular Therefore, in simple terms, as with any obligation measure is an element in deciding whether that where the implementation causes difficulty and the measure is covered by the Agreement" (para. 14.73). implementation program is voluntary, there is no Therefore, the TRIMs agreement is not confined to incentive to comply with the obligation. This could policies targeted at foreign firms; it, like the GATT, be an issue to be taken up in the context of a review is ownership neutral. The importance of the of the agreement, especially if new disciplines are to Indonesia panel decision and of the subsequent be considered. Nevertheless, it should be kept in panel decision on Canadian policies is that these mind that this is a problem for only a third of the findings clarify the relationship between the GATT countries in less than a quarter of the notifications. 174 Trade-Related Investment Measures Structural Adjustment as a Defense sessions have been held between members interest- ed in implementation and members that have The preamble of the TRIMs agreement states that the requested extensions. Some of these sessions relate agreement takes into account the trade, development, to notifying countries' plans to bring their laws into and financial needs of developing countries. In this conformity. As a rule, none of the notifying coun- respect, four of the applications for extensions (by tries had developed an implementation plan or Argentina, Malaysia, the Philippines, and Thailand) identified alternative policies that could be used to cited the financial crises that had hit the East Asia and achieve the same objective.11 Latin America regions.A fifth (Colombia) cited a par- Furthermore, in the cases involving the automo- ticular circumstance of structural adjustment, from tive sector, notifying countries have, in general, not illegal to legal farming. These two kinds of cases, with- bound most of their tariff lines. This would create out question, are specific to developing countries and an opportunity, given the caveats identified above, appear to be legitimate grounds for an extension. for the use of price-based measures to replace quan- In the Indonesia panel dispute the arbitrator took tity-based measures. That avenue does not appear up this precise point. In this case he was required to to have been pursued, perhaps because of the incen- rule on the length of time that it would take Indone- tive compatibility of voluntary schemes and also sia to implement the panel ruling and bring its laws perhaps because any increase in tariffs would not be into conformity with the TRIMs agreement. The interpreted positively by investors (domestic and complaining parties (the EU, Japan, and the United foreign) or by other WTO members. States) argued that structural adjustment should An issue related to when a policy should be not be considered a defense because it is part of any removed is the sequencing of reform. TRIMs are obligation to liberalize.9 typically used in conjunction with other policies. The arbitrator's final ruling was that Indonesia Furthermore, given the existence of certain policies, should be allowed 12 months: 6 months for admin- TRIMs could have positive welfare effects. One fac- istrative consultations and another 6 months tor that was not taken into account during the because it was a developing country. Canada was Uruguay Round negotiations was how the removal given less time to implement. Thus, the precedent of certain TRIMs, without addressing companion was confirmed that developing countries require policies, would affect trade. For example, local con- more time than industrial countries. The decision tent schemes are usually combined with a subsidy. also appears to set a precedent regarding structural The TRIMs agreement disciplines trade policy adjustment as a defense. Indeed, the rulings in this instruments but not the subsidy policies. One view and previous cases indicate unambiguously that would be that liberalization should not proceed, without the cover of the TRIMs agreement, any since incentives have not been disciplined. The WTO member with policies that are in the TRIMs other view, of course, is that both should be disci- annex would have to bring these laws into conform- plined at the same time. ity within a 12-month period.10 Nevertheless, the central issue remains: in order A second issue related to structural adjustment is to implement obligations, members need to have a whether a particular local content scheme is effective. solid understanding of what they committed to and The Philippines has regularly failed to meet its local a solid vision of how to implement these obliga- content targets (Abrenica 2000). Malaysia and Pak- tions. This would include the timing and sequenc- istan, by contrast, exceeded their targets by a signifi- ing of liberalization policies to suit countries' own cant margin,suggesting that any structural adjustment needs. would not pose an immediate problem given that the local content scheme is not affecting the production Interface between Preferential Trade Agreements decisions of firms (Ali 2000; Tyndall 2000). and Multilateral Trade Rules The general perception of preferential trading Timing and Sequencing of Policies: Implementation agreements (PTAs) is that they involve a degree of Plans liberalization that goes beyond the obligations at Since the transition period for implementation has the multilateral level. Indeed, this is precisely the closed, a number of bilateral question-and-answer case, especially in the context of tariff and services 175 S E L E C T E D T R A D E P O L I C I E S A F F E C T I N G M E R C H A N D I S E T R A D E liberalization. Regarding investment, however, there (in the context of right of establishment) and per- are a number of cases in which the investment pro- formance requirements. Such an approach would visions lag behind the TRIMs agreement. The Aus- view the existing TRIMs agreement as a basic tralia­New Zealand Closer Economic Relations framework within which to deal with performance Agreement, which is widely agreed to be one of the requirements that are inconsistent with GATT arti- most forward-looking preferential trade agree- cles and for which rules on market access would be ments, does not even have an investment provision, required. In addition, some attention would also yet it has liberalized services and goods trade (Scol- have to be focused on the definition of FDI and the lay 1996). scope of the dispute settlement mechanism. In The TRIMs agreement simply specifies a transi- essence, a new architecture would be required to tion period of five years for developing countries deal with investment issues. without foreshadowing any conflict with PTAs. The EU articulated this view in the context of the Argentina, however, in its request for an extension, preparations for the Fourth Ministerial Conference cites specifically its intention to develop an automo- in Seattle in 1999. Needless to say, some developing tive component in the MERCOSUR trade agree- countries would be opposed to such an approach. ment. Similarly, the transition period provisions in Furthermore, the issue would have to be NAFTA are inconsistent with those in TRIMs. approached carefully, since commercial presence is already part of the General Agreement on Trade in Services (GATS), and in a positive list manner. It Lack of Criteria for Extensions could be argued that, for consistency, if market Perhaps the most obvious issue that has arisen is the access issues were to be part of an investment lack of any criteria for an extension of the transition instrument, a similar approach, using a positive list periods. The only possible reference point is the by sector, might be required. suggestion that the transition period vary with the level of development, with the least-developed Renovation countries allowed two more years than developing countries. Clearly, the extension issue would have Another option would be to renovate the existing been much easier to resolve had the criteria for architecture by adding an extension or reducing granting extensions been unambiguous and trans- obligations. This option would be preferable to the parent. new architecture option, since it would automati- cally preclude any discussion of market access issues. Still, it may prove not to be an easy path for Approaches to the Review of the TRIMs negotiation, since the debate about the trade effects Agreement of investment measures was not resolved during the Article IX of the TRIMs agreement requires a Uruguay Round. review of the agreement no later than January 1, To circumvent this problem, one approach might 2000. Such a review had not begun at the time of be to adopt a traffic light system as in the SCM writing, in part because of the linkage between the agreement. TRIMs that were deemed to be trade review and a new trade round and also because the distorting and directly inconsistent with the existing agreement has yet to be implemented in its entirety. provisions of Article 2 would be classified in a red This section examines some of the options in the box, those that were not inconsistent would be in a context of a review of the agreement. green box, and those on which there is a debate as to their effect would be in a yellow box. While not novel, the approach could be used to Full Negotiation on Investment: A New Architecture accommodate some of the concerns of developing The obvious option, given that the TRIMs was a countries. The traffic light approach could use crite- compromise agreement, is to bring the negotiations ria other than trade effect to determine the alloca- about investment back full circle to the original tion of policies among the different boxes. Indeed, mandate provided for in Uruguay in 1986. That the idea of a "development" box separate from the wording was broad enough to accommodate an triple-box traffic light approach has been mooted in instrument dealing with both market access issues other negotiations. Perhaps the only caveat is that 176 Trade-Related Investment Measures the criteria for allocating policies, whether trade the multilateral trading system on how to deal mod- effects, development dimension, or something else, estly with issues related to investment. One of the should be made clear at the outset. major steps forward has been greater clarity about When examining whether the annex list should interpreting GATT rules as they relate to policies be extended, high priority should be given to export aimed at favoring one industry over another. At the performance requirements. Currently, the wording same time, a third of the WTO members that were allows the use of this policy, since it does not restrict required to implement their obligations failed to do trade. It does, however, clearly affect trade and so, suggesting that the agreement is far from perfect. should be addressed. Another issue to reexamine is Moreover, as is discussed in Chapter 42, by Hoek- the application of local content policy in light of its man and Saggi, there is no consensus as to how to use elsewhere. For example, the recent Trade Policy move investment issues forward in the WTO, if at all. Review for Canada highlighted that these policies This chapter has identified a number of key issues are used at the subnational level. that have made implementation of the TRIMs Renovating the existing architecture need not be agreement problematic. These issues are divided confined simply to an extension; it could also into two areas: ambiguity in the wording of the include cutting back the existing agreement. Indeed, TRIMs agreement, which has made interpretation this approach would assume that the existing agree- of obligations difficult, and lack of capacity on the ment has gone too far and, given the implementa- part of some developing countries to fully under- tion difficulties, needs to be curtailed. This stand the scope and implications of these obliga- approach would reduce the policies listed in the tions. These issues have created a tension between annex list or lengthen the transition periods along the generally accepted notion of efficiency and the the lines requested by some members in their Seat- broader definition of development. Adherence to tle proposals, or both. the latter may require conceding the former. How- ever, economic theory and a body of empirical evi- dence provide strong support for the proposition No Change that neutral policies designed to enhance the effi- A third approach might be to leave the TRIMs ciency of investment are better than targeted gov- agreement as it is until all WTO members have ernment intervention at attracting foreign completed implementing their obligations. This investment and enhancing its contribution to devel- standstill approach would be acceptable to the small opment (see Chapter 42). number of countries that have requested extensions of the transition period but is unlikely to receive Notes much support from the industrial country mem- bers and perhaps a significant majority of develop- The chapter has benefited from the comments and discussion of ing country members that have conscientiously participants at the Special WTO Seminar on Implementation, held implemented their obligations. on May 20, 2000. 1 Canada: Administration of the Foreign Investment Review Act (BISD 30S/140, 1984). Conclusions 2 The countries making this argument include Cuba, the The debate on the inclusion of investment issues in Dominican Republic, Egypt, El Salvador, Honduras, India, the multilateral trading system that started in the Indonesia, Malaysia, Nigeria, Pakistan, Sri Lanka, and Uganda. late 1940s continues today. The TRIMs agreement 3 Brazil versus the United States in the context of patent protec- that was part of the Uruguay Round package was an tion (G/TRIMS/D/17). attempt to address some of the issues related to 4 This interpretation was confirmed in the July 1, 1998, report of investment policies. In the end, however (as con- the panel on Indonesia: Certain Measures Affecting the Automobile firmed by two panel decisions), the agreement sim- Industry. Paragraph 162 of the report states, "This reinforces the ply addresses trade-distorting policies, regardless of conclusion that the TRIMs agreement has an autonomous legal whether they are targeted at foreign or domestic existence, independent of that from Article III." enterprises. 5 Part of this is explained in para. 14.7 of the Indonesia panel The six years of experience with the TRIMs agree- decision, which states, "Indonesia also supports the argument ment has been an invaluable learning experience for put forward by India, a third party, that the TRIMs agreement 177 S E L E C T E D T R A D E P O L I C I E S A F F E C T I N G M E R C H A N D I S E T R A D E is basically designed to govern and provide a level playing field 8 In the implementation of the Agreement on Textiles and for foreign investment and that therefore measures relating to Clothing, the flexibility of the wording has allowed some internal taxes or subsidies cannot be construed to be trade WTO members to back-load their implementation so that related investment measures." the greatest liberalization does not occur until the latest pos- sible date. 6 The Uruguay Round Agreement changed significantly the con- cept of special and differential treatment by allowing for differ- 9 Bora and Neufeld (2000), in a study of how the five affect- ent transition periods. Previously, special and differential ed Asian countries used tariffs to respond to the financial treatment applied to market access and measures that would crisis, found that only Thailand raised tariffs above the violate the most-favored-nation principle. The proposal by the bound levels, and only in a few lines. Furthermore, the tar- 12 countries asks for an exemption from an existing obliga- iffs that were raised were typically on luxury products, tion--indeed, an obligation that was in the GATT before the which the authors interpret as evidence that the role of tar- Uruguay Round. Furthermore, the request also implies that the iffs during the crisis was one of revenue raising as opposed panel rulings in both the GATT Canada case and the two WTO to protection. cases should be ignored. 10 Canada was awarded eight months as a "reasonable period of 7 Another view could be that the outcome was the maximum time" (WT/DS142/12). that some developing countries could accept, given that some industrial countries actually wanted the investment provisions 11 This was confirmed through personal interviews with delegates in the Uruguay Round to go beyond existing GATT disciplines. from the notifying countries. 178 20 G A R R Y P U R S E L L Local Content duction facilities and to reduce car imports. The most direct and widely used means of doing Policies so has been to impose quantita- tive restrictions on imports while at the same time offering international auto producers Australia's Experience opportunities to establish local factories, subject to the condi- with Automobiles tion that they go beyond assem- bly of imported, completely knocked down (CKD) packs and incorporate specified levels T of "local content" in the form of he automobile industry worldwide domestically produced components. is technology-intensive as regards The experience of many countries reveals that both its processes and its products, and it is charac- such trade-related investment measures not only terized by considerable economies of scale and a involve very high economic costs to consumers, high degree of specialization in component manu- government budgets, and the economy in general facture. Largely for these reasons, the global indus- but also, on balance, have retarded rather than try has increasingly become internationalized, with advanced indigenous technological capabilities. component production for individual models locat- This chapter summarizes the long and well-docu- ed in many countries and assembly concentrated in mented Australian experience with local content large domestic markets or in countries that are a plans for autos and draws some lessons from this base for regional exports. At the same time, con- experience that are relevant for developing coun- sumers demand a large variety of models, at com- tries using or considering similar policies. petitive prices. Imports of cars therefore account for high shares of the total supply even in the largest Local Content Policies in Australia: national markets, including the United States and A Brief History the European Union (EU). Nevertheless, economic nationalism and the Australian policies aimed at creating a domestic auto belief that the automobile industry is a transmitter manufacturing industry started in the mid-1930s, par excellence of the latest industrial technologies when extra protection against competing imports have led many countries, at some stage of their eco- and a variety of subsidies were offered to induce nomic history, to attempt to become fully or pre- General Motors to produce a local car, the Holden. dominantly self-sufficient in car production. In Until 1960, production of the Holden and of subse- pursuit of this goal, they have sought to persuade quent cars with substantial Australian content by international auto firms to establish domestic pro- British Motors and Ford was supported by local con- 179 S E L E C T E D T R A D E P O L I C I E S A F F E C T I N G M E R C H A N D I S E T R A D E tent arrangements, concessional loans, and, crucial- become one of the most highly protected industries ly, by the way Australia's general system of import in Australia. On average, effective protection of Aus- licensing was applied to the industry. In 1960 the tralian manufacturing was then about 20 percent general import licensing system was abolished, and and had been consistently declining for almost 20 competition from imported automobiles began to years. In the auto industry, however, nominal pro- threaten the market shares and profitability of the tection was about 85 percent--that is, ex-factory established producers. In response to intense lobby- prices of locally produced cars were about 85 per- ing by these producers, in 1965 policies were adopt- cent above the duty-free prices of imported cars. ed that greatly increased the level and complexity of Effective protection of the value added of the auto assistance to the industry over the next 20 years. producers participating in the local content plans A central element of these policies was a series of was estimated to exceed 250 percent. On average, "motor vehicle manufacturing plans" under which nominal protection of local manufacturers of origi- local producers were provided with tariff conces- nal equipment components was estimated at about sions on imported components if they met specified 67 percent, and effective protection to their value levels of local content in the vehicles they produced. added was estimated to be 162 percent. The first of these plans required lower levels of local Profit rates in the component sector were well content for small-volume producers than for high- above the general level of the rest of Australian man- er-volume producers. This discouraged large-scale ufacturing but, on average, were considerably lower production and led to a proliferation of models and in the producer/assembler sector, even though some a corresponding fragmentation of production large producer/assemblers consistently earned very among component suppliers. In 1975 the small-vol- high profits. High protection thus reflected and ume provisions were phased out, and a single local resulted in high production costs rather than high content requirement of 85 percent was introduced. profits. A primary reason was that the system Despite the tariff concessions on imported inputs, encouraged the fragmentation of the market and the the import tariff protecting the industry was not loss of scale economies. In 1985 approximately sufficient to make all producers profitable, and in 380,000 cars were produced under the local content 1966 it was increased from 35 to 45 percent. By scheme, but this production was divided among 5 1975, the industry had convinced the government companies operating 8 manufacturing and assembly that the 45 percent tariff was inadequate, and quan- plants and producing 13 different basic car models. titative import restrictions were introduced that Small production runs (by international standards) limited imported cars to a market share of 20 per- for this large number of models led to small produc- cent. These import restrictions were supposed to be tion runs and high costs for many local component temporary, but in fact they were extended for 13 suppliers. In addition, the penalties for not meeting years, until April 1988. Although import controls local content commitments meant that a number of became the main protective instrument, in 1978 the components that would otherwise have been tariff on imported cars was increased again, to 57.5 sourced from producers in other countries were pro- percent. In 1982 an "export facilitation" scheme was duced in Australia at very high cost and were sold to introduced under which exports of cars or compo- the assemblers at correspondingly high prices. nents earned credits that enabled firms to reduce For at least 20 years, economists, as well as reports their local content below the otherwise mandatory by the Tariff Board and the Industry Assistance level of 85 percent and reduced the tariffs they paid Commission, had warned that the policies being on imported components. While this measure was followed toward the auto industry would have the intended to partially reverse the increasing isolation consequences that in fact became very apparent by of the Australian industry from the world auto the mid-1980s.1 In 1985 the government finally industry, because the industry's output was still began gradually winding down assistance to the protected by import licensing the scheme represent- industry--a policy that has since been consistently ed a further increase in the protection of the indus- followed. The measures have included: try's value added--that is, in effective protection. By 1985, after 20 years of the local content plans · The replacement of quantitative import controls and the tariff and nontariff measures that were on imported cars by tariff quotas, which were needed to make them viable, the auto industry had phased out in 1992. 180 Local Content Policies: Australia's Experience with Automobiles · Penalties (introduced in 1986) for low-volume librium model of the Australian economy, a simu- production by firms participating in the local lation in which auto tariffs were reduced from 27.5 content scheme. to 5 percent yielded estimated increases in real · The abolition, in 1989, of the local content GDP of between 0.4 and 1 percent. Since protec- scheme and the setting of import tariff rates on tion of the auto industry had been much higher original equipment components at the same level than 27.5 percent in the past, this exercise suggests as the tariff rates on cars. that the local content policy and other policies · Reduction of the tariff on imported cars and on applied to this industry had imposed even higher components from 57.5 to 45 percent in 1988, fol- costs on consumers and the economy for many lowed by a regular yearly reduction of 2.5 per- years. centage points in every subsequent year. These reductions stopped when the tariff reached 15 Lessons from the Australian Experience percent in 2000, but in principle there is to be a further reduction to 10 percent in 2005. A number of lessons of the costly Australian experi- · The introduction of retraining arrangements for ence with auto industry protection are worth noting. workers displaced by the reforms. 1. Restricting car imports and attempting to force As a consequence of these reforms, protection of self-sufficiency in production through local con- the industry is now much lower than in the past-- tent programs is likely to lead to a fragmented although it is still considerably higher than average market structure with a large number of models, protection for most other Australian industries, most of which are produced at low volumes. In which has also declined since 1985. In 1995 average 1980 in Australia the five major auto firms sub- effective protection of car production for domestic ject to the local content plan (GM, Ford, Mit- sale was estimated to be 31 percent, and average subishi, Nissan, and Toyota) produced 14 effective protection of auto component production different models, with annual sales of each model for the domestic market to be 55 percent. Reduced ranging from 1,392 to 68,204. On the fringes of protection levels had been made possible by lower the industry, three firms (Leyland, Renault, and costs associated with a decline in the number of Volvo) that were not subject to the local content basic car models, from 13 to 5. These models were plan assembled five models at very low volumes, produced by four assemblers operating four plants. about 1,700 per model. The remaining supply of By 1996, average model and plant volumes had autos was imported under a quota equivalent to about doubled from the 1985 figures, and consider- 20 percent of the market, but these imports were able rationalization and cost-cutting had occurred subject to a 57.5 percent tariff and were sold at in the component industry. Total car production prices that, on average, were approximately 85 was about 15 percent lower than in 1985, but the percent above the world price. Without excep- domestic industry's share of the domestic market tion, all the locally produced and assembled had declined from 77 to 55 percent. Meanwhile, models were produced at scales that were far exports of both cars and components had increased lower than the levels required to exhaust substantially and now account for significant shares economies of scale in the various processes, of domestic production. including, in particular, manufacture of engines, Australia's attempts to become self-sufficient in transmissions, and body panels. auto production were extremely expensive for con- 2. Especially at low volumes, as local content rises, sumers and for the economy as a whole. In 1995, the cost of the components that must be pro- after the local content plan had been abolished and duced locally to meet these requirements also tariffs had been reduced to 27.5 percent, the Indus- rises. In 1980 Toyota produced about 20,000 try Commission estimated that protection of the units of its Corolla model. In evidence presented industry was equivalent to a tax of about 3,700 to the Industries Assistance Commission, Toyota Australian dollars (US$2,960) on each car sold and explained how its estimated "duty needs" for 43,000 Australian dollars (US$34,400) for each protection against a comparable import person employed in the auto assembly and compo- increased as local content increased, as shown in nent sector. Using an economywide general equi- the table. 181 S E L E C T E D T R A D E P O L I C I E S A F F E C T I N G M E R C H A N D I S E T R A D E Local content, including assembly and marketing costs (percent) 37 60 70 75 80 85 Required import tariff (percent) 19 41 46 51 59 60 Toyota also presented evidence of the costs of Decisions on the nature of the local content producing various components in Australia, com- rules and the details of how they were applied to pared with production costs of the same compo- individual firms were crucial for the profitability nents in Japan. These data indicated a cost and survival of the auto firms, and efforts to disadvantage ranging from about 50 percent to influence these decisions became a major activity more than 500 percent. Mitsubishi and Ford also for their managers. In the interest of "fairness," presented evidence showing a wide range of cost the rules were adjusted to take account of the par- differences, although less marked than those indi- ticular situations of groups of firms or individual cated by Toyota. firms. As noted earlier, special reduced local con- The Australian experience thus clearly demon- tent programs were applied to low-volume pro- strates the cost of indiscriminate local content ducers between 1966 and 1975, which gave them rules that require local sourcing of components to considerable cost advantages by allowing them to meet the requirements regardless of their produc- import components that the larger producers tion cost. In the Australian case, these costs were were obliged to buy or to produce domestically. incorporated in the cost of the finished cars and These programs were replaced by provisions that were reflected in the selling prices of the cars, so allowed local content to be averaged across a that the higher the cost of a component to the car number of models by individual producers. By assembler, the more the component contributed to comparison with new entrants, this gave a deci- meeting the local content requirement. Depending sive advantage to incumbent producers with at on how "indigenization" percentages are defined, least one reasonably large-volume model, if they the same perverse incentive, whereby high-cost decided to introduce other, small-volume models. components satisfy indigenization requirements Later, to offset this effect, special low transitional more easily than do low-cost components, is likely local content arrangements were made for Toyota to be found in developing countries. and Nissan when they joined the local content 3. The local content policies, and the policies that sup- program and began production. ported them, were strongly countercompetitive. In 1981 the Industries Assistance Commission Owing to the local content requirements, many noted the "efforts to improve their competitive component suppliers had captive markets, and position by model rationalization and coopera- their market power was only limited by the poten- tive arrangements between producers" (Australia, tial ability of the assemblers to set up production Industries Assistance Commission, 1981: 125). in-house. For a number of years even this option Nevertheless, it observed that "manufacturers are was limited by a separate car component manufac- continuing to invest in separate and parallel pro- turing program. Under this scheme, components duction facilities in what is, by international stan- produced by independent manufacturers that met a dards, already a very fragmented industry." All minimum specified local content (usually 85 per- this was made possible by increases in tariffs to cent) were deemed to have 100 percent local con- accommodate rising industry costs, and for 10 tent when used by a vehicle producer under the years import quotas were applied to imported plan, even though the same component produced cars so as to ensure an 80 percent share of the in-house was subject to the general local content market to the local industry, regardless of its pro- requirement applicable to the producer.2 In this duction costs. The import quotas were extended and various other ways, the administering authori- to include completely and partly knocked down ty (the Federal Department of Business and Con- packs. At first, complex rules were established for sumer Affairs) contributed to the effective the allocation of quotas, based primarily on pre- cartelization of the domestic auto industry.3 For quota imports but with numerous exceptions. example, in the 1970s the department blocked Later, the quotas were auctioned, but the auction Honda from establishing an assembly plant that rules were extremely complex and became the would have operated outside the local content plan. focus of intense lobbying. 182 Local Content Policies: Australia's Experience with Automobiles 4. Contrary to the objectives of the original promot- Toyota, more than five times the world cost of the ers of these schemes, they retarded rather than components that could now be imported. This promoted technological change in the auto indus- made it worthwhile to export at prices that were try. A constant complaint of the local producers far below production costs. In 1981 the Industries was that the local content requirements made it Assistance Commission pointed out: too expensive for them to introduce new compo- nents and production techniques in their local There would be little rationale for export facil- operations. They consistently lobbied for lower itation without the inward-looking orientation local content ratios while advocating the same or and high marginal assistance associated with tighter limits on imports of finished cars. In par- high local content provisions. . . . Export facili- ticular, the local content programs seriously tation will increase government direction and retarded the introduction of smaller, fuel-efficient control of the industry and add to what is vehicles into the Australian market (Australia, already a complex and administratively costly Industries Assistance Commission, 1981: 125). assistance package. . . . There would be little The industry only began to catch up with the rest gain in predicating longer term restructuring of the world after the policy was reversed in 1985 of automotive production on the development and, especially, after the local content plans were of high cost exports which would require con- abandoned, in 1989. According to a submission to tinuing high subsidy. (Australia, Industries the Industries Commission by Toyota in 1996, Assistance Commission, 1981: 128)6 "The Government's car policy since 1984, by reducing protection, has required the car manu- 7. The transactions costs associated with schemes of facturers to progressively reduce the gaps in their this kind are likely to be extremely high, both for cost, quality and delivery performance" (quoted the government bodies involved in formulating in Australia, Industry Commission, 1997: 232). policy and administering the schemes and for the 5. Total employment in and associated with the auto auto and component producers and the many industry was reduced rather than increased as a other participants in the automobile market. result of the schemes and the policies that sup- Defining and administering local content require- ported them. In 1996 total employment in auto- ments involved the Department of Business and motive and auto component manufacturing was Consumer Affairs in a great deal of microman- 70,300, whereas total employment in car retailing, agement of the industry and was extremely time- repair, and the sale of auto replacement parts and and resource-intensive. For example, the depart- tires was 295,800, more then four times as great.4 ment's responsibilities included: The Industries Commission cited estimates of own-price elasticities of the demand for private · Examining the detailed cost and sales records cars in other countries that ranged from about of each producer. ­0.7 to ­1.6, and its staff gave a conservative esti- · Making frequent changes to accommodate mate for the Australian market of ­0.5. Even low-volume and new producers. using this lower estimate of the demand elasticity, · Setting and administering local content condi- it can be inferred that over time, the reduction in tions for component production. employment in car dealerships, repair shops, and · Adjusting local content requirements to allow the like as a consequence of the increased protec- for exchange rate appreciations that affected tion and higher car prices associated with the the various producers in different ways.7 local content plan would have far exceeded any · Drawing up and administering rules on "com- plausible estimate of increased employment in ponent reversion"--requests for producers to auto assembly and auto component production.5 switch from local sourcing to importing par- 6. The addition of "export facilitation" increased ticular components. rather than reduced the economic costs of the sys- · Exercising oversight of the prices charged by tem. Credits that firms received for exports component suppliers to assemblers to ensure allowed them to reduce their local content. The that "local content" was not artificially inflated. benefits to the firms of marginal reductions in · Monitoring the import prices of components local content were extremely high--according to to ensure that assemblers were not underin- 183 S E L E C T E D T R A D E P O L I C I E S A F F E C T I N G M E R C H A N D I S E T R A D E voicing imported inputs to help meet local of withdrawing from it that were politically content requirements. acceptable and administratively feasible. · Setting the rules for and administering the 8. Finally, once the local content programs became export facilitation policies adopted in 1979.8 established and major automobile producers, · Extending the general auto industry controls to component suppliers, trade unions, and other car "derivatives" such as panel vans and small groups came to rely on them, it became extremely buses using the same engines and other major difficult to remove them. As noted above, detailed components as cars and to four-wheel-drive critiques by economists and other experts were vehicles. These vehicles were initially not sub- already being made and were well known in the ject to local content and related policies, and 1960s, and by the early 1970s these criticisms were the import tariffs applied to them were lower consistently reflected in reports on the industry than car tariffs. As a result, their prices declined by the Industries Assistance Commission and in in relation to car prices, leading consumers to numerous press articles. Despite all this, the sys- choose them instead of cars. To limit this shift, tem had generated its own momentum, and it controls were widened to cover the derivatives. steadily became more protective and economical- ly costlier. In 1985 policy finally reversed course, Over the entire history of the local content and protection began to be wound down, but it plans, there were continuing conflicts between the took 13 years of sustained effort to reduce it to the auto producers, the component producers, the present much lower level, which, even in 2001, is trade unions, and many other groups with an still well above the levels found in nearly all other interest in the industry and the policies that major Australian industries. affected it. These disputes were heard and report- There are various explanations for the ability of ed on at length in numerous sessions of the Aus- the auto industry to sustain political support for tralian Tariff Board and its successor its special treatment over such a long period: organizations. Between 1965 and 1996 there were eight major hearings and reports, most of which · The populist appeal of high national content in took more than a year to complete and involved a well-known and visible consumer product. evidence presented by dozens of interested · The distinctly nontransparent nature of the groups. For example, work on the July 1981 protection resulting from the local content report of the Industries Assistance Commission programs. started in March 1979, and evidence for it was · The development of a strong vested interest in presented by 90 different parties--including the continuation of the system by the govern- assemblers, component producers, importers, ment officials responsible for its administration. auto distributors, raw material suppliers, trade · The determined lobbying of local businessmen unions, trade associations, professional associa- and the large international firms allied with tions, state governments, municipalities, and them, which entrenched themselves in the many others. Special ad hoc bodies were also set Australian market behind the protection of the up to provide advice on auto industry policies; local content programs. The international auto these included the Car Industry Council, estab- firms had ample resources that they used to lished in 1983, and the Automotive Industry influence the two principal Australian political Authority, created in 1985. parties at both the federal and state levels. Gen- In response to these continuous pressures, there eral Motors, which had a history of support was not a year after 1965 in which significant from local content programs going back to the changes were not made in the local content rules mid-1930s, had a key role in this regard. themselves or in the tariff and other policies that · The industry's unionized and influential work supported the system. Looking back, it is apparent force. that, over many years, a great deal of talent, intel- · The concentration of most of the investment lectual energy, and administrative and managerial and employment in the manufacturing side of resources--not least in the private sector--was the industry in a few places--in Melbourne wasted in first creating and building up this eco- and the nearby town of Geelong, and on the nomically costly edifice and then in devising ways outskirts of Adelaide. Even as late as 1996 the 184 Local Content Policies: Australia's Experience with Automobiles Victorian and South Australian state govern- Notes ments, the municipalities in which the major This chapter is based on Pursell (2001). auto plants are located, and politicians from these places were still lobbying strongly against 1 The Tariff Board mutated over time. In 1973 it became the Industries Assistance Commission, in 1989 the Industry further reductions in protection. By contrast, Commission, and in 1998 the Productivity Commission. the consumer and general national interest in The changing names reflect significant changes in emphasis lower car prices and in a more efficient indus- and coverage, from tariffs and industrial protection in the try was diffuse and difficult to mobilize.9 beginning to, eventually, policies for enhancing productivi- ty and efficiency across all industries (including services All of these political-economy reasons for the industries). staying power of the structures created by the local 2 This rule deterred in-house production, since any imported content plans and their supporting policies are like- parts for the component would reduce the vehicle producer's ly to be important in other countries that start local content as defined in the plan. along similar paths to those traversed in Australia. 3 The Department of Business and Consumer Affairs was respon- sible for the administration of the local content plans during the 1970s and 1980s. The Department of Industry and Com- Some Implications for Developing merce was responsible for policy aspects. Countries 4 In 1979 employment in auto and auto component manufac- The local content requirements and other aspects of turing was 62,368, while employment in motor dealerships the new auto policies currently being implemented and tire retailing alone was 165,700. in a number of developing countries have obvious 5 In 1979, when domestic car prices were about 85 percent similarities with Australia's past policies. The mar- above world prices, a cut in protection to 20 percent would ket for cars in many countries will be too small to have been equivalent to a reduction in car prices of about 35 allow economies of scale to be captured. For exam- percent and, assuming a demand elasticity of ­0.5, an increase ple, India's car market is currently about the same in annual final car demand of 17.5 percent. Over a moderate size as that of Australia; 411,000 cars were produced time span--say, five years--the employment effects from sell- ing more new cars and servicing a larger total stock of cars in India in 1996­97, while Australia produced quickly begin to exceed plausible reductions in employment in 490,000 in 1996. The potential for costly fragmenta- auto and component production. tion is great, with more than 20 domestically pro- duced models being planned in India, versus 5 6 Despite the opposition of the Industries Assistance Commis- sion, "export facilitation" was continued by the government models in Australia. More generally, local content and is one element of the early policy package that still exists. regimes involve considerable potential for detailed, However, it lost much of its impact after the local content plan complex, ad hoc, and nontransparent government was abolished and, subsequently, as tariffs were reduced. intervention in the industry. 7 In particular, Japanese producers requested and received spe- The expensive and economically damaging record cial treatment to offset the reduction in their local content of local content policies for automobiles in Australia ratios that resulted from appreciation of the yen. suggests that the present efforts of a number of developing countries at the WTO to legitimize or 8 The export facilitation rules were extremely complex and were changed frequently. For example, initially, local content credits indefinitely extend trade-related investment meas- were earned on the basis of the gross value of exports, but this ures (TRIMs) are not in their own economic inter- was soon changed to net foreign exchange earnings from est. But the Australian experience also shows the exports after deducting the cost of imported components. The strength of the nationalist and protectionist instincts concept was further refined to deal with local components into which these policies play, as well as the tenacity that themselves, at first, second, or even further remove, used imported components or materials. of the interests that are created by the policies and that would oppose their removal. As became appar- 9 After losing out for 13 years, the forces supporting special ent in Australia, because of its transparency, relative treatment of the auto industry had a victory in 1997. In a simplicity, and relative freedom from lobbying and report that year the Industry Commission recommended that the tariff reductions should continue until the tariff reached 5 administrative discretion, tariff-based protection is percent in 2005, but the government rejected this recommen- preferable to TRIMs and would be much more con- dation. Instead, it decided that the reductions would stop in ducive to the economically efficient development of 2000 at 15 percent, with a provision that there would be a fur- this important industry in developing countries. ther reduction to 10 percent in 2005. 185 21 H A N A A K H E I R - E L - D I N Implementing the also tend to increase with the stage of processing. Thus, the average tariff on fibers in indus- Agreement on trial countries is about 1 per- cent, but tariffs on clothing Textiles and often exceed 20 percent, thus enhancing the effective protec- Clothing tion to higher value added prod- ucts in these countries. In the Uruguay Round devel- oping countries managed to negotiate a compromise agree- ment to integrate and liberalize T trade in textiles and clothing extiles and clothing are important over a period of 10 years, beginning on January 1, industries in developing countries, 1995. The Agreement on Textiles and Clothing and they contribute significantly to manufacturing (ATC) is the transitional agreement that regulates production, employment, and trade in such trade in textiles over the 10-year MFA phaseout economies as China, Hong Kong (China), India, the period. Both importing industrial countries and a Republic of Korea, and Pakistan. Traditionally, large number of developing country exporters were developing countries have protected these indus- in favor of this transition period, to allow domestic tries through tariffs and quantitative restrictions. industries to prepare for the expected increased Until the Uruguay Round, this domestic protection competition resulting from freeing of trade in tex- was somewhat justified by the protection accorded tiles. This chapter reviews the main elements of the to textile and clothing industries in industrial coun- ATC, assesses its implementation to date by indus- tries. Through a set of bilaterally negotiated agree- trial countries, considers the efficacy of the WTO ments under the Multifibre Arrangement (MFA), dispute settlement mechanism as an enforcement industrial countries--principally Canada, the Euro- device, and discusses the principal concerns of pean Union (EU), Norway, and the United States-- developing countries regarding implementation of applied widespread and restrictive quotas against the ATC. imports from developing countries. This violated the fundamental GATT principle of nondiscrimina- Main Elements of the Agreement on tion and the injunction against the use of quantita- tive restrictions.1 Textiles and Clothing In addition, imports of textiles and clothing are The Agreement on Textiles and Clothing is a multi- restricted by tariffs that in industrial countries are, lateral trade agreement under the WTO, but it has on average, more than double those on other manu- distinctive features that differentiate it from other factures (15 percent, as against 6 percent). Tariffs WTO agreements, as well as from the MFA. The 186 Implementing the Agreement on Textiles and Clothing MFA involved separate agreements between certain step removal of existing quotas (described by the GATT contracting parties to waive their GATT agreement as "integration") and through accelerated rights and obligations by applying quantitative expansion of the remaining nonintegrated quotas restrictions selectively against specified countries, ("liberalization"). Integration involves two groups of thus violating the nondiscrimination principle. The countries: those that maintained quotas under the ATC is an integral part of the multilateral trading MFA (principally, the United States, the EU, Canada, system, administered and supervised by the WTO. It and Norway) and other WTO members that have applies equally to all WTO members and is binding chosen to retain the right to use the special safe- on all of them. Another difference is that the MFA guards provision of Article VI of the ATC. contained a provision for the accession of non- Integration is to be carried out in three stages. In GATT members such as China. In contrast, acces- the first stage, which began on January 1, 1995, sion to the WTO automatically implies membership WTO members must integrate 16 percent of the in the ATC, and non-WTO members are not cov- total volume of their 1990 imports. In the second ered by ATC provisions. stage, which started on January 1, 1998, an addi- A distinctive feature of the ATC is its fixed time tional 17 percent of the total volume of 1990 span of 10 years, which cannot be extended. The imports must be integrated, followed by another 18 ATC is therefore a transitional agreement. The percent in the third stage, which commences on agreement's main elements are its product coverage, January 1, 2002. Finally, on January 1, 2005, the the program of liberalization, the treatment of remaining 49 percent of the total volume of 1990 existing trade restrictions, the application of transi- imports must be integrated. The choice of products tional safeguards, the fulfillment of commitments to be integrated is left to the importing country but under GATT rules, the supervision of ATC imple- must include at least one item from each of four mentation, and dispute settlement. major product groups: yarns and tops, fabrics, made-ups, and clothing. Two comments are in order. First, a more reason- Product Coverage able proposition would have been to choose a refer- The product coverage of the ATC is specified in an ence year as close as possible to the ATC's initiation annex to the agreement in terms of the six-digit level in 1995 (depending on the availability of import of the Harmonized Commodity Description and data), rather than 1990. Second, leaving nearly one- Coding System (HS). The ATC covers all of Section half of all imports to be integrated at the end of the 11 of the HS, with the exception of raw fibers. It also transition period does not ensure a smooth and covers certain lines from other chapters of the HS painless process of integration and thereby contra- that embody textile materials, such as luggage, dicts one of the purposes of a transition period. umbrellas, watch straps, and parachutes. The ATC's Some avenues of flexibility in exceeding quota product coverage is wider than that of the MFA in limits that applied under the MFA were carried over that it includes products of pure silk as well as those to the ATC. These include transfer of 6 percent of made from vegetable fibers. The ATC Annex the unfilled quota volume from the previous year to includes both HS lines that were restrained under the current year (carryover), prior utilization of 6 the MFA and lines that were unrestrained.According percent of next year's quota (carryforward) and to estimates by the International Textiles and Cloth- transfer of quotas from one product to the other ing Bureau (ITCB), imports of HS lines that were within the limit of 6 percent of the quota requested not restrained under the MFA accounted for 33.6 to be increased (swing). These flexibility advantages percent of total imports in 1990. The proportion of are usually transferred to quota beneficiaries in HS lines not covered by quantitative restrictions in cases of tight quotas. the United States was 36.8 percent; in Canada and Concurrently with the process of integration, Norway the share was considerably higher. products remaining under restriction are allowed an increase in growth rates above those agreed on under the MFA. Quotas for such products are to be Liberalization Program increased by an additional 16 percent in the first The gradual elimination of quantitative restrictions stage, by 25 percent in the second stage, and by 27 under the ATC is to be achieved through the step-by- percent in the third. Small suppliers whose restric- 187 S E L E C T E D T R A D E P O L I C I E S A F F E C T I N G M E R C H A N D I S E T R A D E tions represented 1.2 percent of total restrictions as Transitional Safeguards of December 31, 1991, are to be accorded an even higher additional growth rate: 25, 27, and 27 per- The ATC permits the introduction of new selective cent in the three stages, successively. Least-devel- restrictions during the transition period by means oped countries are eligible to receive this treatment, of a temporary safeguard mechanism (Article VI). as well. They cannot, however, claim it as a right, This right is available to all WTO members. Mem- since application of the higher growth rates is sub- bers that have not applied MFA restrictions in the ject to the qualification "to the extent possible." past are required to give notification as to whether This process of increasing the negotiated growth or not they wish to use this right. The transitional rates, sometimes called the "growth-on-growth" safeguard can only be applied to products that have provision, is a significant liberalization element in not been integrated into the GATT. Once integrat- the ATC. ed, safeguard actions can be taken under Article For certain groups of countries, the ATC provides XIX of the GATT. for more favorable treatment in fixing quota levels, The ATC stipulates that transitional safeguards growth rates, and flexibility provisions than the should be used as sparingly as possible. They can be minimum requirements set out above. In addition invoked only in certain situations: the invoking to least-developed countries and small suppliers, country must determine that an increase in total these beneficiaries include wool-producing coun- imports of the product has caused serious damage or tries that are dependent on the export of wool prod- a threat of actual damage, and the causation of this ucts and countries that are highly dependent on the damage or threat of damage has to be attributed to a export of processed imported products. specified country on the basis of a sharp and substan- tial increase in imports from that country. Changes in relevant economic factors such as output, produc- Treatment of Existing Restrictions tivity, capacity utilization, employment, market Article II of the ATC requires that WTO members share, and the like have to be taken into account. give notification of all bilateral restrictions existing Use of the ATC safeguard mechanism has been as of December 31, 1994, the day before the entry restrained in comparison with experience under the into force of the ATC. Notifications are to include MFA. The ATC and MFA mechanisms differ in details on the base level, growth rates, and flexibility three respects. First, the MFA concept of market dis- provisions that apply to each quota. These restric- ruption was not confined to serious damage to the tions are subject to the ATC until the products are domestic industry, as is the case with the ATC, but integrated. The notified quotas constitute the totali- was also invoked to prevent risks of market disrup- ty of the restrictions; all quotas that are not notified tion. Second, under the MFA the finding of market must be terminated. Article VI of the ATC, however, disruption could be limited to consideration of allows for the imposition of transitional safeguards. increased imports from a particular source, while Moreover, textile products may also be subjected to the ATC requires total imports to be taken into regular antidumping and safeguard actions once account in determining the existence or threat of products have been integrated. serious damage. Third, under the MFA the finding The ATC requires that only the MFA restrictions of market disruption was based on a sharp and sub- contained in the bilateral agreements be notified stantial rise in imports at a significantly low price. and carried over into the transition period. Unilat- The ATC has disregarded the price factor. eral actions, which were not covered by the MFA, Having determined the existence or threat of seri- were to be removed within one year after January 1, ous damage and the exporting country responsible 1995. All WTO member countries were required to for it, the invoking country must consult with the notify any non-MFA restrictions, whether consis- exporting country. The consultations may lead to tent with GATT or not, to the Textiles Monitoring agreement on the imposition of restraints. The mat- Body (TMB). All such restrictions not justified ter has to be reported to the TMB, in any case, and if under GATT were to be either brought into con- there is no agreement, the matter is to be referred to formity with the GATT within one year or phased the TMB. Restraints can remain in place for a maxi- out over the transition period according to a pro- mum period of three years, or until the product is gram presented to the TMB. integrated into GATT, whichever occurs first. 188 Implementing the Agreement on Textiles and Clothing The quota level under the ATC should not be less The list of items notified by the EU and the Unit- than the actual level of trade reached in the first 12 ed States to the TMB indicates that, up to the end of months of the 14-month period preceding the 2001, integrated products were either of little month when the request for consultations was importance to the major importers or had not orig- made. Should the quota remain in place for more inally been restrained by quotas. The same observa- than a year, it must be increased annually by at least tion applies to the integration program notified by 6 percent. The ATC also provides for flexibility ele- the United States for the third stage. As Table 21.2 ments in these quotas. shows, of 750 U.S. quotas, only 2 were removed in stages 1 and 2, and 11 were removed on a preferen- tial basis for Romania alone. For the EU, which has a Assessment of ATC Implementation total of 219 quotas, 14 were eliminated by integra- Under the MFA many industrial countries, includ- tion in stages 1 and 2, and no early elimination has ing the United States, the EU, Canada, Austria, Swe- been reported. Canada eliminated 29 of 295 quotas den, Finland, and Norway, restricted imports of during the first two phases. Norway, by contrast, has textiles and clothing from developing countries. eliminated all quantitative restrictions except for (Japan and Switzerland, important industrial three quotas on fishing nets from Indonesia, importers, never imposed such quotas.) As of 1995, Malaysia, and Thailand. only the United States, the EU, Canada, and Norway Much criticism has been voiced regarding the continued to use quotas to restrict their imports of phaseout programs of the United States, the EU, textiles, but these economies represented almost and Canada, especially by developing country one-half of the global market. In 1997 total trade in exporters of textiles and clothing. Since quotas have textiles and clothing stood at about US$332 billion; varying significance for trade, it is useful to look at U.S. imports amounted to US$63 billion, the EU's to the integration programs from the perspective of US$65 billion, Japan's to US$23 billion, and Cana- the extent of restrained trade that has been freed of da's to about US$6 billion. The EU and the United quota restrictions as a result of integration. As States are the largest markets for textiles and cloth- Table 21.3 makes clear, the products that have been ing and the two main users of the quota system. freed of quota restrictions represent only small The ATC calls for multilateral review of progress shares of the total restrained imports of the two achieved, before the end of each stage, to assess the largest WTO members--about 6 percent for the implementation of the integration and liberaliza- United States and less than 5 percent for the EU. tion processes. Review of implementation to date Although 33 percent of trade has been integrated to shows that little real progress has been achieved. fulfill the minimum legal ATC requirement, the Table 21.1 summarizes the integration programs for process has contributed little toward the realization the combined stages 1 and 2 of WTO members that of the main objectives of the ATC--the progressive maintained restrictions under the MFA. These pro- phasing out of quotas and the liberalization of grams cover a period of seven years, or 70 percent of trade. On the positive side, Canada has eliminated the ATC transition period. It is obvious that the many (but not all) restrictions on sensitive prod- products selected for integration have been concen- ucts such as tailored-collar shirts and children's trated in lower value added items such as tops, blouses and shirts. Norway has eliminated all its yarns, and fabrics. Clothing products represent only quotas except for the three on fishing nets men- a small share of the total. tioned above. Table 21.1 Results of Integration Programs for ATC Stages 1 and 2 (integration as percentage of volume of 1990 imports) WTO member Yarns and tops Fabrics Made-ups Clothing Total United States 16.46 4.15 8.73 3.90 33.24 EU 16.04 9.47 5.27 2.53 33.31 Canada 10.26 6.43 15.50 2.78 34.97 Norway 10.10 14.26 14.66 4.32 43.34 Source: TMB notification as reported in ITCB, IC/W/2I9, July 21, 2000. 189 S E L E C T E D T R A D E P O L I C I E S A F F E C T I N G M E R C H A N D I S E T R A D E Table 21.2 Number of Quotas Eliminated by Integration in ATC Stages 1 and 2 Total number Number of quotas eliminated WTO member of quotas By integration By early elimination Total United States 750 2 11a 13 EU 219 14 0 14 Canada 295 29 0 29 Norway 54 0 51 51 a. Quotas eliminated only in respect to Romania. Source: TMB notifications, reported by ITCB, IC/W/219, July 21, 2000. Table 21.3 Restrained Trade Freed of Quotas, 1995­97 (percentage of imports) WTO member 1995 1996 1997 United States By volume 6.23 6.03 6.00 By value 6.40 6.14 6.12 EU By volume 4.74 4.92 4.77 By value 4.28 4.34 4.18 Source: ITCB, WT/GC/W/283, 2000. As mentioned, the ATC requires that restraining example, under the integration process proposed countries increase the existing growth rates of quo- by the EU, most quotas would be left in place until tas by no less than 16, 25, and 27 percent successive- close to the end of the 10-year transition period. ly over the specified three stages. These rates may Only 52 of the 219 EU quotas (less than 24 per- appear impressive, but in reality they did not yield cent) would be dismantled during the three substantial added access because the base growth stages--a very modest phaseout indeed! To begin rates allowed under the MFA were modest. Table with, in 1990 only 58.3 percent of imports was 21.4 reports the additional growth in access for the actually under quota restrictions. This allowed the first two stages (seven years) of the transition period. EU to avoid integrating restrained products that With the implementation of the third stage of were of any significance. Furthermore, although the ATC, neither the integration proposals nor the under its proposal the EU would have integrated proposed additional growth factors significantly 51 percent of the base year imports, this would improve access or liberalize trade in textiles. For represent only 12.3 percent of 1995 total imports Table 21.4 Expanded Market Access Attributable to Increases in Quotas (Stages 1 and 2 Combined) (percentage of imports) Annual average Average pre-ATC Total increase increase in WTO member growth rate in access access, 1995­2001 United States 4.61 6.36 1.03 EU 3.44 4.49 0.73 Canada 5.26 7.53 1.22 Source: ITCB, document WT/GC/W/283, 2000. 190 Implementing the Agreement on Textiles and Clothing and 21 percent of overall restrained imports. The tions do not result in agreement on satisfactory bulk of the restrained products (about 79 percent) solutions, the problem may be referred to the TMB. will have to be liberalized at the end of the transi- The ATC provides for legal action against the tion period. Another indication of the constrain- exporter or importer under domestic laws in case of ing effect of the remaining restrictions is the fact false declarations for purposes of circumvention. that whereas total EU imports increased by 31 per- Actions taken in fulfilling commitments under cent between 1995 and 1999, imports from WTO the ATC have to be notified to the TMB. (A summa- members under quota restrictions expanded by ry is sufficient if these actions have been notified to only 20 percent.2 other WTO bodies.) If actions have not been taken to fulfill specific commitments, the TMB has to be informed. These commitments are related to the The Textiles Monitoring Body achievement of improved market access for textile The Textiles Monitoring Body (TMB) has the man- products and are entered into to ensure the applica- date of supervising the implementation of the ATC. tion of policies in the areas of dumping, subsidies, It has a dual function: it examines the conformity of and piracy of trademarks and designs, with a view all actions taken with the ATC, and it addresses dis- to establishing fair and equitable trading conditions putes among members. The ATC requires that and avoiding discrimination against textile imports changes in rules and procedures, or changes in the (ATC Article VII.1). category system that affect the implementation or administration of quotas, not upset the balance of Developing Country Concerns rights and obligations between members, adversely affect or impede the access of a member, or disrupt Implementation of the ATC to date indicates that trade under the ATC. Should such changes occur, little actual liberalization is to be expected before the parties concerned are to consult. If consulta- the end of the 10-year transition period. Experience tions fail to reach a mutually satisfactory solution, also shows that other policy measures can be easily the matter is to be referred to the TMB for decision. employed to limit the impact of liberalization. The TMB thus acquires the function of resolving Thus, safeguards and antidumping measures have disputes among the parties. been used to restrict exports to both the EU and the The TMB is a standing body made up of a chair- United States. Some believe that quotas may be a man and 10 members drawn from countries that better alternative (see Ozdem and Demirkol 1994), are broadly representative of the WTO. It is almost given that industrial countries, especially in the EU, equally divided between industrial and developing actively use contingent protection to protect their countries. With the exception of four permanent industries. Rules of origin can also be used to industrial country representatives, membership restrict trade. Changes in U.S. rules of origin for rotates. This asymmetric treatment favors the textile and clothing products, implemented as of industrial country group. The TMB relies principal- July 1, 1996, had adverse effects on sales to countries ly on notifications by WTO members. Its decisions that produced for export to the American market are taken by consensus, and it has therefore not and created an incentive for such exporters to always been able to resolve disputes. In such cases obtain their materials from countries not subject to members may invoke general WTO dispute settle- quota restraints (ITCB 1999). ment mechanisms. The implications of the ATC for developing coun- ATC Article V.1 requires that members establish tries depend principally on the relative importance the necessary legal provisions and administrative of restricted markets for their exports, on the signif- procedures to prevent circumvention by transship- icance of textiles and clothing in their external ment, rerouting, false declarations, and falsification trade, on future trends in competitiveness, and on of documents. When the existence of such circum- their own policies in this sector. Industrial country vention is established through investigation and members of the WTO should complement their cooperation among parties, members may agree to commitment to phase out quantitative restric- deny the entry of goods, to charge the goods to the tions--whether imposed under the MFA or other- quota of the true country of origin, or to introduce wise--with reductions in most-favored-nation restraints on the country of transit. If the consulta- (MFN) tariffs on textiles and clothing. 191 S E L E C T E D T R A D E P O L I C I E S A F F E C T I N G M E R C H A N D I S E T R A D E Removal of yarn quotas in the EU and the United More generally, prospects for growth in the area States under the ATC will open up international of garment-making may be quite promising if markets and will expose developing country exports countries benefiting from proximity to the EU or to increased competition from countries with effi- the United States focus on higher-quality niche cient yarn industries and large export capacities that products. "Since clothing products are becoming have fully utilized their quotas. In the EU, for exam- almost as perishable as fruits or vegetables, with ple, India and Pakistan can be expected to challenge new fashions being introduced every few weeks or yarn exports by small and less efficient producers months, being close to final demand is an advantage from the Mediterranean region, Brazil, and Korea. that needs to be exploited" (Dean Spinanger, Kiel Other potential competitors are not likely to present Institute of World Economics, personal communi- a serious challenge in the EU, as they were far from cation, 2000). Opportunities for foreign and local filling their quotas in 1994­96.3 Keen competition in private investment could be increased through fabrics is likely from Malaysia and Thailand, which establishment of export-processing zones that offer exceeded their quotas in the EU. Other competitors modern industrial infrastructure, duty-free treat- from Asia, South America, Russia, and Central and ment of imports, access to new technologies, and Eastern Europe (CEE) have not filled their respective technical and vocational training for regional work- quotas to the EU and, other things being equal, are ers. Partnership agreements with the EU or with not likely to threaten the export performance of members of the North American Free Trade Agree- other countries.4 Elimination of quotas--which ment (NAFTA) can enhance such activities by pro- provide guaranteed access to the market for small viding better access to international fashion designs, and less efficient producers­will open the market to production techniques, accessories, patterns, dyes, more efficient, relatively large suppliers that have and finishes and to marketing and advertising ser- exceeded their quotas or are close to filling them. vices. This could have a significant effect on An important export opportunity available to improving competitiveness by facilitating the cre- countries that already have partnership agreements ation of distinct and differentiated products for with the EU or the United States is the potential niche markets both within regional markets and in increase in outsourcing. Clothing produced in these the rest of the world. countries with EU or U.S. fabrics will enjoy free Further reduction in MFN tariffs on textiles and access to EU and U.S. markets, given prevailing clothing under WTO auspices will benefit develop- rules of origin. European or U.S. investors may ing countries in markets where they do not enjoy increasingly engage in subcontracting activities in preferential treatment and where they were the regional partner economies by creating new restrained by quotas under the MFA. Substantial productive units and supplying existing ones with export opportunities exist for efficient suppliers the fabrics, accessories, designs, and know-how to that are subject to high MFN tariffs ("peaks"). produce high value added products to be exported Export opportunities will also expand after the to their home markets. complete elimination of quantitative restrictions It appears that, after the conclusion of partner- under the ATC. Beneficial effects may be important ship agreements with the CEE countries, total out- for items that face binding quotas in the United sourcing increased significantly, to about 18 percent States but are likely to be minimal for exporters that of total CEE exports to the EU in 1993, up from 10 were not subject to the MFA or were not facing percent in 1989. For garments alone, such activities binding quotas in their export markets. accounted for around 74.5 percent of CEE exports The extent to which exporters of textiles and to the EU.5 (By contrast, the shares were 12.2 per- clothing can effectively benefit from the opportuni- cent in Morocco and 16.5 percent in Tunisia in the ties created by the complete implementation of the same year; see World Bank 1995.) Although these ATC will depend on their ability to improve their activities may improve the efficiency of domestic relative competitiveness over the transition period. textile industries and promote exports, the Moroc- Factors such as labor, transport, capital, and trans- can experience suggests that they may lead to dual- actions costs, as well as the real exchange rate, will ism in the economy by promoting the installation of be significant determinants in this respect. production units that are not integrated with the Finally, although the MFA did not directly restrict rest of the domestic economy. trade in fibers, its phaseout may be expected to have 192 Implementing the Agreement on Textiles and Clothing a favorable impact on fiber production by increas- have used dispute settlement to contest measures ing the long-term demand for, and thus the price of, restricting trade in textile products that violated the textile fibers. Two distinct effects are likely: an out- ATC, including transitional safeguards. put effect arising from increases in the volume of The EU and the United States have fulfilled only textile and clothing output, and hence fiber input, the minimum legal requirements of the ATC, offer- and a substitution effect resulting from the elimina- ing very little in terms of progressive phasing out of tion of distorted incentives, created under the MFA, quotas and liberalization of trade in areas that are of to use certain types of fibers. For cotton producers, importance for market access by developing coun- the substitution effect may be relatively large, as it tries. The strategy has been one of postponing inte- has been reported that the MFA resulted in an gration and liberalization of key products until the implicit tax of around 20 percent on cotton relative very end of the transition period. This raises con- to man-made fiber products (Martin 1996). cern about the feasibility of implementing the inte- gration commitments. It is worth noting that tariffs on clothing will remain significantly higher than the Conclusion manufacturing average for most countries and that The ATC is designed to strengthen the multilateral domestic producers of garments and textile prod- trading system by integrating textiles and clothing ucts in industrial and emerging market economies into the WTO. In principle, it cannot be renewed or may use antidumping and safeguard actions more extended after the 10-year transition period has frequently in the future. The abolition of the MFA expired. At that time, WTO-inconsistent restraints will increase the incentive to use antidumping as an will no longer exist, and protective measures must instrument for restricting imports. By slowing be taken in accordance with WTO provisions. adjustment, the back-loaded implementation strat- The ATC embodies a comprehensive framework egy of the major restricting countries may make it of procedures for strengthening its implementation. difficult to implement the ATC on schedule. Anoth- The TMB ensures strict compliance with ATC pre- er important point is the great potential gains from scriptions, and compliance is further reinforced by liberalization for consumers of textiles and clothing the applicability of the WTO Dispute Settlement in major markets such as the United States and the Understanding to this sector. Costa Rica, India, and EU (see Box 21.1). Both considerations mean that Thailand are among the developing countries that there is a very good case for accelerating integration. B O X 2 1 . 1 T H E C O S T O F T R A D E P R O T E C T I O N O F T E X T I L E S A N D C L O T H I N G I N T H E E U R O P E A N U N I O N The textile and clothing industries in EU countries trade in textiles and clothing and examine the (aside from some Southern Rim members) are impact of an accelerated implementation of EU now of relatively minor importance in terms of obligations under the ATC. Looking only at direct both production and exports. Overall, these effects (partial equilibrium analysis), they calcu- industries account for slightly more than 5 per- late that in 1997 EU consumers paid roughly ECU cent of exports and less than 5 percent of manu- 12 billion more for textile and clothing products facturing value added. One might therefore than they would without the quotas and tariffs. expect protection of textile and clothing markets Use of general equilibrium analysis yields very in the EU to be a minor issue, but this is not the similar costs to EU consumers from higher textile case. The EU, like the United States, has back- and clothing prices (for both imported and loaded the implementation of the Agreement on domestic goods): ECU 12.7 billion. Textiles and Clothing (ATC). If the implementation of the ATC had been Francois, Glisman, and Spinanger (2000), using accelerated, so that integration would have been both partial equilibrium and general equilibrium complete by December 31, 1997, EU economies analysis, estimate the costs of EU restraints on would have gained over ECU 25 billion per year (continued) 193 S E L E C T E D T R A D E P O L I C I E S A F F E C T I N G M E R C H A N D I S E T R A D E B O X 2 1 . 1 ( C O N T I N U E D ) (measured in 1997 ECU, and including indirect As for the distribution of the costs of protection efficiency effects). Almost ECU 6.5 billion of these across population groups, the quota prices for annual gains simply follow from a recapture of children's clothing are noticeably higher than for ATC quota rents. Other gains stem from adult clothing, even though adult clothing typi- increased investment and reallocation of cally carries higher retail prices. With quota prices resources to areas with higher returns. The sum for children's clothing in 1997 about 200 percent of the yearly discounted net income gains from a higher than for comparable adult clothing, the full 1997 implementation is more than ECU 160 magnified impact on families with children is billion. obvious. Recalculating these results (ECU 12.7 billion for The conclusion to be drawn is that liberaliza- consumers, plus ECU 12.3 billion from the gain in tion under the ATC should be effected as quickly efficiency and other factors, for a total of ECU 25 as possible and should be comprehensive. Delay- billion) and applying them to an EU family of ing such liberalization to the very end of the ATC four, we find that the average gain resulting from transition period (January 1, 2005) would accelerated implementation (doing away with impose continuing significant direct costs on quotas and tariffs) amounts to ECU 270 per year consumers and substantial income losses for the per household. Moreover, these figures imply economy as a whole. The resources lost could that the annual cost of each job saved in the tex- have been invested in production potential-- tile and clothing industries by delayed implemen- human or physical capital--in sectors where the tation is, on average, ECU 28,000 for the textile EU has clear comparative advantages. Thus, industry and ECU 41,000 for the clothing indus- quite apart from the benefits that would accrue try. Since the industry as a whole is contracting, to efficient producers of clothing in developing Francois, Glisman, and Spinanger (2000) argue countries, liberalization is in the interest of that the cost of EU protection for the textile and importing countries. clothing industries could well approach the full value added of these industries by the time quo- Source: Prepared by the volume editors, based on Fran- tas are removed in 2005. cois, Glisman, and Spinanger (2000). Notes The percentage of quota utilization was 6 percent in Brazil, 51 percent in Peru, 56 percent in Thailand, and 77 percent in 1 On the export side, the last incarnation of the MFA (MFA IV) Korea (see Clément and others 1996). covered Argentina, Bangladesh, Brazil, China, Colombia, 4 The rates of quota utilization for the main exporters of fabrics Costa Rica, the Czech Republic, the Dominican Republic, the to the EU for 1994­96 were as follows: Argentina, 34 percent; Arab Republic of Egypt, El Salvador, Fiji, Guatemala, Hon- Brazil, 28 percent; Bulgaria, 94 percent; Czech Republic, 90 duras, Hong Kong (China), Hungary, India, Indonesia, percent; Egypt, 74 percent; Hong Kong (China), 16 percent; Jamaica, Kenya, the Republic of Korea, Macao (China), Hungary, 37 percent; India, 93 percent; Indonesia, 80 percent; Malaysia, Mexico, Oman, Pakistan, Panama, Peru, the Philip- Korea, 46 percent; Malaysia, 101 percent; Pakistan, 98 per- pines, Poland, Romania, Singapore, the Slovak Republic, cent; Peru, 24 percent; Poland, 28 percent; Romania, 34 per- Slovenia, Sri Lanka, Thailand, Turkey, and Uruguay. cent; Singapore, 5 percent; Slovak Republic, 44 percent; 2 EU Integration Plan for ATC Stage 3, ITCB, IC/W/219, July 21, Thailand, 108 percent; and Turkey, 71 percent (see Clément 2000. and others 1996). 3 India filled 107 percent of its yarn quota to the EU in 1994­96, 5 It should be noted that the main textile and clothing exports while Pakistan and Indonesia filled 150 and 130 percent of of the CEE countries are wool and man-made fiber products, their respective quotas during the same period. Argentina, while the bulk of exports from most Mediterranean countries however, covered only 33 percent of its quota in 1994­95. consists of cotton manufactures. 194 22 J . M I C H A E L F I N G E R Safeguards more to the national economic interest than they take away-- from ordinary protection. The prescriptive part of the Making Sense of chapter draws lessons from the GATT/WTO Provisions GATT/WTO experience--lessons that a government might take into Allowing for Import Restrictions account when deliberating the structure of a safeguard mecha- nism that would help it manage pressures from particular indus- tries for "exceptional" import pro- T tection as the government rade liberalization is not rocket sci- implements a liberalization program or works to main- ence. Any program that significantly tain a policy of openness to international competition. opens the domestic market to international competi- There is more to designing a sensible safeguard tion will require a degree of fine-tuning. Any govern- mechanism than simply finding what trade restric- ment that maintains a liberal trade policy will be tions GATT/WTO rules allow. In this regard, these subject to occasional pressures for exceptional treat- rules are too generous; imposing all the restrictions ment--for example, temporary protection for a par- that are permitted would isolate an economy from the ticular industry. Thus, part of the politics of global system. The challenge a government faces is to safeguarding a generally liberal trade policy is to have identify, among the many processes the GATT/WTO in place a policy mechanism for managing such pres- allows, a safeguard system that makes economic and sures; that is, for considering petitions for protection political sense--one that distinguishes between that are exceptions to the general thrust of policy. restrictions that will and will not advance the national This chapter is about such policies. The analytical economic interest and one whose political dimen- part reviews the use of GATT/WTO rules that spec- sions will help support the government's concern to ify how and when a member country may introduce integrate its economy into the global system. a new trade restriction or replace an old one. (I do not distinguish one instance from the other.) GATT Experience with Safeguard Provisions I draw two major lessons from this analysis: Although the GATT is perhaps best known as the 1. GATT/WTO rules are fungible. At different times, patron of agreements to remove trade restrictions, it members have used different instruments to han- includes a number of provisions that allow coun- dle safeguard issues. tries to impose new ones. Twenty of them are listed 2. GATT/WTO rules do not distinguish economi- in Table 22.1, and the list could be longer. Article cally sensible trade restrictions--those that add XX, for example, includes 10 subcategories. The 195 S E L E C T E D T R A D E P O L I C I E S A F F E C T I N G M E R C H A N D I S E T R A D E Table 22.1 Frequency of Use of GATT Provisions That Allow Trade Restrictions, 1948­94 Instrument Frequency of use 1. Provisions for renegotiating previous concessions and commitments Periodic (three-year) renegotiations at the January 1955­March 1994: 206 renegotiation initiative of the country wanting to increase procedures, 128 of these under Art. XXVIII.5. a bound rate, Arts. XXVII.1 and XXVIII.5 Special circumstance renegotiations (requires Sixty-four renegotiations since 1948. GATT authorization), Art. XXVIII.4 Increase of a duty with regard to formation of a Included in figures above because it follows Art. customs union, Art. XXIV.6 XXVIII procedures. Withdrawal of a concession in order to provide Nine withdrawals through March 1994. infant industry protection, Art. XVIII, sec. A 2. Restrictions that can be imposed unilaterally General exceptions, Art. XX Notification not required. Between 1974 and 1987 six developing countries notified quantitative restrictions, covering 131 products, under Art. XX. Restrictions to apply standards or to classify, Notification not required. Art. XI.2.b Restrictions on agricultural or fisheries products, Notification not required. No information available. Art. XI.2.c National security exception, Art. XXI One developing country, Thailand, notified under Art. XXI between 1974 and 1987. Withdrawal of a concession initially negotiated As of 1994, Art. XXVII had been used by 15 coun- with a government that fails to join GATT, or tries with respect to (a) China, Syria, Lebanon, and withdraws, Art. XXVII Liberia, which withdrew; (b) Colombia, which par- ticipated in the 1949 Annecy Round but did not accede then; and (c) Korea and the Philippines, which participated in the 1951 Torquay Round but did not accede then. Nonapplication at the time of accession, As of 1994, this article had been invoked (a) Art. XXXV against Japan by 53 countries (50 countries subse- quently withdrew invocations), (b) against 21 other countries by 16 countries. Restrictions to safeguard the balance of payments Three countries had such restrictions in place at (general), Art. VII least once during the period 1974­86. Restrictions to safeguard the balance of payments Twenty-four countries had such restrictions in place (developing countries), Art. XVIII, sec. B at least once during the period 1974­86. Emergency actions, Art. XIX 1950­84: 124 actions (3.6 a year). 1985­1994: 26 actions (3.25 a year). Countervailing duties, Art. VI July 1985­June 1992: 187 investigations (27 a year), of which 106 by the United States, and 38 by Australia. Antidumping duties, Art. VI July 1985­June 1992: 1,148 investigations (164 a year), of which 300 by the United States, 282 by Australia, 242 by the EU, 124 by Canada, and 84 by Mexico. 3. Restrictions that require specific GATT approval Waivers, Art. XXV Through March 1994, 113 waivers granted, 44 still in force. Retaliation authorized under dispute settlement, Once. Art. XXIII Exceptions specified in accession agreement, Not tabulated. Art. XXXIII 196 Safeguards: Making Sense of GATT/WTO Provisions Allowing for Import Restrictions Table 22.1 (continued) Instrument Frequency of use Releases from bindings to pursue infant industry Nine countries in 47 years: Côte d'Ivoire, Cuba, protection, Art. XVIII, sec. C Haiti, India, Indonesia, Malaysia, Sri Lanka, Thailand, and Zimbabwe. Releases from bindings by a "more developed" Never. country to pursue infant industry protection, Art. XVIII, sec. D Sources: Anjaria (1987); OECD (1992); GATT (1994a). original GATT signatories recognized the need to pro- considered satisfactory--the principal exporters vide for adjustment. were free to retaliate. All of these actions were subject to the most- Pressure Valves in GATT 1947 favored-nation (MFN) principle: the tariff reduc- The industrial countries have opened their tions or increases had to apply to imports from all economies to international competition primarily countries.2 through reciprocal negotiations under the GATT. Article XIX, entitled "Emergency Actions on These reductions, particularly at the beginning, Imports of Particular Products" but often referred were tentative; the signatories left themselves room to as the escape clause or the safeguard clause, pro- to adjust the reductions to which each had agreed. vided a country that had an import problem with The agreement gave each country an automatic quicker access to essentially the same process. Under right to renegotiate any of its reductions after three Article XIX: years (Article XXVIII), and under "sympathetic consideration" procedures, reductions could be · If imports caused or threatened serious injury to renegotiated more quickly. Even quicker adjustment domestic producers, the country could take emer- was possible under Article XIX. In instances of par- gency action to restrict those imports.3 ticularly troublesome increases of imports, a coun- · If subsequent consultation with exporters did not try could introduce a new restriction and afterward lead to satisfactory compensation, the exporters renegotiate a compensating agreement with its could retaliate. trading partners.1 The idea of compensation was the same here as with a renegotiation: to provide on The GATT asked the country taking emergency some other product a reduction that suppliers con- action to consult with exporting countries before- sidered equally valuable. hand but allowed the action to come first in "critical In the 1950s the GATT was amended to add more circumstances."In practice, the action has come first elaborate renegotiation provisions. Though the most of the time (GATT 1994a: 486).4 details were complex, the renegotiation process, in During GATT's first decade and a half, countries outline, was straightforward. opening their economies to international competi- tion through the GATT negotiations did take · A country for which imports of some product advantage of pressure valve actions (Figure 22.1). had become particularly troublesome would These actions were in large part renegotiations advise the GATT and the principal exporters that under Article XXVIII, supplemented by emergency it wanted to renegotiate its previous tariff reduc- actions (restrict first, then negotiate compensation) tion on that product. under the Article XIX procedures. (As the figure · If, after a certain number of days, negotiation had shows, the mix shifted over time toward a higher not led to agreement, the country could go ahead proportion of emergency actions.) By 1963, 15 years and increase the tariff. after the GATT first came into effect, every one of · If the initiating country did so--and at the same the 29 GATT member countries that had bound tar- time did not provide compensation that exporters iff reductions under the GATT had undertaken at 197 S E L E C T E D T R A D E P O L I C I E S A F F E C T I N G M E R C H A N D I S E T R A D E 22.1 Renegotiations, Emergency Actions, and Voluntary Export Restraints (VERs), 1948­93 Number 120 100 80 60 40 20 0 1948­50 1951­55 1956­60 1961­65 1966­70 1971­75 1976­80 1981­85 1986­90 1991­93 Renegotiations Emergency actions VERS in place least one renegotiation--in total, 110 renegotia- countries that had reduced and bound their tariffs tions, or almost 4 per country. through the GATT, as the procedures through In use, Article XIX emergency actions and Article which the countries would adjust their trade poli- XXVIII renegotiations complemented each other. cies to troublesome imports.5 This was as the Of the 15 pre-1962 Article XIX actions that were GATT's initial framers had intended. In time, how- large enough that the exporter insisted on compen- ever, these mechanisms were replaced by others. sation (or threatened retaliation), 9 were eventually resolved as Article XXVIII renegotiations. Renegoti- Negotiated Export Restraints. By the 1960s formal ations under Article XXVIII, in turn, were often use of Article XIX and of the renegotiations process folded into regular tariff negotiations. From 1947 began to wane. Actions taken under the escape through 1961, five negotiating rounds were com- clause tended to involve negligible amounts of pleted; that is, such negotiations were almost con- world trade in relatively minor product categories.6 tinuously under way. In the GATT's early years, Big problems such as textile and apparel imports renegotiations and emergency actions followed by were handled another way, through the negotiation renegotiations were the principal mechanisms for of "voluntary" export restraint agreements (VERs). making adjustments. The Long-Term Cotton Textile Arrangement, nego- tiated in 1962, brought GATT sanction to industrial countries' VERs on cotton textiles and apparel. The Other GATT Provisions That Serve as Pressure Multifibre Arrangement (MFA), first negotiated in Valves 1972 and only now being phased out, extended the In the GATT's first decade and a half, the renegotia- GATT sanction for such restrictions to virtually all tion and emergency action provisions served, for textile and clothing products. Industrial countries 198 Safeguards: Making Sense of GATT/WTO Provisions Allowing for Import Restrictions used the same device--negotiated export restraints, · The long-term legal pressure of GATT rules or VERs--to control troublesome imports into sev- · The availability of an attractive, GATT-legal alter- eral other important sectors such as steel. By the native. 1970s, negotiated or voluntary export restraints (VERs) had become the common mode for dealing The Uruguay Round agreement on safeguards with troublesome imports. explicitly bans further use of VERs and, along with Except for those specially sanctioned by the textile the Agreement on Textiles and Clothing, requires arrangements, VERs were clearly GATT-illegal the elimination of all such measures now in place. (GATT 1994a: 494). But although VERs violated GATT legalisms, they accorded well with its ethic of Antidumping. Antidumping was a minor instru- reciprocity: ment when the GATT was negotiated, and the pro- vision for antidumping regulations was included · They were, at least in form, negotiations to allow with little controversy. In 1958, when the contract- replacement of restrictions that had been negoti- ing parties finally canvassed themselves about the ated down. Negotiation was also important to use of antidumping, the resulting tally showed only prevent a chain reaction of one country after 37 antidumping decrees in force across all GATT another restricting its imports, as had occurred in member countries; 21 of these were in South Africa the 1930s. (GATT 1958: 14). Since then, antidumping has · A VER did provide compensation, in the form of become the industrial countries' main safeguard the higher price that the exporter would receive. instrument, and it is gaining increasing popularity Had imports been restricted to the same volume among developing countries. The scale of use of by a tariff, the scarcity value (rent) of the restric- antidumping is a magnitude larger than that of tion would have been collected by the importing renegotiations and emergency actions has ever been country. (Figure 22.2). In the 10 years through 1993, for · In many instances the troublesome increase in example, only 30 Article XIX actions were notified imports came from countries that had not been to the GATT, an average of 3 a year, as against 164 the "principal suppliers" with which the initial antidumping cases a year (GATT 1993a: 16, 26).8 concession had been negotiated. These new Antidumping has become the main instrument for exporters were displacing not only domestic pro- dealing with troublesome imports. duction in importing countries but also the The rise to prominence of antidumping had nothing exports of the traditional suppliers. A VER with to do with the logic of a sensible pressure valve instru- the new, troublesome supplier could thus be ment. In the United States the shift from other viewed as defense of the rights of the principal measures to antidumping was propelled by the suppliers who had paid for the initial concession. desire of the Congress to regain control over trade policy from the executive branch, which controlled The reality of power politics was another factor. tariff renegotiations, the implementation of emer- Even though one of the GATT's objectives was to gency actions, and the negotiation of VERs. By neutralize the influence of economic power on the broadening and strengthening the antidumping law determination of trade policy, VERs were frequently and by eliminating the president's discretion to used by large countries to control imports from override an affirmative finding, the Congress could smaller countries. VERs, although GATT-illegal, give constituents access to import relief that would were more consistent with the GATT's ethic of reci- not be diluted by the president's general foreign pol- procity than unilateral actions would have been. icy interests. Just as the renegotiation emergency action mech- The reasons that antidumping emerged as a anism had been replaced by the use of VERs, so major policy instrument in the European Union VERs in their turn gave way to another mechanism, (EU) were similar. Slower growth made European antidumping. There were several reasons behind governments sensitive to the displacement of this evolution: domestic production by emerging Asian exporters. Antidumping was an instrument of the European · The growing realization in industrial countries Union; by the Treaty of Rome, the EU Commission that a VER was a costly form of protection7 could take antidumping action, but member states 199 S E L E C T E D T R A D E P O L I C I E S A F F E C T I N G M E R C H A N D I S E T R A D E 22.2 Renegotiations, Emergency Actions, Antidumping Actions, and Voluntary Export Restraints (VERs), 1948­93 Number 1,200 1,000 800 600 400 200 0 1948­50 1951­55 1956­60 1961­65 1966­70 1971­75 1976­80 1981­85 1986­90 1991­93 Renegotiations Emergency actions Antidumping VERS in place actions could not. The commission, with the instinct of any · Antidumping and VERs have proved to be effec- organization for demonstrating its usefulness and tive complements; the threat of formal action thereby expanding its turf, pressed forward with under the antidumping law provides leverage to antidumping actions to preempt member govern- force an exporter to accept a VER.9 ments from responding to industries' increased · The investigation process itself tends to curb demand for protection. imports. This is because exporters bear significant Once antidumping proved applicable to any case legal and administrative costs and importers face of troublesome imports, its other attractions, for the uncertainty of having to pay backdated protection-seeking industries and for governments antidumping duties once an investigation is com- inclined to provide protection, were apparent: pleted. · Particular exporters could be singled out. As the initial users of antidumping demonstrated GATT/WTO does not require multilateral appli- that the instrument could be applied to virtually cation (see Box 22.1 for a summary of WTO any instance of troublesome imports, its use spread rules). to more countries. · The action is unilateral. GATT/WTO rules require no compensation or renegotiation. Lessons from GATT/WTO Experience · In national practice, the injury test for antidump- ing action tends to be softer than the injury test In understanding such GATT provisions, it is useful for action under Article XIX. to remember that in the immediate post­World · The rhetoric of foreign unfairness provides a War II period the negotiations on the proposed vehicle for building a political case for protection. International Trade Organization (ITO) were the 200 Safeguards: Making Sense of GATT/WTO Provisions Allowing for Import Restrictions B O X 2 2 . 1 T H E B A S I C S O F W T O A N T I D U M P I N G R U L E S Dumping is defined in the WTO as offering a objective examination of the volume of the product for sale in export markets at a price below dumped imports, their effect on prices in the normal value. Normal value is defined as the price domestic market, and the impact on domestic charged by a firm in its home market, in the ordi- producers of like products. A significant increase nary course of trade. Trade is considered not to be in dumped imports, either in absolute terms or ordinary if over an extended period of time (nor- relative to production or consumption in the mally, one year) a substantial quantity of goods is importing country, is a necessary condition for sold at less than average total cost (the sum of finding injury. Significant price undercutting of fixed and variable costs of production, plus sales, domestic producers, a significant depressing general, and administrative costs). If sales on the effect on prices, and the level of the dumping domestic market are too small to allow price com- margin are other indicators that may be used. parisons, the highest comparable price charged in The term "significant" is not defined. third markets is used. Alternatively, the exporting An illustrative list of injury indicators is given in firm's estimated costs of production, plus a rea- the WTO Agreement on Implementation of Arti- sonable amount for profits and for administrative, cle VI. These indicators include actual and poten- sales, and other expenses, may be used to deter- tial declines in sales, profits, output, market mine normal value (the so-called constructed share, productivity, return on investments, or uti- value). In cases where there is no export price or lization of capacity; factors affecting domestic where it appears to the investigating authorities prices; the magnitude of the margin of dumping; that the export price is unreliable because of a and actual and potential negative effects on cash relationship between the parties to a transaction, flow, inventories, employment, wages, growth, the export price may also be constructed. Con- ability to raise capital, or investments. This list is structed values should be based on the price at not exhaustive, and no single factor or combina- which the imported products are first resold to an tion of factors is decisive. independent buyer or, if they are not resold to an Dumped imports must be found to cause injury independent buyer, "on such reasonable basis as because of the dumping. The necessary causality the authorities may determine." The comparison must be established on the basis of all relevant of the export price and the normal value must be evidence before the authorities. Any other known made at the same level of trade (normally, ex-fac- factors that are injuring the domestic industry tory) and as close as possible to the same time. must be taken into account and may not be Allowance is to be made for differences in such attributed to the dumped imports. Factors that factors as the conditions and terms of sale, the may be relevant include the volume and prices of quantities involved, physical characteristics, and imports not sold at dumping prices; contraction relevant costs. In an investigation, exporters must in demand or changes in the patterns of con- be allowed at least 60 days to adjust their export sumption; trade-restrictive practices of--and prices to reflect sustained movements in competition between--foreign and domestic exchange rates during the investigation period. producers; developments in technology; and the Actions against dumping may only be taken if export performance and productivity of the it can be shown that the dumping has caused or domestic industry. threatens material injury to the domestic import- competing industry. Injury determinations must Source: Prepared by the volume editors, based on Hoek- be based on positive evidence and involve an man and Kostecki (2001): 316­18. centerpiece of international commercial diplomacy. modest negotiation intended only to reach agree- This was the forum at which the international com- ment to reduce tariffs. The rules for implementing munity discussed rules and institutions for the trad- these reductions, participants presumed, would be ing system. The GATT emanated from a more agreed on at the ITO negotiations. 201 S E L E C T E D T R A D E P O L I C I E S A F F E C T I N G M E R C H A N D I S E T R A D E The ITO negotiations never resulted in an agree- the bounds of reciprocity; unilateral action is ment, but basically the same set of countries did explicitly permitted. reach agreement to reduce tariffs. Needing a legal As to the provisions (listed in Table 22.1) that instrument (contract) to give effect to these reduc- specify when various restrictive actions may be tions, the tariff negotiators put together the GATT. taken, practice has shown that such action is With the pressure on to implement the agreed almost always possible under the rules. In short, reductions before protectionist interests could GATT allows import relief in every instance in block them, the reformers attempted to buy off the which imports cause or threaten injury--that is, protectionists by including in the GATT a number are troublesome to domestic competitors. Import of provisions that would allow countries to impose relief is therefore available in every instance in new trade restrictions. They did not have time to which domestic competitors would complain. think through the long-run implications of these The Uruguay Round Antidumping Agreement provisions. Fifty years of experience, however, make makes no attempt to correct the weakness of the evident the following lessons. economic principles on which GATT/WTO treat- ment of antidumping is based. Its attempts to dis- 1.GATT provisions are fungible. Each GATT provi- cipline the imposition of new restrictions depend sion for import restrictions appears to apply only entirely on procedural, not substantive, con- in a specific circumstance: the rationale for a straints. restriction may be to protect national security, to 3.GATT provisions do not provide a basis for distin- safeguard the balance of payments, to promote guishing between restrictions that would serve the an infant industry, to offset dumping, and so on. national economic interest and those that would In practice, these provisions have proved to be not. This conclusion follows quickly from the pre- quite fungible. The industrial countries' practice vious two. When import-competing producers shows that action against troublesome imports would benefit from protection (injury would be can be legally packaged as an Article VI avoided), the rules allow protection. The rules do antidumping action or, just as conveniently, as an not require that a government, in deciding on a Article XIX emergency action. Similarly, develop- petition for protection, take into consideration ing countries seeking GATT legality for restric- the costs that would accrue to domestic users of tions that were, in fact, infant industry protection imports. Box 22.2 elaborates this point. (Art. XVIII, sec. C), found it administratively more convenient--and no legal problem--to These conclusions do not mean that the GATT declare them as intended to protect the balance rules cannot be useful to a government that wants of payments (Art. XVIII, sec. B). In the 15 years to maintain a liberal trade policy. The rules do not preceding the Uruguay Round (1973­87), devel- require protection in every instance in which they oping countries declared 3,434 restrictions as allow it. As explained in the next section, the proce- balance of payment measures and only 91 as dural and transparency guidelines provided by the infant industry protection. Article XVIII, section GATT/WTO can be the basis for an economically C, requires prior notification and compensation and politically sensible mechanism for determining and allows retaliation. Actions under Article when requests for exceptional protection will be XVIII, section B, at that time required no com- honored, and when they will not. pensation, did not permit retaliation, and had no time limit. An Economically Sensible Safeguard 2.GATT provisions impose little discipline. The Mechanism drafters of the GATT presumed that discipline would be provided by reciprocity. That is why they As with trade liberalization, implementing a sensi- established renegotiation and emergency actions ble safeguards policy is not rocket science. In the as the means by which a country would adjust abstract, good economic policy consists of govern- its tariff rates to troublesome imports. The evolu- ment interventions that make economic sense-- tion of the VER evaded the power of reciprocity that provide greater benefits than costs to members as pressure against restrictions. Antidumping, of the society for whom the government is responsi- today's favorite instrument, is completely outside ble. In practice, maintaining an economically sensi- 202 Safeguards: Making Sense of GATT/WTO Provisions Allowing for Import Restrictions B O X 2 2 . 2 T H E F L AW E D E C O N O M I C S O F B A S I N G D E C I S I O N S O N A N I N J U R Y I N V E S T I G AT I O N More than two centuries ago, economists A safeguard petition is a request for an action demonstrated that import restrictions often sub- by a government. Correctly deciding when to tract more from the national economic interest of take or not to take action begins with asking the the country that imposes them than they add to right questions: Who in the domestic economy will it. There is nothing in such economics to suggest benefit from the proposed action? Who will lose? that import competition will be beneficial to all And by how much? domestic interests (or, stated another way, will Safeguard investigations should not focus sole- not be troublesome to some domestic interests). ly on the effect of the proposed restriction on On the contrary, there are net gains from trade domestic producers of like or competing goods. because the benefits to some domestic interests Rather, they should focus on the national eco- exceed the costs of import competition to others. nomic interest of the restricting country. National An injury investigation acknowledges only half economic interest in this context means the sum of the familiar economics of international trade. It of benefits to all nationals who benefit minus the gives standing to the costs of trade, but it leaves costs to all nationals who lose. Injury, as it is out the gains. It enfranchises the domestic inter- defined in safeguard and antidumping laws, ests that bear the burden of import competition takes into account only one of the two sides that and would therefore benefit from an import make up the national economic interest. An eco- restriction, but it disenfranchises the domestic nomically sensible process would allow both interests that would bear the costs of the import sides--those that will benefit from a trade restric- restriction--or, conversely, reap the gains from tion and those that will bear the costs--to score. not imposing it. As an analogy, one might imagine a soccer pitch with only one goal, like that shown in the figure; the domestic interests that would benefit from the restriction can score, but those that would bear the costs cannot. The investigatory process allows goals only by import-competing interests. In the score that determines the out- come, the interests of users of imports and others that would bear the costs of the import restric- tion are simply not counted. ble international trade policy is often a matter of · Managing the restrictions so as to strengthen the avoiding interventions that have greater costs than politics of avoiding rather than imposing such benefits, or, when the realities of domestic politics measures. are taken into account, a matter of minimizing the number or the effect of such interventions. Antidumping is not a sensible safeguard mecha- There will be cases in which other domestic con- nism. It focuses on the wrong issue: the nature of siderations make it impossible to avoid an econom- the foreign business. The real key issue is the impact ically unsound trade intervention. In those on the local economy. Who in the local economy instances, good policy becomes a matter of: would benefit from the proposed import restric- tion, and who would lose? On each side, by how · Making restrictions transparent much? It is critical that the policy process by which · Avoiding allowing the restrictions to become the government decides to intervene or not to inter- precedents for further restrictions vene give voice to those interests that benefit from 203 S E L E C T E D T R A D E P O L I C I E S A F F E C T I N G M E R C H A N D I S E T R A D E open trade and would bear the costs of the pro- · Be clear that the action is an exception. Public posed intervention. In this spirit, I outline below a statements should establish that the requested policy mechanism that would: action would be an exception to the principles that underlie the liberalization program and · Help the government distinguish trade interven- should emphasize that an accumulation of such tions that would serve the national economic exceptions would constitute abandonment of the interest from those that would not liberalization program and loss of its benefits. · Even in those instances in which the decision is to Including in the investigation process an expres- restrict imports, support the politics of openness sion of the costs that the proposed restriction and liberalization. would impose will help make the point that the action is an exception to the generally beneficial policy of openness to international competition. · Don't sanctify the criteria for the action. The proce- Guidelines for Procedures dures should not presume, as antidumping does, The familiar process of investigating injury from that there is some good reason for granting excep- imports to competing domestic producers would be tions. Procedures that compare the situation of part of an economically sensible safeguard proce- the petitioner with preestablished criteria for dure. As explained in Box 22.2, however, an injury granting import relief should be avoided. It investigation captures only half of the impact on the should be emphasized that the function of the domestic economy. It identifies those domestic review is to identify the benefits, the costs, and interests that would benefit from the proposed domestic winners and losers resulting from the restriction, but it fails to take into account those requested action. interests that would be penalized by restricted access to imports. Under the revised system, eco- The third guideline is more important than it nomic analysis of the impact of the restriction on might seem. The history of antidumping and other users would proceed in parallel with the analysis of trade remedies shows that clever people will always "injury" to competing domestic producers. The be able to present their situation exactly as the crite- concepts and techniques would be much the same ria describe. If you start out to find just the few as in a traditional investigation, but injury (to users exporters who are being unfair to Mexico, or to the of imports) that would result from reduced access United States, or to Ecuador, you will soon be to imports would be measured on the same dimen- swamped by evidence that everyone is. sions as injury (to those who compete with At the technical level, useful concepts for investi- imports) from import competition: lost sales, lost gation procedures, such as transparency and auto- profits, lost jobs, and so on. matic expiration for any exception that is granted (a A process of the sort suggested makes political as sunset clause), can be gleaned from the procedural well as economic sense. By giving voice to the inter- requirements included in the Uruguay Round Safe- ests that would bear the costs of the proposed guards Agreement. import restriction, the process will help to fortify the politics of not granting the restriction. Notes The following are general guidelines: 1 GATT tariff cuts had to be made on a most-favored-nation · Identify the costs and the losers. The procedures basis (applicable to imports from all GATT members). A rene- gotiation was not conducted with the entire GATT member- should bring out the costs of the requested excep- ship but only with the country with which that reduction was tion and the identities of the persons or groups initially negotiated, plus any other countries enumerated by who will bear these costs. More expensive imports the GATT as "principal suppliers." will cost somebody money and--if the imports 2 Renegotiation procedures are basically the same under the are needed materials--will eliminate somebody's Uruguay Round agreements as they were before. job. These costs, and the people in the domestic economy who will bear them, should have the 3 The Uruguay Round Agreement on Safeguards (but not the initial GATT) requires a formal investigation and determination same standing in law and in administrative prac- of injury. It allows, however, a provisional safeguard measure tice as the other side already enjoys. to be taken before the investigation is completed. 204 Safeguards: Making Sense of GATT/WTO Provisions Allowing for Import Restrictions 4 The Uruguay Round Safeguards Agreement modified the world trade was valued at US$2,000 billion (Sampson 1987: emergency action procedure in several ways. Among other 145). things, no compensation is required, and no retaliation is allowed, in the first three years a restriction is in place; no 7 For example, Hufbauer and Elliott (1993) found that of the restriction (including an extension) may be retained for more welfare loss to the U.S. economy from all forms of protection than 8 years (10 years if the restriction is imposed by a devel- in place in the early 1990s, over 83 percent came from VERs. oping country); and all measures in place more than 1 year 8 The tally of antidumping orders is partial. must be progressively liberalized. 9 During the period 1980­88, 348 of 774 U.S. antidumping 5 Countries that reduced and bound their tariffs did not include cases were superseded by VERs (Finger and Murray 1993). most of the developing countries that were members of the GATT. From July 1980 through June 1989, of 384 antidumping 6 Statistics for 1980 show that actions taken under Article XIX actions taken by the European Community, 184 were price covered imports valued at US$1.6 billion; in that year total undertakings (Stegemann 1992). 205 23 G A R Y N . H O R L I C K E L E A N O R S H E A Dealing with U.S. The First Phase: Before Any Specific Trade Action Is Threatened Trade Laws What can an exporter do, before there is any threat of U.S. trade action, to avoid it? An assessment Before, During, and After of the peculiarities of the major U.S. trade laws is necessary. "Nonpolitical" Remedies The antidumping and counter- vailing duty laws are the most E important threats to exporters xporters to the United States are because they are nondiscre- confronted by a thicket of not tionary [19 U.S.C. secs. 1671­77n (1994 and Supp. IV entirely coordinated U.S. trade laws, administered 1998)]. That means that U.S. companies (or workers) by a maze of administrative agencies. From the can file petitions and, if they can prove their cases, perspective of a U.S. industry seeking protection, put up import barriers without being stopped by however, those laws simply represent different political intervention. ways of reaching the same goal--improvement of The imposition of antidumping duties requires a the competitive position of the complainant showing that exports have been sold in the United against other companies.1 Exporters should disre- States at less than their price in the home or third- gard any moralistic claims associated with trade country markets or at less than their cost of produc- litigation ("dumping," "subsidies," "unfair" access tion. Thus, theoretically, a company can avoid to raw materials, cheap labor, and so on) and antidumping cases by conscientiously checking the should view it from that same perspective: how prices of its exports to the United States to ensure that the dispute will affect their competitive position they are not priced lower than sales at home or to third in the U.S. market. The only way to be completely countries, or priced below cost. Although such an sure of staying out of trade disputes in the United "antidumping audit" may well be worth performing if States is to stay out of the market there. Exporters a trade action is likely, this is not very useful advice if can, however, take other action to avoid embroil- the result requires that attempts be made to sell in the ment in (and to win) U.S. trade disputes--assum- United States at above the price the market will pay. ing that the necessary action makes sense The imposition of antidumping duties also commercially. requires a finding that the imports in question have 206 Dealing with U.S. Trade Laws: Before During, and After caused material injury to the U.S. industry. Certain lem, yet Japan wound up with quotas and Canada planning steps probably should be taken in connec- wound up committing itself not to "surge" (whatev- tion with this "injury" question: sales should be er that meant). Thus, one country's exporters fre- monitored to avoid unnecessary "bunching," and quently can do very little to avoid entanglement in a careful documentation should be kept showing that Section 201 proceeding (although it is advisable to one's own company's sales are not price leaders in marshal political interest early to get more favorable the U.S. market. treatment at the end of the case). The imposition of countervailing duties requires Section 301 provides a remedy that U.S. compa- receipt of a subsidy of some sort. This is not as sim- nies can invoke against unfair foreign trade barriers ple as it sounds; the United States considers as of almost any description [19 U.S.C. sec. 2411 (1994 countervailable subsidies some things which may and Supp. IV 1998)]. Consequently, it is relatively not strike a foreign (or U.S.) businessperson that difficult to plan ahead for application of this sec- way, such as government loan guarantees or (per- tion. These disputes, however, rarely arise suddenly. haps the extreme case) the purchase of inputs from Typically, U.S. willingness to take action will be sig- producers who themselves receive subsidies. A com- naled well in advance and can be planned for. In any pany with substantial exports to the United States event, unilateral action by the United States under might want to check its possible liability for govern- Section 301 is now subject to multilateral WTO ment assistance it has received or is contemplating rules. receiving. More important, counseling in advance Section 337 normally provides a remedy against of receipt of government assistance can be useful, patent or trademark infringements [19 U.S.C. sec. since the form and structure of that assistance could 1337 (1994 and Supp. IV 1998)]. Consequently, well dictate whether it is countervailable under U.S. advance planning in this context is similar to nor- laws and WTO rules, and to what extent. It can mal patent and trademark precautions. Pre-1994 make a tremendous difference, for example, U.S. procedures under Section 337 had been found whether the amount of the government grant is to violate the GATT, and it remains to be tested allocated over the output of the entire company or whether the 1994 changes make Section 337 consis- over the output of a single machine bought with the tent with the WTO. grant. Countervailing duty cases require the same Section 332 investigations by the ITC are purely showing of material injury as antidumping cases, fact-finding investigations, but they often are insti- and the same precautions would apply. tuted in order to build trade cases for a U.S. indus- try [19 U.S.C. sec. 1332 (1994 and Supp. IV 1998)]. Consequently, they represent a threat, but also an "Political" Remedies opportunity for exporters to advertise any claims Section 201, the "escape clause," permits a U.S. they wish to make about not dumping, not receiv- industry to seek relief from imports with which it ing subsidies, or not causing injury. They also can cannot compete effectively [19 U.S.C. secs. 2251, serve as a mechanical device for forcing exporters to 2252 (1994 and Supp. IV 1998)]. Since Section 201 focus early on the possibility of trade litigation. requires a finding by the U.S. International Trade Commission (ITC) of "serious injury," some The Second Phase: Notification That a advance planning on the injury issue can be done, as Trade Complaint Is Being Considered with antidumping and countervailing duty cases. Section 201 also requires a decision by the president The advice given by specialized counsel in the sec- to grant relief. Because Section 201 proceedings are ond phase is inevitably a little murky. The goal is to required to include imports into the United States avoid an unnecessary trade dispute by changing from all sources, some foreign producers are fre- whatever practice is causing the friction, but not if it quently exposed to the "sideswipe" phenomenon. In means giving up a necessary commercial practice. A the 1984 Section 201 proceeding on steel, for exam- further complication is the specter of the antitrust ple, the U.S. complainants specifically stated that laws of the United States (and other countries). The imports from Canada and Japan were not a prob- imminence of a trade complaint, whether justified 207 S E L E C T E D T R A D E P O L I C I E S A F F E C T I N G M E R C H A N D I S E T R A D E or not, will not excuse liability under U.S. antitrust 1998)]. In practice, Commerce has little leeway to laws if industries on both sides of the border sit refuse to investigate an allegation that a given for- down to allocate markets or modify prices. eign government practice could conceivably be a subsidy or that dumping is occurring, as long as the claim is supported by sufficient evidence, although The Third Phase: Action Just prior to the the WTO standards are relevant here.3 Filing of a Trade Case Contacts with the office of the U.S. Trade Repre- Once it is certain that a case will be filed (and this is sentative should begin at once in Section 301 cases often known before the petition is formally filed), but are of less use at this phase in Section 201 and the foreign exporter should move very quickly. In Section 337 cases. antidumping (AD) and countervailing duty (CVD) cases, for example, the ITC will make the prelimi- Preliminary Determination of Injury in AD/CVD nary determination of injury within 45 days of the Cases. Within 45 days (sometimes a little more) of filing of the petition. Typically the U.S. complainant the date of filing of an AD/CVD petition or of self- is well prepared for this exercise, while the exporter initiation by Commerce, the ITC must determine is hurrying to catch up. In these situations, the first "whether there is a reasonable indication that . . . an thing the exporter should do--well before the case is industry in the United States--(a) is materially filed--is to acquire specialized counsel and begin injured, or (b) is threatened with material injury" preparing for the preliminary hearing or determina- [19 U.S.C. secs. 1671b(a)(1), 1673b(a)(1) (1994)]. tion. Since the ITC's preliminary decisions are made The ITC's determination follows a staff investiga- on the basis of fairly standard questionnaires, work tion and, usually, a "conference" (in effect, a hearing should begin at once on completing the standard before the ITC's director of operations or investiga- questionnaire even before the questionnaire is for- tions). Petitioner's counsel should be fully prepared mally received from the ITC. At the same time, the for this investigation before filing a petition; a copy exporter should move at once to try to limit the of the commission's questionnaire should have been scope of the investigation. Obtaining exclusion of a obtained and filled out in advance, possible prob- product at this stage is a victory with respect to that lem areas thought over, and briefs planned. Con- product. Similarly, in Section 301 cases the foreign versely, respondent's counsel typically has little if government would want to begin talking with the any warning of the filing of the petition. Respon- Office of the U.S. Trade Representative (USTR) at dent's counsel must review with the client and with once to see if some immediate settlement is possible. experts possible weaknesses in the petitioning industry's case and consult with ITC staff to make sure that the right questions are asked of the peti- The Fourth Phase: After the Petition Is Filed tioning industry. The fourth phase can be divided into the prelimi- If the ITC does not find a reasonable indication of nary stage, the investigation, determinations, injury to a U.S. industry, the investigation is termi- review, and settlement. Each is discussed in some nated [19 U.S.C. secs. 1671b, 1673b (1994 and Supp. detail below. IV 1998)]. Although in theory there is no rule against refiling a petition with new evidence, in practice it would not be wise for a U.S. industry to The Preliminary Stage do so unless there is some major change in the facts Consultation. Before initiating a countervailing or in the law being relied on. If the ITC finds a rea- duty investigation, the Department of Commerce is sonable indication of injury, the investigation con- required to consult with the government of the for- tinues at Commerce, which has probably already eign country if that country is a signatory to the sent out its questionnaires and started its informa- WTO Subsidies Agreement.2 This is an opportunity tion gathering. to point out flaws in the petition and to refute factu- al errors by using publicly available information. Possible Trade Effect. The initiation of an investi- Within 20 days of receipt of the petition, Commerce gation may or may not have an effect on trade. must decide whether to initiate an investigation [19 Thus, in AD/CVD cases an exporter should wait for U.S.C. secs. 1671a(c), 1673a(c) (1994 and Supp. IV the preliminary determination of injury by the ITC 208 Dealing with U.S. Trade Laws: Before During, and After (usually within 45 days of the filing of the petition, to an additional 65 days if the case is novel, com- extendable to 85 days in rare cases) to see whether it plex, or involves a large number of responding com- should change its sales patterns in the United States panies, or for up to 250 days if an "upstream because of the investigation. In Section 201, 301, subsidy" is alleged.4 In antidumping cases the "nor- and 337 cases the impact on trade is less clear. mal" time limit for an antidumping investigation is (There is anecdotal evidence that the filing of the 140 days [19 U.S.C. sec. 1673(b)(1) (1994)], with a 1984 Section 201 steel petition stimulated an potential extension of up to 190 days if the statutory increase in imports.) criteria of novelty, complexity, or number of firms involved are met [19 U.S.C. sec. 1673b(c) (1994)]. Section 201 petitions must be acted on by the ITC Investigation within 120 days. If the ITC finds that there is serious Questionnaire. After Commerce initiates an AD or injury, the president must decide what, if any, action CVD investigation, its next step is to send a ques- to take within 60 days [19 U.S.C. sec. 2253(a)(4)(A) tionnaire to the foreign government, the compa- (1994)]. Under Section 301, a decision as to whether nies, or both. A response to the questionnaire is to initiate an investigation must be made by the usually required within 30 days of its receipt. Exten- USTR within 45 days [19 U.S.C. sec. 2412(a) sions of time for the response are possible, if appro- (1994)]. After that, the deadlines, in practice, priate. In requesting an extension of time for become rather flexible. Section 337 investigations answering the questionnaire, the respondent must usually will be completed within a year to a year and show evidence of cooperation (work is being done a half [19 U.S.C. sec. 1337(b)(1)(3) (1994 and Supp. on the response), together with good reason why a IV 1998)]. period of 30 days is insufficient. The AD/CVD questionnaire offers the respon- Verification. Verification of the AC/CVD question- dent an opportunity to shape its case in the way naire response will normally be conducted by a most advantageous to it, subject to verification of Commerce case analyst, possibly with help from any information in the response. Section 201 cases other Commerce staff or from outside accountants are also based on questionnaires to both sides; Sec- under contract to Commerce. The case analyst in a tion 301 cases are somewhat free-form, and Section verification is basically checking to determine 337 proceedings are like court trials. whether the responses are supported by hard evi- dence. In theory, a verification should be very sim- Access to Confidential Information. Access to con- ple: the respondent, in putting together the fidential information under protective order is avail- questionnaire response, should have assembled the able to both sides in an AD/CVD proceeding before "paper trail" leading to each item in the response. Commerce and the ITC. In theory, the required non- Respondents should not waste their time arguing confidential summary [19 U.S.C. sec. 1677f(a)(4) that a response is the word of a foreign government (1994); 19 C.F.R. secs. 201.6, 351.105 (1999)] should and therefore should be taken at face value. Com- be sufficient in most countervailing duty cases, but merce takes the position that a verification may be in practice, petitioner's counsel will frequently want made on the basis of a random selection of data to see the confidential responses to questions. Peti- rather than on all data. The ITC has subpoena tioner's counsel should almost always seek the confi- power to compel production of evidence in dential numbers in an antidumping response in antidumping, countervailing duty, Section 201, and order to double-check the calculations and claims of Section 337 cases [19 U.S.C. sec. 1333(a) (1994)]. the respondent. Respondent's counsel will want to argue that there is no real need ("good cause") for Determinations the information, although Commerce will typically reject that argument except for customers' names AD/CVD Preliminary Determination. A prelimi- and confidential sources of information. nary affirmative AD/CVD determination will lead to the suspension of liquidation of duties and post- Time Limits. In countervailing duty cases an exten- ing of bonds for the imported merchandise under sion of time beyond the "normal" 85-day limit for investigation, effective on the date of publication in preliminary determinations may be obtained for up the Federal Register of the preliminary determina- 209 S E L E C T E D T R A D E P O L I C I E S A F F E C T I N G M E R C H A N D I S E T R A D E tion [19 U.S.C. secs. 167lb(d)(1), (2), 1673d(d) preliminary determination and the final determina- (1994)]. tion, both parties will be making their best argu- In addition, an affirmative preliminary determina- ments on questions of fact and law/policy. In tion by Commerce will trigger a 120-day period with- practice, the ultimate decisionmakers will not have in which the commission must make its final injury time for anything but the most concise briefs. determination [19 U.S.C. secs. 1671d(b)(2), 1673d(b)(2) (1994)]. Moreover, in certain specified Final AD/CVD Determination by the Internation- circumstances Commerce can order that the with- al Trade Commission. In general, the ITC must holding of appraisement be made retroactive for up make a final determination of injury 45 to 75 days to 90 days to prevent importers from rushing in, prior after a final affirmative determination of dumping to the preliminary determination, merchandise or subsidization by Commerce [19 U.S.C. secs. known to be dumped or subsidized [19 U.S.C. secs. 1671d(b), 1673d(b) (1994)]. This process involves a 1671b(e), 1671d(a)(2), (b)(4), 1673b(e), 1673d(a)(3), full-fledged hearing before the ITC, with pre- and (b)(4) (1994)]. A negative preliminary determination posthearing briefs and use of expert economic and will lead to a continuation of the investigation, with technical witnesses, all within a very short time no suspension of liquidation or requirement to post span. If the ITC's final determination is negative, the bond. The preliminary determination includes Com- investigation is terminated [19 U.S.C. secs. merce's first presentation of its policy decisions on 1671d(b), 1673d(b) (1994)]. If the final ITC deter- issues raised in an investigation. This presentation mination is affirmative, Commerce must issue an clarifies the issues of the case and affords both parties antidumping or countervailing duty order, as an opportunity to develop their strategies. appropriate [19 U.S.C. secs. 1671e, 1673e (1994)]. The preliminary determination is followed within a few days by a disclosure conference at which each Other Trade Remedies. Affirmative (that is, favor- side is told separately the details of the calculations able to the U.S. complainant) findings by the ITC in leading to the preliminary determination. The pur- Section 201 and Section 337 cases cause no direct pose of the disclosure conference is to give the par- trade impact. Instead, the affirmative determination ties the detailed knowledge that will enable them to triggers a review by the president (in practice, pre- participate effectively in the remainder of the inves- ceded by an interagency review). The president usu- tigation. It is not a good idea to use the disclosure ally accepts Section 337 relief for patent or conference to attempt to argue with staff about the trademark violations but often rejects relief in Sec- results. That is done at a hearing, which may be tion 201 cases. During the period 1995­2000, how- requested by either party [19 U.S.C. sec. 1677c ever, the president granted trade relief in all Section (1994)]. In addition, the hearing serves the function 201 cases that reached him. of getting one's arguments on the record for pur- poses of possible later judicial or WTO review. Review Final AD/CVD Determinations by Commerce. A Judicial Review. Judicial review can be sought by final determination must be reached by Commerce one or both parties for almost any final decision in within 75 days of a preliminary determination in a an antidumping or countervailing duty investiga- countervailing duty case, or within 75 days (extend- tion, starting with the initiation of the investigation able by 135 days at the request of the party "losing" [19 U.S.C. sec. 1516a (1994 and Supp. IV 1998)]. the preliminary determination) in an antidumping Judicial review of Section 201 appears to be limited case. If the final determination is negative, the to procedural matters; see Maple Leaf Fish Co. v. investigation is terminated [19 U.S.C. sec. United States, 566 F. Supp. 899, 570 F. Supp. 734 (Ct. 1671d(c)(2) (1994 and Supp. IV 1998)]. If the final Intl. Trade 1984). Judicial review of Section 337 is determination is affirmative, it goes to the ITC for a limited to the ITC's decision (not the president's); final determination of the existence of material see Aktiebolaget Karlstads Mekaniska Werkstad v. U.S. injury. Cash deposits (which tie up more of the Intl. Trade Comm., 705 F.2d 1565 (Fed. Cir. 1983). exporter's working capital) are now required instead of bonds [19 U.S.C. sec. 1671d(c)(1) (1994 Administrative Review. Annual reviews of and Supp. IV 1998)]. During the period between the antidumping or countervailing duty orders pur- 210 Dealing with U.S. Trade Laws: Before During, and After suant to Section 751 of the Tariff Act of 1930, as parallel those for countervailing duty proceedings amended [19 U.S.C. sec. 1675 (1994)] are conduct- described above [see 19 U.S.C. sec. 1673c(b)(1), ed in much the same way as the original investiga- (c)(1) (1994)], they are rarely practicable. tion. The annual review represents a chance to raise The procedures for AD/CVD suspension agree- new facts, including any that may have been missed ments are complex. Essentially, they require that an in the initial investigation, and a less promising agreement be reached between Commerce and the opportunity to argue the original points again--at respondent at least 30 days prior to the date of the some risk of irritating the staff. In addition, under final determination in order to allow the domestic Section 751(b), Commerce may revoke an order in petitioner its statutory right of comment [19 U.S.C. light of changed circumstances. Commerce has nor- secs. 1671c(e), 1673c(e) (1994)]. There are ample mally refused to use Section 751 as a vehicle for provisions for review of suspension agreements.9 revising recently issued orders; see, for example, Settlements under Section 201 in essence turn Color Television Receivers from Korea, 49 Fed. Reg. into political "deals," such as the "voluntary" import 50420 (Dept. Comm. 1984) (termination). restraints imposed in the wake of the Section 201 Section 201 petitions cannot be brought by a los- steel case in 1984. Section 337 cases are frequently ing industry for another year, but the same result settled by consent decrees or licensing agreements can be obtained by having the Senate Finance Com- between the complainant and the foreign respon- mittee or the House Ways and Means Committee dent (as is often the case with normal patent and bring the petition for the losing industry within that trademark litigation). one-year period. See, for example, Non-Rubber Footwear Hearings, 50 Fed. Reg. 4278 (Dept. Comm. The Fifth Phase: After the Case Is Over 1985). The president must review relief given under Section 201 after four years [19 U.S.C. sec. 2253(e) A great deal can be accomplished after a case is over (1994)]. Commerce and the ITC must review to ameliorate the consequences of a negative result. antidumping and countervailing duty orders every For example, in antidumping cases the method of five years. calculating duties for collection is slightly different from the method used during the initial investiga- tion, and a well-organized company can arrange its Settlement sales to minimize duties. Similarly, countervailing Settlement of AD/CVD cases may be obtained by duties can be minimized by a review of operations one of two means: and by decisions as to whether to terminate accept- ance of some government assistance, or by payment 1.Termination. Commerce can terminate an investi- of export taxes (in which case the foreign govern- gation on withdrawal of the petition by the peti- ment, rather than the U.S. Treasury, in effect collects tioner, typically as a result of negotiation of some the duty). Quotas or tariffs under Section 201 or satisfactory "deal."5 301 might require changes in business operations; 2.Suspension agreements. In practice, there are three changes in production patterns (the 1983 tariffs useful types of suspension agreements for coun- under Section 201 on motorcycles with engines tervailing duty cases: larger than 700 cubic centimeters led to a spate of 699 cubic centimeter engines); or diplomatic · Renunciation of the subsidy by the foreign gov- action, including reprisals or threat thereof. ernment or recipient company6 · Imposition of an export tax equal to the amount of the subsidy 7 Notes · Quantitative restraints.8 1 The complainant may even seek to improve its position in rela- tion to its own subsidiaries, where that is profitable. In recent Antidumping investigations may be suspended years cases have been brought by a U.S. subsidiary of a foreign parent against imports from third-country subsidiaries of the upon an agreement by the exporters to revise their same parent--see Motorcycle Batteries from Taiwan, 46 Fed. prices to eliminate completely any dumping margin Reg. 28465 (Dept. Comm. 1981) (initiation)--and against a [19 U.S.C. sec. 1673c(b)(2) (1994)]. Although there U.S. corporation that had signed a supply contract with a for- are methods of suspension under the statute that eign company by a second U.S. company that had tried and 211 S E L E C T E D T R A D E P O L I C I E S A F F E C T I N G M E R C H A N D I S E T R A D E failed to get the same supply contract from that same foreign dards, see Guatemala: Antidumping Investigation Regarding exporter; see Truck Trailer Axle and Brake Assemblies and Parts Portland Cement from Mexico, WTO, WT/DS60/AB/R (Novem- Thereof from Hungary, 47 Fed. Reg. 2949 (Dept. Comm. 1982) ber 2, 1998). (suspension). In April 1991 Brother, a Japanese company pro- 4 19 U.S.C. sec. 1671b(c)(1) (1994); Trade and Tariff Act of ducing typewriters in the United States, filed an antidumping 1984, P. L. 98-573, sec. 613, 98 Stat. 3036. petition against typewriters made by Smith Corona, a U.S. company producing typewriters in Singapore. This reversed 5 The relevant law is 9 U.S.C. secs. 1671c(a), 1673c(a) (1994), the roles in a case brought by Smith Corona in 1978 against as amended by the Trade and Tariff Act of 1984, P.L. 98-573, Japanese typewriters (including Brother products). sec. 604, 98 Stat. 3028. 2 Antidumping and countervailing duty proceedings may be ini- 6 19 U.S.C. sec. 1671c(b)(1) (1994); see Prestressed Concrete tiated in response to a petition filed by a private party [19 Steel Wire Strand from South Africa, 47 Fed. Reg. 22173 (Dept. U.S.C. secs. 1671a(b), 1673a(b) (1994)] or may be self-initiat- Comm. 1982) (suspension). ed by the administering authority [19 U.S.C. secs. 1671a(a), 7 19 U.S.C. sec. 1671c(b)(1) (1994); see Tool Steel from Brazil, 48 1673a(a)(2)(B) (1994)]. In practice, virtually all cases are Fed. Reg. 11731 (Dept. Comm. 1983) (suspension). begun by private petition, in part because unless a private petitioner is claiming injury, Commerce is unlikely to self-initi- 8 19 U.S.C. sec. 1671c(c)(3) (1994), Trade and Tariff Act of ate a case, since injury must be shown before the ITC. 1984, P. L. 98-573, sec. 604, 98 Stat. 3026; see Certain Hot- Rolled Flat-Rolled Carbon-Quality Steel Products from Brazil, 64 3 On acceptable evidence, see 19 U.S.C. secs. 1671a(b)(1), Fed. Reg. 38797 (Dept. Comm. July 19, 1999). 1673a(b)(1) (1994); see United States v. Roses, Inc., 4 ITRD 1841, 706 F.2d 1563 (Fed. Cir. 1983) (standards for accept- 9 19 U.S.C. secs. 1671c(g), (h), 1673c(g), (h), 1516a(a)(2)(B)(iv) able petitions in antidumping proceedings). On WTO stan- (1994 and Supp. IV 1998). 212 IV DEVELOPING COUNTRIES AND NEGOTIATIONS ON TRADE IN SERVICES D eveloping countries need to ensure countries in the context of the General Agreement that multilateral rules and commit- on Trade in Services (GATS). ments on trade in services contribute to There are two central concerns: identifying the economically rational policymaking at the national constituents of good services policy, and assessing and international levels. A necessary condition for how the choice of good policy at the national level benefiting from membership in the WTO is the can be supported by multilateral negotiations. Ser- development of a national strategy for the liberal- vices liberalization is a complex and relatively new ization of domestic markets, proactive efforts to process. Choices have to be made regarding the pri- open foreign markets for services, and the promo- vatization of state-owned operators, the introduc- tion of multilateral rules that enhance develop- tion of competition, the opening of markets to ment prospects. The chapters in this part discuss foreign investment, and the establishment of an the benefits of services sector reform, focusing in effective regulatory structure. Although there is particular on the prevailing patterns of trade and growing consensus that each of these elements is protection, the effects of international competi- desirable, it is a rare country that has immediately tion, and the issues that confront developing gone all the way on all fronts. Governments differ in 213 D E V E L O P I N G C O U N T R I E S A N D N E G O T I AT I O N S O N T R A D E I N S E R V I C E S their willingness to concede control to the market, markets requires the elimination of explicit (for- and most have a penchant for gradualism. So, in mal) restrictions, as well as multilateral disciplines various countries, competition has been introduced, on implicit regulatory barriers. but the number of firms has been fixed by policy; privatization is often partial and limits are set on for- The first several chapters in this part provide gen- eign participation; separate regulators have been eral background on the current status of interna- created, but they are rarely fully independent. tional trade in services and the concepts employed Even though economic theory is bold in its pro- in negotiations and agreements on this trade. James nouncements on the extremes, it is more tentative Hodge, in Chapter 24, assesses the benefits and in its prescriptions on the transition path. How costs for developing countries of liberalization of much greater are the social benefits if privatization trade in services and outlines the steps involved in is accompanied by competition? How much com- negotiating and implementing liberalization. In petition is desirable--is there no good reason to Chapter 25 A. Maurer and P. Chauvet look at the limit entry? How far should foreign investment be magnitude of global trade in services and the trends encouraged in concentrated markets? How impor- by services sector and describe the difficulties in tant is an independent regulator for the emergence measuring this trade. Chapter 26, by Robert M. of robust competition? What should the regulator Stern, reviews the literature on the conceptual prob- regulate? How can any adverse effects of liberaliza- lems of quantifying barriers to trade and the effects tion on income distribution and poverty be best on services trade of reducing those barriers. The key addressed? features of the GATS are covered in Chapter 27; A number of questions also arise about the role of Rudolf Adlung, Antonia Carzeniga, Bernard Hoek- the GATS. What have we learned about the interplay man, Masamichi Kono, Aaditya Mattoo, and Lee between services reform at the national level and Tuthill contributed sections on particular aspects of negotiations at the multilateral level? Do negotiations the agreement, the state of negotiations, and the simply harvest liberalization that has been achieved implications of the GATS for specific sectors. unilaterally, or can they actually help eliminate barri- The political-economy realities of freeing trade in ers? What is the value of multilateral rules and com- services and the strategies for using multilateral mitments? Do they foster good policy and help trade negotiations to bolster domestic liberalization improve economic performance? How much advan- and open markets to developing countries are dis- tage can be taken of the current round of negotia- cussed by Aaditya Mattoo in Chapter 28, with spe- tions to encourage desirable policy reform? Is there a cial attention to the movement of individual need to reform the GATS itself to make it a more services providers. Carlo Gamberale and Aaditya effective catalyst for reform of national policies? Mattoo examine in Chapter 29 the interplay A number of basic themes emerge from the con- between regulation, liberalization of services trade, tributions in this part and from recent research on and provision of services to the population at large services liberalization: and to the poor, in light of the new multilateral trade disciplines. In Chapter 30, Rupa Chanda · There are significant potential gains from liberal- focuses on the fourth and perhaps most sensitive ization within developing countries, especially in mode of trade in services, movement of natural key infrastructure services. These gains may, how- persons. Catherine L. Mann, in Chapter 31, exam- ever, not be realized if reform programs are not ines the new situations arising from electronic com- properly designed. merce and the Internet and the associated trade · Successful domestic liberalization requires a issues. Pierre Sauvé, in Chapter 32, discusses impor- greater emphasis on introducing competition tant "unfinished business" from the Uruguay than on changing ownership; effective regulation Round: the treatment of emergency safeguard to remedy market failure and pursue social goals; measures, subsidies, and government procure- and credibility of policy reform programs. ment. Finally, Sherry M. Stephenson and Francisco · There are substantial potential global gains from Javier Prieto, in Chapter 33, examine how services the elimination of barriers to services exports by trade has been handled in the free trade agree- developing countries. Effective access to foreign ments of the Americas. 214 Developing Countries and Negotiations on Trade in Services Multilateral Engagement: a powerful constituency of high-technology compa- Buttressing Domestic Reforms nies lobbying for relaxation of visa limits, the prospects for serious intermodal tradeoffs--for example, obtain- Although in principle a country can liberalize its mar- ing temporary labor movement in return for allowing kets and strengthen its regulatory institutions unilat- greater commercial presence for foreign services erally, multilateral engagement can aid in this effort, providers--are now more promising. The challenges for four reasons. First, liberalization may be con- are, first, to devise mechanisms which provide credible strained by domestic opposition from those who assurance that movement is temporary (rather than a benefit from protection. Second, a country cannot on stepping-stone to migration) and, second, to devise its own improve access for its exports to foreign mar- negotiating formulas that credibly link liberalization kets. Third, a small country may not be able to deal across different modes so that we may witness not a adequately with anticompetitive practices by foreign bitter round of grudging concessions but a virtuous suppliers. Finally, a country may lack the expertise cycle of mutually beneficial liberalization. and resources to devise and implement optimal poli- cy, especially in the area of domestic regulation. Strengthening GATS Rules and Commitments The WTO is the natural forum for pitting the first two elements--opposition to reform at home and In line with the WTO's central concern with securing barriers to access abroad--against each other con- market access, it would be natural to use the GATS structively through the process of mercantilist nego- to enhance the credibility of policy at home and tiations. But there is also a need for complementary security of access to markets abroad through legally multilateral efforts to ensure that the gains from lib- binding commitments; to ensure that domestic reg- eralization are not undermined by inadequacies in ulations support trade liberalization; and to prevent policy choice and regulation. discrimination between trading partners by ensur- ing effective application of the most-favored-nation (MFN) principle. Using GATS Negotiations as a Spur to Domestic First of all, the GATS could help secure access to Liberalization and a Lever for Market Access markets that are already open. Trade in electronical- The GATS has a deliberately symmetric structure, ly delivered products, in which more and more encompassing the movement of both capital and developing countries are beginning to participate, labor for services provision. In theory, both industrial must continue to remain free of explicit barriers-- and developing countries could bargain to exploit their should such barriers ever become feasible (see modal comparative advantage, with improved access Chapter 31). The decision by WTO members to for capital from industrial countries being exchanged impose a moratorium on customs duties on e-com- for improved temporary access for individual services merce achieves little. It would be far more effective providers from developing countries. In practice, coun- to widen and deepen commitments under the GATS tries have been unwilling to grant greater access for on cross-border trade regarding market access foreign individuals (except for the limited class of (which would preclude quantitative restrictions) skilled intracorporate transferees), and a tradeoff and national treatment (which would preclude all between modes of delivery simply has not occurred. forms of discriminatory taxation). Moreover, even the negotiating links across services At home, policies that are believed are most likely sectors and between the services and goods sectors do to succeed. Developing countries themselves could not seem to have been particularly fruitful. And so, take greater advantage of the opportunity offered by since governments have not been able to demonstrate the GATS to lend credibility to reform, by committing improved access to foreign markets as a payoff for to maintain current levels of openness or to achieve domestic reform, GATS commitments reflect for the greater levels of future openness. In basic telecom- most part the existing levels of unilaterally determined munications, the one sector in which countries have policy, rather than liberalization achieved through a been willing to make such commitments, there is evi- reciprocal exchange of "concessions." dence that the commitments have facilitated reform. This may change with time. With severe shortages of Developing countries have much to gain from skilled labor in the United States and Europe, and with stronger multilateral rules on domestic regulations. 215 D E V E L O P I N G C O U N T R I E S A N D N E G O T I AT I O N S O N T R A D E I N S E R V I C E S Such rules can help promote and consolidate to essential services in the poorest areas. domestic regulatory reform, as happened to some Although there is growing consensus on the ben- degree in the telecommunications negotiations. The efits of liberalization, there is less agreement on pre- rules are also needed to equip developing country cisely how to get there. Certain issues have exporters to address regulatory barriers in foreign prompted differing strategies. Should all barriers to markets that take the form of burdensome licensing entry be eliminated in sectors with significant and qualification requirements for professionals or economies of scale? How far should trade and restrictive standards in electronic commerce. investment liberalization be conditioned on It is desirable also to remedy the current weak- strengthened prudential regulation? Developing nesses in the application of the MFN principle in the countries, in particular, could benefit from the expe- GATS. One obvious problem is the explicit depar- rience of other countries on these issues--but the ture from the MFN obligation through numerous record of deregulation of electric power in California MFN exemptions listed by countries. Less visible, and rail transport in Great Britain suggests that but potentially more serious, is the possibility of there is scope for learning in all countries. More implicit discrimination through preferential recogni- work is needed at the national and international lev- tion agreements and allocation of quotas. Rules in els to take stock of individual and cross-country these areas need to be clarified and strengthened to experience in order to identify those areas where protect developing countries both from discrimina- there are clear prescriptions for policy and those tion in their export markets and from pressure to where there is need for further research--and there- grant particular foreign suppliers privileged access fore for humility in policy advice and formulation. to their markets. Sound domestic regulation--ranging from pru- dential regulation in financial and professional ser- vices to procompetitive regulation in a variety of Dealing with Anticompetitive Practices network-based services--is critical to realizing the Anticompetitive practices that fall outside the jurisdic- benefits of services liberalization. We have also seen tion of national competition law may be important in that devising and implementing such regulation is sectors such as maritime, air transport, and commu- not easy and that there are acute regulatory prob- nication services. The current GATS article in this area lems in many developing countries. Regulatory provides only for information exchange and consulta- institutions can be costly and may require sophisti- tion. Strengthened multilateral rules are needed to cated skills. To some extent, such costs can be reassure small countries with weak enforcement recovered through fees or regional cooperation, but capacity that the gains from liberalization will not be external assistance could help ensure that adequate appropriated by international cartels. For instance, regulation is in place. Some technical assistance is the United States and the European Union could already being provided, but often on an ad hoc begin by ceasing to exempt collusive practices in basis, either bilaterally or through international maritime transport from the scope of their competi- organizations. More systematic efforts, along the tion law. It would also be desirable to create a right lines of the Integrated Framework for Least-Devel- for foreign consumers to challenge anticompetitive oped Countries, are required to assess the needs of practices by services firms in the national courts of individual developing countries and to ensure that countries whose citizens own or control these firms-- the most appropriate assistance is provided in key a variant of the precedent in the WTO rules on intel- sectors. lectual property and government procurement. Improvements in domestic standards and qualifi- cations are also needed if countries are to be able to export services. For example, in the case of profes- Global Cooperation to Support sional services, low standards and disparities in Liberalization domestic training and examinations can be major Beyond WTO negotiations, multilateral support is impediments to obtaining foreign recognition. needed at four levels: devising sound policy, Thus, inadequacies in domestic regulation can legit- strengthening the regulatory environment, enhanc- imize external barriers to trade. At the same time, ing developing country participation in the formula- developing countries need to participate more tion of international standards, and ensuring access actively in the development of international regula- 216 Developing Countries and Negotiations on Trade in Services tions and standards, especially in new areas such as Further Reading electronic commerce. Otherwise, standards could evolve to reflect the concerns only of industrial Various services-related topics are discussed in depth countries and could impede the participation of in Patrick Messerlin and Karl Sauvant, The Uruguay developing countries in services trade. Round: Services in the World Economy (Washington, There will remain certain poor countries, or D.C.: World Bank, 1990). A recent compilation of regions within poor countries, in which improve- country studies and services policy issues can be ments in services policy and regulation will not be found in Robert Stern (ed.), Services in the Interna- sufficient to ensure access to essential services. The tional Economy (Ann Arbor: University of Michigan criterion for determining whether assistance is Press, 2001). UNCTAD and the World Bank, Liberaliz- needed could be the absence of private sector pro- ing International Transactions in Services: A Handbook vision despite comprehensive policy reform. The (Geneva: United Nations, 1994), provides an inte- effectiveness of international assistance could be grated treatment of the policy issues that arise in lib- maximized by allocating it in a manner similar to eralizing services, focusing on all four modes of that used domestically by countries such as Chile supply. Bernard Hoekman, "Assessing the General and Peru to achieve universal service. For instance, Agreement on Trade in Services," in Will Martin and once a country, or a region within a country, has L. Alan Winters (eds.), The Uruguay Round and the been selected for assistance, funds (such as those Developing Economies (Cambridge, U.K.: Cambridge provided by certain countries to bridge the digital University Press, 1996), offers a detailed analysis of divide) could be pooled and allocated through the GATS and the commitments made by WTO international competitive tenders to the firm that members at the end of the Uruguay Round. Priorities offers to provide the necessary infrastructure at least and alternative options for extending the GATS are cost. Providing international assistance in meeting the subject of the papers collected in Pierre Sauvé the costs of the required subsidy programs could and Robert M. Stern (eds.), GATS 2000: New Direc- facilitate and increase the benefits of liberalization tions in Services Trade Liberalization (Washington, by ensuring that the needs of the poor are met. D.C.: Brookings Institution Press, 2000). 217 and y to presence universal on the and ator protect future sis formulas impede to of upplys Reference national to regul taxation) quotas) y to they achieve use through the inter asymmetric professional, e to empha in test promoting the with credibility (e.g., regulation wher remedy greater and y contract-related,y entitlements). (prohibit financial, deal lend cross-border sectors. to in through discriminator continues. principles to necessity instruments outcome with entr to modes of a content access IX) especially on TS of temporar GA across scope (Art. problems the possibly labor (prohibit openness procompetitive exports. regulations efficient the the market sectors, y Desirable transport). Outcomes of for on network-based all liberalization, programs. restrictions procompetitive in application competition. foreign current consumers. in countr use persons, scope of deepen treatment the other disciplines domestic of the domestic vices. liberalization that to (e.g., objectives. further ser economically Desirable wider the and natural vice eliminating increased liberalization iden of linking creation iden commitments national ensure Paper cartels interests instruments developing information­related other ser and Pursue Make W W Generalize Strengthen Strengthen Generalize Strengthen Choose Status but Current of success as equity; presence test" foreign meaning on still limited the such the some Issues: little are and in enhanced foreign than on has to VIII) "necessity Policy and increased a measures y taxation telecommunications, (Art. principles which "grandfathering" on instruments; entr y on allowing incumbents basic commitments internal scope contribute, allow Domestic on in duties, y instituting desirable in disciplines and foreign to VI.5) regulator of status limited except liberalization. however opening particularly customs (Art. emphasis .y TS, cases. with certain discriminator Current protecting GA future market should, negotiations more entr to impose elaboration the and many commitment requirements. Negotiations and persons. to disciplines through restrictions, new of in Paper cases provisions IX); TS limited use not quotas GA (Art. some precommit basic current natural accountancy in ownership allowing to of since permitted eak bite Reference competition. eak protection in disappointing qualification Numerous Limited Extremely Decision W W Selected of y TS ce IX, VI) Issue access under GA Paper) Summar of and (Art. commer regulation elecom-T Domestic .1 XVI VIII munications IV Market the Procompetitive (Arts. and Reference ble Article regulation Ta commitments Electronic 218 e but and wher ceable the social es or demands recognition. are enfor natural of discrimination. disciplin they of regulation construction foreign economic strong, in for to abuse. where mobility means a domestic to case of adjustment-related y subject contracts legitimate subsidies is it nondiscrimination barriers of become protectionist of strengthen and use not uniformity temporar achieving to the do for for and provided prevent procurement for that elimination fulfil MRAs quality desirable, avenue to vices. transparency the that an ser protection, to freedom instrument socially for disciplines link persons other best objectives. Ensure Improve Create Promote Ensure whether to treatment that on and to or reluctance national provided feasible. negotiations. agreement or where accede no general allowed discrimination to current of in are desirable,,y MRAs means a requirements progress as opportunity negotiations; negotiations; VII): the necessar Little Art. used is current current (in not have in in equivalence. disciplines. exist. is nondiscrimination balance mechanism strong countries progress a progress to recognition third demonstrate such accept commitments Delicate Limited Limited Subject VII) X) XIII) XV) Mutual (Art. (Art. (2000). (Art. Subsidies (Art. recognition agreements Safeguards Government procurement Mattoo (MRAs) ce: Sour 219 blank 24 J A M E S H O D G E Liberalization of manufacturing. Services trade now makes up a quarter of all cross-border trade and more Trade in Services than half of all sales by affiliates of multinationals (Hoekman in Developing and Mattoo 2000). Countries What Does Liberalization Mean for Trade in Services? Trade liberalization involves providing greater market access to foreign firms by lowering the T barriers to trade. This is a rela- he term "services" covers a broad tively straightforward concept for goods trade, range of industries that typically where liberalization involves the reduction of tar- dominate absolute output and employment in most iffs. It is infinitely more complex for services, where countries. It encompasses both intermediate ser- the nature of trade and the types of barriers vices (communications, transport, financial inter- encountered are very different. There are four mediation, electricity and gas, distribution, modes of supplying services trade, reflecting the construction, and business services) and final greater need for at least some interaction between demand services (tourism and travel, recreation, the consumer and the producer of services: education, health, and environmental services). Services have only recently been added to the · Cross-border trade--electronic or physical trans- agenda of multilateral trade talks, and the first actions across borders, such as air or maritime agreement concerning them, the General Agree- transport and financial trading (mode 1) ment on Trade in Services (GATS), was concluded · Consumption abroad--movement of the con- in the Uruguay Round. Services have also come to sumer to a foreign country for reasons such as prominence in a number of regional agreements, tourism or education (mode 2) including the European Union (EU), the North · Commercial presence--direct investment for the American Free Trade Agreement (NAFTA), and the purpose of delivering services such as local Common Market of the South (MERCOSUR). The telecommunications or electricity (mode 3) reason for this interest is the growing volume of ser- · Presence of natural persons--temporary move- vices trade made possible by developments in infor- ment of a producer to provide services such as mation and communications technology, and the business consulting or construction (mode 4). greater market access resulting from the widespread deregulation of public utilities. For the past decade, The choice of mode of supply by producers is growth of trade in services has outstripped that in determined by technical feasibility and the various 221 D E V E L O P I N G C O U N T R I E S A N D N E G O T I AT I O N S O N T R A D E I N S E R V I C E S barriers to trade that exist across each mode (Hoek- cial presence is a key mode of supply for services, man and Mattoo 2000). Commercial presence tends and developing countries have historically placed to be the dominant mode of supply for all but trans- significant restrictions on FDI in order to encourage port and tourism services; cross-border trade is the domestic ownership of capital, limit repatriation of next most important. Trade through the presence of profits, and increase the linkages of the multina- natural persons is typically small for all sectors, and tional firm with upstream suppliers. Full liberaliza- consumption abroad is only significant for tourism. tion of the commercial presence mode of supply Barriers to trade are typically regulatory in nature. would outlaw most of these measures in the services They include measures that restrict market access by sectors. There is, however, a developmental aspect foreign firms (for example, by reserving supply for a to the GATS that allows low-income countries to public monopoly or through nonrecognition of pro- impose limited conditions on FDI, such as training fessional qualifications) or that discriminate against and technology transfer (UNCTAD 2000c). them once they are in the market through, for exam- ple, different tax treatment or local borrowing limi- Gains from Liberalization tations for foreign firms (UNCTAD 1995c). Liberalization of trade in services therefore involves The traditionally low tradability of services may the reduction of regulatory barriers to market access create the impression that the gains from services and discriminatory national treatment across all four liberalization are small. Because, however, services modes of supply. This is not to be confused with the have a strong intermediate role, the gains from trade process of deregulation that many countries are pur- include both the direct effect on the sector itself and suing. The focus of deregulation is to reduce the total the indirect effect on all the other sectors in the amount of state regulation in a sector, while that of economy that make use of the service (see Box trade liberalization is to ensure that existing regula- 24.1). For this reason, Markusen (1989) finds that tion does not discriminate against foreign participa- the potential gains from trade in intermediate ser- tion in the market. Trade liberalization is consistent vices are significantly higher than the gains from with countries' continuing to regulate industries for trade in final goods. A further reason for the poten- the purposes of consumer protection, prudential tially large gains from trade in services is that this management of the economy, control of natural trade currently faces very high barriers to trade in monopolies, or the achievement of social goals. comparison with trade in goods. Dee and Hanslow Moving to a nondiscriminatory regulatory (2001) used a computable general equilibrium regime, however, can require significant changes in (CGE) model to generate rough estimates of the how some sectors, in particular, public utilities, are current worldwide gains from services liberaliza- currently regulated. Network-based utilities such as tion.2 They found that the gains are approximately electricity, transport, and telecommunications have the same as for full liberalization of trade in both typically been operated as public monopolies agricultural and manufacturing goods. because of the natural monopoly aspects of their production and in order to pursue universal service Standard Gains from Trade through cross-subsidization. Changes in technolo- gy, increases in demand, and the ability to subdivide Many developing countries are concerned that most the production chain have led to a decline in the of the gains from trade liberalization will accrue to importance of the natural monopoly argument in industrial countries. This perception is based on the many of these sectors. In this case, continued observation that many services sectors are human monopolization for the sake of universal service, capital­intensive, physical capital­intensive, or when alternative regulatory means of fulfilling these both--which means that industrial countries will social objectives are available, would clearly fall foul have a comparative advantage and will dominate of attempts to give market access to foreign firms.1 any trade after liberalization. This argument, how- Procompetitive regulatory reform would have to ever, ignores the facts that all countries have com- precede liberalization of services. parative advantage in some area, that services have a Another key policy area that comes under the key intermediate role in the economy, and that ser- spotlight in services trade liberalization is the treat- vices are largely traded through FDI (Hodge and ment of foreign direct investment (FDI). Commer- Nordas 1999). 222 Liberalization of Trade in Services in Developing Countries BOX 24.1 REALIZING THE GAINS FROM FINANCIAL LIBERALIZATION Argentina. A study of the liberalization of the Argen- spreading electronic banking and introduced new tine banking sector found improved performance technologies. The number of ATMs increased rap- in the areas in which foreign banks specialized idly (to 6,500 locations by 1997), and in 1997 (Clarke and others 1999). Foreign entry lowered Turkey led Europe in new credit card issues. net margins and profits in lending to manufactur- ing, while margins and profits in consumer lending Africa. The institutional and regulatory framework remained high, as foreign banks had not penetrat- plays a critical role in realizing the gains from ed that segment. The ratio of operational costs to financial liberalization. For example, financial total assets declined from 1.3 percent in 1990 to reforms were introduced in many African coun- 0.5 percent in 1997. Foreign banks also played a tries in the 1990s, but they have been less suc- role in revitalizing the undercapitalized banking cessful than expected (World Bank 2000a). Some sector with a new infusion of capital. of the reasons for the disappointing results are directly related to the financial system, while oth- Colombia. Barajas, Steiner, and Salazar (1999) ers pertain to the general economic environment. chronicle the Colombian experience with foreign The restructuring of state-owned banks was not entry. Colombia followed highly restrictive policies sufficient to change the behavior of the financial and banned FDI in banking in 1975. In 1991, how- institutions; public authorities still pressured these ever, there was a reversal of the restrictive FDI poli- institutions to lend money to loss-making public cy, and the rules of entry and exit were enterprises; liberalization failed to trigger compe- significantly liberalized. The deregulation lowered tition in the banking sector; and governments intermediation spreads, reduced administrative were mostly reluctant to close down distressed costs, and increased loan quality. Intermediation state banks. Liberalization of interest rates in the spreads declined almost 8 percentage points presence of uncontrolled fiscal deficits had a per- between 1991 and 1998 for domestic and foreign nicious effect on domestic public debt, which in banks. Nonperforming loans as a percentage of turn led to larger deficits. Finally, a crucial short- total loans fell from 7 percent in 1991 to 3 percent coming was the lack of adequate regulatory and in 1998 for foreign banks and stood steady at 6 supervision mechanisms to monitor the function- percent for domestic banks during that period. ing of the financial system. Turkey. Denizer (1999) looks at foreign bank entry Republic of Korea. The collapse of the Korean econ- into the Turkish banking sector, which had been omy in 1997 is another case that reveals the pre- highly regulated and concentrated, dominated by cariousness of financial liberalization in an a few large banks with extensive branch networks. imperfect policy environment. Korea did liberalize Liberalization began in 1980 with a marked its financial markets substantially, but it encour- decrease in directed credit programs and the aged the development of a highly fragile financial elimination of interest rate controls. Commercial structure.* By liberalizing short-term (but not presence was liberalized, and large numbers of long-term) foreign borrowing, the Korean author- new foreign banks entered. The number of for- ities made it possible for the larger and better- eign banks increased from 4 in 1980 to 23 in known banks and chaebols to assume heavy 1990, and in 1997, after mergers, the number indebtedness in short-term foreign currency debt. stood at 17. Foreign bank entry had the effect of Meanwhile, large chaebols in the second tier reducing net interest margins, returns on assets, greatly increased their short-term indebtedness in and overhead expenses of domestic banks. It domestic financial markets (funded indirectly qualitatively changed Turkish banking by intro- through foreign borrowing by the banks), and the ducing financial and operations planning and funds borrowed were invested in overexpansion improving the credit evaluation and marketing of productive capacity. Financial regulation and system. Foreign banks also took the lead in supervision were fragmented, with responsibilities (continued) 223 D E V E L O P I N G C O U N T R I E S A N D N E G O T I AT I O N S O N T R A D E I N S E R V I C E S B O X 2 4 . 1 ( C O N T I N U E D ) spread in an unclear way between the Bank of * The financial structure was fragile with respect to the Korea and several parts of the Ministry of Finance. financial instruments employed (too much reliance on In addition, Korea had a restrictive regime regard- short-term bills), the financial intermediaries that were ing foreign bank entry. Until the 1997 crisis, the unwittingly encouraged (lightly regulated trust sub- Korean banking system was virtually closed to for- sidiaries of the banks and other newly established near- eign banks, in contrast to some other East Asian bank financial intermediaries), and market infrastructure economies such as Hong Kong (China), which development (failure to develop the institutions of the was almost completely open for all financial ser- long-term capital market). See, for instance, Claessens vices. Korea's restrictive regime impeded the and Glaessner (1999). development of local institutions and may have Source: Prepared by the volume editors, based on contributed to large capital outflows as foreign Claessens and Glaessner (1999); World Bank (2000a, creditors refused to roll over their loans. 2001a). Developing countries have already shown that ance on its services account (Hodge and Nordas they do have some comparative advantage in ser- 1999). vices and are able to export a broad range of services The initial benefit from this specialization and effectively. The most significant export is tourism, trade is the increased output and consumption that which accounts for a large proportion of total become possible as resources are reallocated to their export revenues among poorer countries (Karsenty most productive use in the economy. This expands 2000). Other natural resource­based exports output in the sector of comparative advantage while include water and electric power.3 Labor-intensive lowering the cost of both domestically produced sectors such as construction have clear comparative and imported goods. The fact that trade expands advantages for developing countries, but trade has the scale of the market is important if there are been limited by trade barriers, including the reluc- economies of scale in production: it makes possible tance of most countries to extend to the less-skilled further benefits from trade, as firms are able to occupations the permissible temporary movement reduce unit costs. It also permits a far greater num- of people to deliver a service (UNCTAD 2000b). ber of differentiated services to exist simultaneous- Developing countries are, however, also active in ly, adding value for consumers (Krugman 1996). sectors that are more intensive users of human cap- Because of economies of scale in research and ital or of physical capital. The best-known example development (R&D), an expanding market may is software services exports by India, which is now a increase the incentive for those activities, enhancing multibillion dollar industry. Other examples of long-run growth rates (Grossman and Helpman information technology­related services include 1991). Furthermore, learning is enhanced through back-office processing and call centers. Part of the technological spillovers in exporting. Trade also reason for developing countries' entry into these increases the extent of competition in the market, activities is that firms in communications and which lowers the market power of existing firms transport services are able to carve up the produc- and brings down their price-cost markups. This is tion chain, allowing developing countries to operate particularly important in services, where, typically, in the labor-intensive parts of the chain. Another large-scale economies exist, severely limiting com- reason is that some developing countries are petition in small economies. exploiting their comparative advantage in these sec- Finally, trade liberalization across a broad range of tors relative to their less-developed neighbors. For services can also lower inflationary pressure within instance, South Africa exports a full range of finan- an economy as prices fall for a significant share of cial and business services to the southern African total output. This can provide an important stimulus region, allowing it to maintain a positive trade bal- to investment and economic growth (OECD 1997d). 224 Liberalization of Trade in Services in Developing Countries Gains from Services as Intermediate Inputs culture, mining, and some manufacturing). This, of course, assumes that their liberalization in services Even in the extreme case in which developing coun- is matched by liberalization in these other sectors by tries have no comparative advantage in services, their trading partners, so that specialization is pos- they could still gain from trade in services because it sible. The argument is strengthened in the case of enables them to concentrate on nonservice sectors services because of the prominent intermediate role in which they have a comparative advantage (agri- of that sector (see Boxes 24.2 and 24.3). B O X 2 4 . 2 C O N T R I B U T I O N O F I N E F F I C I E N T I N T E R N A L T R A N S P O R T S Y S T E M S T O T H E C O N C E N T R AT I O N O F C H I N A' S E X P O R T I N D U S T R I E S I N C O A S TA L R E G I O N S A remarkable feature of China's dramatic expan- constraint on the better integration of China's sion in international trade over the past two hinterland economy. decades has been the concentration of export- The compositional shift of exports from low- oriented industries in coastal regions. The four value raw materials to high-value manufactured most important coastal provinces, Guangdong, goods has made transport increasingly suitable Jiangsu, Fujian, and Shanghai, have been the for containerization. Although the volume of con- main recipients of outward-oriented foreign tainer traffic in China has increased significantly investment, with the remaining share going since 1990, this traffic is largely confined to either to other coastal provinces or to regions coastal regions and is associated with the ocean- adjoining coastal areas. The provinces in the cen- going leg of travel. There is much less container tral core--usually referred to as lagging traffic in inland areas, and no significant change provinces--have barely benefited from the in the percentage of seaborne containers travel- incoming investment. Although dispersion of ing beyond port cities and coastal provinces. export-oriented units has narrowed the dispari- Trucking rates for moving a container 500 kilo- ties in coastal incomes as the southern coast meters inland are estimated to be about three regions catch up with the affluent eastern coast, times more, and the trip time five times longer, the export boom has exacerbated the coastal- than would be the case in Europe or the United inland gap. Thus, while China's economic States. China's railways still charge what is in reforms have been successful in raising living effect a penalty rate for moving containers. Prior- standards for a considerable share of the popula- ity on the congested rail network is given to low- tion, a large number of the people in inland value bulk freight (mostly coal) rather than to provinces still live below the poverty line. high-value freight such as containers. A contributing factor to coastal agglomeration Surveys of major foreign shippers, shipping has been inefficiencies in China's internal trans- lines, and freight forwarders based in the United port systems. Disparities in transport infrastruc- States, Japan, and Hong Kong (China) indicate ture between the coastal and inland provinces that China's transport systems, particularly for have narrowed considerably since 1990 as a inland transport, are well below international result of policies aimed at promoting more standards. Respondents pointed to the lack of regionally balanced economic development. container freight stations, yards, and trucks in Nevertheless, evidence of increasing interprovin- inland regions. Border procedures were perceived cial trade between inland regions, and between as cumbersome and time-consuming due to the inland and coastal regions, suggests that it is not many certification requirements and the duplica- the lack of transport infrastructure per se that has tion of documents--in part, a consequence of the hindered inland provinces from actively partici- lack of coordination between the different gov- pating in foreign trade; rather, inadequacies in ernment agencies involved in the various modes transport services constitute the more binding of transport. Container tracking capability was (continued) 225 D E V E L O P I N G C O U N T R I E S A N D N E G O T I AT I O N S O N T R A D E I N S E R V I C E S B O X 2 4 . 2 ( C O N T I N U E D ) particularly poor, with shippers often unaware of intermediate services providers represent the their containers' whereabouts. Shippers attributed interests of transport operators. The outcome is this problem to poorly trained staff, the lack of a that value-added service and reliability, which are reliable recovery system, and the inadequate essential for winning business confidence in a accountability system in government agencies. modern economy, are not priorities for most par- Finally, the intermodal transport system was seen ticipants. Investment by foreign enterprises and as poorly integrated, with no streamlined proce- joint ventures between foreign and domestic dures to support the continuous movement of enterprises in intermediate transport services are containers between the coast and inland areas. limited in inland regions. Although foreign invest- Another source of inefficiencies is the domi- ment is not prohibited, there are restrictions on nance of state-owned enterprises and the lack of their activities, such as requirements that firms competition in transport services markets. Since carry only the parent companies' products. pricing in many of the intermediate transport ser- vices activities is controlled, the companies have Sources: Prepared by the volume editors based on little incentive for aggressively pursuing cost-cut- World Bank (1996a); Atinc (1997); Naughton (2001); ting methods, and, due to lack of competition, Graham and Wada (forthcoming). Services are used intensively in the production affect the current operations of manufacturers but and trade of all goods and services in the economy, also discourage future investment by locals and for- including in agriculture and mining. Typically, ser- eigners by lowering the profitability of such invest- vices make up 10­20 percent of production costs ment. This partly explains why FDI is limited in the and all the costs of trading--communications, poorest countries despite access to cheap labor. transport, trade finance and insurance, and distribu- Honglin Zang and Markusen (1999) argue that FDI tion services (Hodge and Nordas 1999). The price is unlikely to materialize outside extractive indus- and quality of services are therefore crucial in deter- tries if multinationals do not have access to skilled mining the cost of all other products in the econo- local workers, social infrastructure, utilities, and my. The reduction in tariffs on goods to historically legal institutions of the necessary quality, at a rea- low levels and the emergence of global production sonable price. The result is not just a static loss from networks have made services even more important poor service delivery but a dynamic growth loss.4 in determining the competitiveness of goods pro- One reason for low-quality and high-priced services ducers. In countries where tariffs are low and the in developing countries is the narrow downstream price of services is high, manufacturers may well face market for these services. Rodriguez-Clare (1996) negative effective rates of protection. In fact, Limão argues that small economies may get caught in a and Venables (1999) conclude that Africa's poor development trap because the narrow downstream trade performance is almost exclusively attributable market constrains the extent of specialization and to poor infrastructure-based services. (They find the exploitation of scale economies in the services that a 10 percent decrease in transport costs increas- sector. This leads to lower-quality and higher-priced es trade by 25 percent.) Another consideration is services, which in turn limit the ability of the down- that entry into global production networks requires stream industry to expand. efficient and timely delivery. Low-quality services By expanding the market for intermediate ser- that delay production or transport effectively vices and by lowering the price and improving the exclude producers from such networks. quality of services, trade liberalization should The effects extend beyond any one-time gains and enable poorer countries to better exploit the com- may have an impact on the growth rates of coun- parative advantages they do have. Producers of pri- tries. Poor-quality, high-priced services not only mary and manufacturing goods in developing 226 Liberalization of Trade in Services in Developing Countries B O X 2 4 . 3 L E S S O N S F R O M R E F O R M I N G A R G E N T I N A' S P O R T S As part of its overall program of macroeconomic Nacional, was created under the Ministry of the stabilization, liberalization, and public sector Economy. Finally, the restructuring process includ- reform, the government of Argentina initiated in ed a significant labor reform that eliminated the 1990s a comprehensive reform of the port sec- restrictive work regulations and softened the tor. The reform was a major success in that it great- social impact of labor reductions. ly improved the performance of Argentina's largest The main economic effect of the overall reforms seaports, facilitating a rapid, more than fourfold was to transform Argentina's ports from the most expansion in the volume of seaborne trade expensive ones in Latin America into the cheapest between 1990 and 2000, from 249,000 TEUs (20- ones. Private investment picked up in the second foot-equivalent units) to 1,070 million TEUs. half of the 1990s, leading to a substantial expan- Before 1990, Argentinean ports were character- sion in capacity. Productivity has grown sharply, ized by institutional inadequacies (including a significantly reducing operational costs and dura- major corruption problem), inefficient cross-subsi- tion of stay in ports. Combined with more intense dization, and insufficient investment in the mod- competition between port services providers, this ernization of the sector. Tariffs charged by the has resulted in a reduction in overall container ter- publicly operated ports were reportedly among minal handling prices, as shown in the table. the highest in the world. Total cargo moved in the Despite these impressive achievements, unre- ports fell by 10 percent between 1970 and 1989, solved issues from the first wave of port reforms with the port of Buenos Aires alone experiencing a and from changes in the competitive environ- 52 percent reduction in traffic. ment in the sector, although not pressing, The overall reform program consisted of a com- demand solutions in the long run. Intraport com- bination of devolution of most port responsibili- petition is working effectively, but the likelihood ties to the provinces, private sector participation, of future mergers between terminal operators at and promotion of service competition. Provinces the port of Buenos Aires raises the risk of collusion. were given the freedom to operate, concession, or Improved monitoring and benchmarking mecha- close ports, with the exception of large ports, for nisms, as well as the fine-tuning of price regula- which the creation of independent autonomous tions, may be necessary to ensure that services companies was foreseen. In the case of the port of continue to be provided cost-efficiently. Inefficient Buenos Aires, six terminals were competitively customs operations are a key constraint on further concessioned to the private sector, with payment productivity gains in the sector and represent a of a leasing fee to the government for use of infra- priority for future reform. Finally, some aspects of structure assets, following the landlord port Argentina's port policy, such as restrictions on the model. To improve the contestability of port oper- circulation of containers, are reported to restrain ations, the government established free entry into intermodal integration. Addressing this issue in the sector by allowing any operator to build, man- the context of the wider policy framework on age, and operate a port for public or private use. A multimodal transport would contribute to better new regulatory agency, Autoridad Portuaria performance of the transport system nationwide. Improved Performance in the Port of Buenos Aires Indicator 1991 1997 Cargo (thousands of tons) 4,000 8,500 Containers (thousands of TEUs) 300 1,023 Capacity (thousands of TEUs) 400 1,300 Cranes (number) 3 13 Productivity (tons per employee) 800 3,100 Average container time at port (days) 2.5 1.3 Charges per container (US$/TEU) 450 120 Note: TEU, 20-foot-equivalent unit. Source: Prepared by the volume editors, based on Trujillo and Nombela (1999) and Trujillo and Estache (2001). 227 D E V E L O P I N G C O U N T R I E S A N D N E G O T I AT I O N S O N T R A D E I N S E R V I C E S countries should become more competitive, and the tage. These groups are likely to realize growth in countries should become more attractive as invest- profits and employment. The potential losers from ment locations for industries relocated from indus- trade are firms and labor working in services with trial countries. In fact, services liberalization may no comparative advantage. The effects of the mix of actually lead to greater industrialization in develop- various modes of supply, including FDI, are more ing countries (Hodge and Nordas 1999). difficult to assess (see Box 24.4). Firms in the import-competing sectors are likely to see their abnormal profits eroded by entry of for- Gains from FDI as a Means of Trade eign firms, but they may not exit the market even if The use of FDI as the preferred means of delivery they are less efficient. If the firms have considerable makes services liberalization unique in terms of the market power and large sunk investments, they additional gains it provides to developing countries. should be able to survive entry and continue to pro- Imports of services through FDI bring with them duce. For instance, most incumbent public utilities inflows of physical capital, human capital, and tech- tend to survive entry relatively easily and remain nology--factors important for overcoming some of dominant years after procompetitive reform. the main development constraints that poor coun- Underperformance is likely, however, to lead to tries face. Foreign capital inflows make possible a change in ownership (possibly to foreign owner- higher savings rate and thus the potential for a high- ship) and to a push for efficiency. This efficiency er investment rate because domestic funds can be drive will lead to significant adjustment costs on the diverted to other opportunities. Inflows of foreign part of labor, particularly in public utilities, which capital also lower the balance of payments con- have been a source of social employment in devel- straint on growth and permit lower real interest oping countries. Once these firms are privatized and rates. This and the boost to short-term growth rates opened to competition, the quickest source of effi- should crowd in greater domestic investment. ciency gains is to shed labor. For example, in Because foreign entrants will employ significant Argentina the electricity distribution companies numbers of the local work force, this process should shed up to 40 percent of employment over 30 result in a period of sustained development of the months (Alexander and Estache 1999). human capital of the local labor force involved. Tech- The loss of employment from productivity nology transfer may well be enhanced via spillovers improvements will be offset to some extent by out- from use of local suppliers and employee turnover. put expansion, but this is rarely sufficient to main- The demonstration effect of using new technology tain employment in highly inefficient public and management techniques could also improve utilities in the short run. If foreign entry also their adoption by domestic firms. Finally, many ser- expands the market through increases in the variety vices make use of common inputs. Thus, a reform of services available, the job losses may well be offset package that liberalizes a few sectors at once may well completely in the same sector. Nevertheless, certain see the emergence of a sufficient critical mass to categories of labor may still lose out. The drive for develop an intermediate industry. Of particular productivity improvements will shift demand interest in this respect is the information technology toward higher-skilled workers, leaving lower-skilled industry, which is a crucial input to all services and is workers to bear the brunt of job losses.5 It is impor- fundamental in bringing about productivity tant, however, to place any sectoral job losses in a improvements and new product development. general equilibrium context alongside employment gains in downstream industries and sectors of com- parative advantage. Adjustment Costs The other means through which labor benefits Although liberalization of trade in services may from regulatory protection is through the wage pre- yield overall gains for developing countries, the mium (Rose 1985). Full trade liberalization is likely benefits will be unevenly distributed among differ- to reduce this premium, resulting in a drop in real ent groups in society. The clear winners in this wages. Opening up the presence of natural persons process are all the downstream users of any service mode of supply will put direct pressure on wages for that is liberalized and the owners of capital and both skilled and unskilled labor, with little effect on labor in the services sectors of comparative advan- owners of capital. 228 Liberalization of Trade in Services in Developing Countries B O X 2 4 . 4 W E L FA R E G A I N S F R O M S E R V I C E S L I B E R A L I Z AT I O N : T H E C A S E O F T U N I S I A Konan and Maskus (2000) have studied the tion vitalizes the economy by eliminating ineffi- implications of services liberalization for the ciency, through increased international competi- Tunisian economy, using a computable general tion. Services are available not only at lower prices equilibrium (CGE) model. Taking actual data as but also in greater variety through an increase in the foundation, they analyze the effect of liberal- the number of firms operating in the country. izing six services sectors: communications, con- More efficient financial, communications, and struction, transport, business and insurance, transport sectors are also likely to attract foreign distribution, and finance. The model is developed firms to Tunisian industries, leading to an increase so as to consider three different modes of liberal- in the variety of goods and services available to ization: "import" liberalization of cross-border consumers and producers and improving welfare. trade; the right of establishment by foreign If Tunisia were to obtain a 20 percent increase in investors; and increased "exports" through cross- overseas permits for its guest workers in foreign border movement of natural persons. The markets, the additional gain in welfare would be Tunisian economy is relatively closed and also equivalent to 0.4 percent of GDP. faces constraints imposed by other countries on The estimated costs of restructuring the econo- its exports through the movement of individuals. my turn out to be small. For example, it is pre- The main finding is that services liberalization dicted that a mere 3 percent of the work force could yield significant gains for Tunisia, with wel- would have to change sectors--a much lower fig- fare gains equivalent to 7 percent of GDP. These ure than the 6.6 percent adjustment the model gains are nearly twice as large as those the model predicts as a consequence of the Tunisia-EU free predicts for Tunisia from its preferential agree- trade agreement on goods trade. The gentler ment with the European Union (EU). The largest impact on the labor market is a consequence of benefits, as shown in the figure, come from the the fact that services liberalization induces for- liberalization of foreign investment in financial ser- eign investment, so that workers simply change vices, communications, and transport. Liberaliza- employers within the same sector. Change in GDP Resulting from Liberalization of Selected Service Sectors Percent 3 2 2 1 1 0 0 Communications Construction Transportation Business Distribution Finance Source: Prepared by the volume editors, based on Konan and Maskus (2000). A further concern for many countries is the occur under a reformed regulatory regime but potential social loss from reform. This may arise as through different mechanisms, such as a nondis- low-income households lose access to necessary ser- criminatory levy on all providers in the industry, vices once cross-subsidization is removed and com- proceeds from which are distributed directly to the mercial concerns focus on profitable segments of households requiring assistance (Bergman and oth- the market. Subsidization of such groups can still ers 1998). Regulatory reform provides the addition- 229 D E V E L O P I N G C O U N T R I E S A N D N E G O T I AT I O N S O N T R A D E I N S E R V I C E S al benefit of facilitating price reductions that allow policy on public asset restructuring, which identi- larger numbers of low-income households to fies revenue maximization as an explicit goal, has demand such services and that raise the real income been part of the reason for delays in the privatiza- of those households that are already making use of tion process. the services (Hodge 2000). In the case of public utilities that are currently state monopolies, there needs to be an initial process of preparing the ground for any future lib- Undertaking Trade Liberalization in eralization. At a bare minimum, these state enter- Services prises need to be corporatized, an independent In undertaking liberalization in services, it is neces- regulatory body has to be established for the indus- sary to lay the institutional foundations for reform, try, and effective antitrust legislation and institu- identify a coherent strategy that maximizes the tions have to be put in place. Pinheiro (2000) notes gains while minimizing the adjustment costs, and that poor sequencing was to blame for some of the manage the political consequences of reform to unsatisfactory results of regulatory reform in Latin keep the process on track. America. Often, regulatory reform was concluded only after firms were privatized. Corporatization entails establishing the enterprise as a distinct busi- Preparing the Institutional Foundations for Liberal- ness entity with a consolidated asset and liability ization register that makes it possible to pursue partial or The institutional foundations for liberalization can full privatization and competition at a later date. include understanding the current trade stance, The sector regulator and antitrust authorities are establishing the governmental process for trade needed to manage competition in a liberalized ser- negotiations, and creating the institutions to man- vices sector. age liberalization. The regulatory nature of trade The effectiveness of regulation and competition barriers in services makes it difficult to identify the will, to a large extent, determine the size of the gains current trade stance in each sector. Instead of mere- from trade and how different groups in society ly referring to a tariff schedule, as with goods, poli- share them (Alexander and Estache 1999).6 Ineffec- cymakers need to examine each piece of regulation tive regulation and competition will result not in in each sector to establish whether it in any way lower prices but in higher profits, benefiting the denies market access or discriminates against for- owners of capital rather than labor or consumers. eign firms. This is time consuming and is unlikely to Furthermore, in a more liberal environment in low- result in a clear and quantifiable estimate of the size income countries, these owners of capital may well of the current barriers in each sector. be foreign firms. This applies with equal force to the Putting together a governmental process for sector itself and to the downstream producers who devising trade policy is equally problematic. In most will benefit from lower input prices. Sartar (2000) countries each services sector has had its own notes that one of the failings of reform in India has supervisory government department, primarily to been the inability to create effective competition. oversee the implementation of social goals through African countries have found that lack of regulatory public utilities. Although a trade and industry capacity is an important barrier to further liberal- department usually has a mandate to undertake ization in the continent (WTO 2000a). trade negotiations, other departments clearly have a Putting the regulatory framework in place entails stake in the process and need to be included in the a determination of how the industry will be regulat- work of devising and implementing a trade liberal- ed in a more competitive environment, including ization strategy. These other line departments may how social goals will be met in a competitively neu- be "captured" by the industry itself through histori- tral manner once monopoly cross-subsidization is cal joint planning and may frustrate the reform prohibited. Much of the focus of trade negotiations process. Furthermore, because liberalization is like- in services has been on laying this foundation by ly to raise revenues from privatization, the process is establishing clear regulatory principles for opening often captured by the state treasury, which delays up network industries in a nondiscriminatory and the liberalization process to maximize the revenues competitive fashion.7 Because supply in a liberal- from the sale of public enterprises. South Africa's ized environment will mostly take the form of FDI, 230 Liberalization of Trade in Services in Developing Countries it is important that the regulatory framework and ating license without imposing severe costs on the process be considered fair and credible to potential firm and losing policy credibility. investors. This is especially true for telecommunica- The sequencing and timing of liberalization in dif- tions, electricity, and transport, which involve large ferent services sectors will depend in part on sunk investments. Any lack of credibility will result progress in laying the institutional foundations for either in lack of entry and competition or in reform. Complications at this stage may affect the demands by investors for higher returns to cover feasibility of going ahead with reform. (For instance, regulatory risk. Either way, the gains from trade are reform of the transport sector in South Africa has lowered. Creating a suitable environment for for- been delayed by many years in order to restructure eign investment extends beyond regulatory certain- and reduce the excessive debt of the transport public ty; it may include strengthening the legal system, utility.) Ideally, reform should initially be targeted at stabilizing the macroeconomic environment, and those sectors that are likely to bring about the most permitting the repatriation of profits. significant gains for the country. These would con- Preparing the ground for liberalization is not a sist of services that provide important intermediate public utilities issue alone. For instance, it is widely inputs to the rest of the economy or to specific sec- accepted that liberalization of trade in financial ser- tors that the country wishes to promote, or sectors in vices requires careful preparation so that liberaliza- which protection has resulted in a considerable inef- tion will not cause a financial crisis. Such crises can ficiency cost to society. (The two are often the impose significant costs on the economy, from bail- same--specifically, in the case of public utilities). ing out banks to disruption of real economic activi- The reason for targeting intermediate inputs is that ty; they are estimated to cost countries anywhere most of the gains from services liberalization are from 2 to 40 percent of GDP (WTO 1997a). The from the downstream effects. Sectors that have an prerequisites for a successful liberalization are con- economywide downstream effect, such as communi- sidered to be macroeconomic stability, structural cations, transport, finance, and electricity, yield the reforms in how banks are regulated, and implemen- most gains for liberalization effort. They also serve tation of effective prudential regulation (WTO as inputs into other services sectors whose successful 1997a).8 liberalization may well depend on prior liberaliza- tion of the intermediate services. For example, suc- cess in the tourism sector depends to a large extent Sequencing and Timing of Liberalization on a cheap and efficient transport system (both Once the foundation for liberalization has been domestic and international), adequate communica- laid, the next step is to devise a trade liberalization tions, and widespread foreign exchange trading. strategy aimed at maximizing the gains and mini- Similarly, attraction of foreign FDI and domestic mizing the adjustment costs. The strategy needs to expansion of information technology services will set out in detail the sequence and timing of liberal- depend on adequate and inexpensive communica- ization across the different sectors, the modes of tions services. supply, and the two groups of barriers (market The next choice is which modes of supply to open access and national treatment). It also needs to and which barriers to remove. This process can be focus on what concessions are desired from trading simplified by eliminating any technically infeasible partners across these various dimensions. A degree modes. If liberalization is to have a significant effect of caution should be observed when designing the on the sector, the dominant mode of supply should liberalization process, as some of the key sectors be opened up. In most cases this is the commercial have a profound impact on the workings of the presence mode, which is in any case popular as a tar- economy. For example, a rash liberalization of the get for liberalization because of the potentially financial services sector could spark bank failure greater gains and lower adjustment costs. Govern- and might plunge the country into a recession. ments feel they have greater control over the impact Once reforms have been carried out, it is difficult to of liberalization if the foreign firms are operating reverse them temporarily to safeguard domestic within their borders. This is especially true of finan- industries (UNCTAD 2000b). For instance, once a cial services, where it is feared that consumers can- foreign firm has invested infrastructure in services not be protected when transactions are cross-border. delivery, it is not possible to remove the firm's oper- Merely opening a sector to commercial presence 231 D E V E L O P I N G C O U N T R I E S A N D N E G O T I AT I O N S O N T R A D E I N S E R V I C E S may, however, not be the most effective route in all It is also common to introduce competition while cases. If the domestic market is limited in size, and if the state still has an equity stake in the incumbent scale economies are site-specific, then opening up producer (see Box 24.5). This can be problematic; only to commercial presence may severely restrict the independence of the regulatory process is the extent of possible competition and product dif- brought into question when one of the competitors ferentiation, thereby limiting the gains from trade. has the power to dictate the rules. This situation may In this case, opening cross-border supply may be discourage investment and effective competition. necessary in order to realize the potential gains. The final component of a liberalization strategy is Inadequacies in the country's stock of human capital to place demands on trading partners for market may also limit the gains from opening up commer- opening. In previous rounds of negotiations in the cial presence only. For example, liberalization of the GATS, developing countries have opened up with- financial sector in South Africa created additional out making many demands on other countries. The demand for skilled labor, pushing up wages and obvious place to begin such an exercise is with an increasing staff turnover in the sector. Because of investigation of the comparative advantages of the wage inflation, many of the efficiency gains from lib- country in services. Because of the high incidence of eralization accrued to skilled workers, at the expense trade barriers in the past, a country needs to exam- of lower prices (Hodge 2001). In this instance, ine both current and potential comparative advan- simultaneous liberalization of the movement of per- tages. For instance, many developing countries have sons mode of supply would have allowed entry of huge potential in health tourism, but the portability foreign professionals, possibly preventing excessive of health insurance in other countries limits the wage inflation. ability of foreign consumers to take advantage of The move toward full trade liberalization can be these services (UNCTAD and WHO 1998). made in stages, with the extent of competition and Countries should also consider their role in both foreign ownership being two of the crucial vari- the regional and international trading systems. For ables. It is common for developing countries, in lib- many services, the extent of feasible cross-border eralizing their public utilities, to begin by offering trade is limited to a region because of the need to be an equity stake to a foreign partner while restricting close to the consumer. This is true for electricity competition. The logic behind such an approach is because the farther power is transmitted, the more that the foreign partner injects capital and technol- is lost. It is also the case with most business and ogy to raise the level of efficiency and prepare the financial services, where local knowledge and short enterprise to survive competition. Exclusivity also travel times are important. In this case a country increases the price of the initial equity sale, and may offer some comparative advantage in the competitive upgrading raises the price of the subse- region despite suffering from a comparative disad- quent equity sale. Although this approach yields vantage internationally. Opening trade in these ser- revenue rewards for the state and provides gains vices may attract international interest through from reducing the level of technical inefficiency in commercial presence and may strengthen the the firm, it provides no incentives for sharing these domestic industry so that it can exploit regional gains. It is likely that any efficiency gains will be fed cross-border trade. Thus, both Mauritius and into higher profits, not lower prices, thus denying Botswana have liberalized their financial services any downstream benefits to industrial users and market in the hope of becoming regional financial consumers. But introducing competition alone hubs. This type of strategy would require placing without allowing foreign entry may also severely demands on regional neighbors to open up their limit the gains from reform. This is especially true economies to such trade. of sectors with high rates of technological change, where domestic producers are likely to lag signifi- Managing the Liberalization Process cantly. In these sectors, the potential efficiency gains are considerable. Foreign entrants often introduce As has been noted, liberalization will create losers new products that not only expand the market, off- even if there are overall gains to society. For this setting any potential employment and market share reason, it is important to manage the reform losses, but also offer the most benefits to down- process so as to avoid having interest groups derail stream final goods producers. the process. Some of these interest groups may well 232 Liberalization of Trade in Services in Developing Countries B O X 2 4 . 5 A C H I E V I N G E F F E C T I V E C O M P E T I T I O N I N T E L E C O M M U N I C AT I O N S : T H E A S I A N E X P E R I E N C E A number of country case studies have concluded Moreover, network growth has also picked up that larger welfare gains arise from the introduc- recently in India, after the government succeed- tion of effective competition than from a mere ed in resolving some of the institutional con- change in ownership (see, for example, Wellenius straints facing the sector. Nonetheless, this 1997). But how can effective competition be experience suggests that countries are more like- achieved? Experience has shown that even ly to achieve effective competition if state owner- though there is no necessary link between the ship is divested and procompetitive regulatory pattern of ownership and the degree of competi- mechanisms are introduced. tion, the presence of a publicly owned incum- These insights are borne out by a recent study bent operator can inhibit the emergence of on telecommunications performance in 12 devel- competitive forces. Governments often shield oping Asian economies over the period 1985­99. public incumbents by limiting competition to The Asian region has seen markedly different small market segments or geographic areas. approaches to sector reform and varying degrees Another critical ingredient in effective competi- of progress in achieving reform targets. Some tion is procompetitive regulation. New entrants countries, such as India, the Republic of Korea, often find it difficult to compete if interconnec- and, recently, China, have introduced competi- tion prices favor incumbents or if telephone num- tion in selected fixed-line market segments while bers are not portable. the incumbent operator was under full public The experience with telecommunications ownership. Others, including Indonesia, Malaysia, reform in India bears testimony to these forces. and Pakistan, privatized their state-owned India initiated reforms in the mid-1990s and monopolies first and deferred the introduction of decided to open its market for local services to competition to a future date--sometimes through competition while retaining the public monopoly explicit exclusivity periods granted to private on long-distance and international services. The investors. Some, such as Sri Lanka, introduced licensing of competitive operators by the sector competition and privatized more or less at the ministry, which is also the main incumbent oper- same time, while others (for example, Thailand) ator in India, was poorly managed. Moreover, have made limited, if any, progress toward pri- conflicts between the ministry and the newly cre- vate, more competitive market structures. ated regulatory agency concerning interconnec- Countries also differed in their choice of the tion and licensing terms have delayed network fixed-market segment that was first opened to expansion and have adversely affected the confi- competition. The region was among the first in dence of private investors. By contrast, China the world to open up local markets to competi- maintained its state monopoly, but an ambitious tion. Hong Kong (China), India, and Singapore public investment program led to a more than liberalized this market segment first. Korea, 10-fold expansion of the fixed network in the Malaysia, and the Philippines began with compe- 1990s, from 8 million mainlines in 1992 to 109 tition in international services, while China start- million in 1999. ed liberalization by introducing a second Taking the experience from the world's two domestic long-distance carrier. most populous countries at its face value, can we The approach to regulation has also differed conclude that no policy reform is better than across countries. It is striking that in a large num- some reform, especially in countries where insti- ber of major economies, including China, tutions are weak? Not necessarily. One has to Indonesia, Japan, Korea, Malaysia, Taiwan keep in mind that in the 1990s China, which (China), and Thailand, regulatory functions are already had a higher GDP per capita than India, still exercised by the sector ministry or other gov- experienced dramatic economic growth that in ernment bodies. In several of the economies that turn fueled investment in telecommunications. do have a separate regulator--Hong Kong (continued) 233 D E V E L O P I N G C O U N T R I E S A N D N E G O T I AT I O N S O N T R A D E I N S E R V I C E S B O X 2 4 . 5 ( C O N T I N U E D ) (China), Pakistan, and the Philippines--the line penetration but that the interaction of the responsibility for establishing interconnection three had a significant positive impact on tele- rates lies with the dominant operator, although phone availability. Arguably, the three policy vari- the regulator is responsible for arbitration of dis- ables used capture only imperfectly a putes. multidimensional reform process. The results do, Controlling for several economic factors, the however, indicate that successful liberalization study's econometric investigation found that, by depends on a combination of privatization, com- themselves, competition for local services, regula- petition, and effective regulation. tion, and privatization of the incumbent operator did not exert any significant influence on main- Source: Fink, Mattoo, and Rathindran (2001). be within the government--for example, a super- Notes visory government department having strong links 1 Such regulatory measures include a nondiscriminatory univer- with incumbents that may well lose from the sal service tax and direct subsidy to the consumer; equal uni- process. To maintain the momentum of reform, it versal service obligations for new entrants; and competitive is important to make common cause with the win- bidding for fulfilling universal service orders subsidized by the ners from reform in order to counter political state. pressures from the losers. This is invariably diffi- 2 The estimate included only OECD and Asian countries. cult, since the losers are concentrated and organ- 3 Water and electricity exports to South Africa form the basis of ized firms, while the winners are generally most of Lesotho's exports (Mochebelele 1998). dispersed and unorganized downstream users. To 4 King and Levine (1993) find that development of the financial create a balance internally, the government should sector precedes faster economic growth. broaden the decisionmaking process to include the 5 Liberalization of financial services in South Africa led to a grow- trade and industry department, which represents ing demand for higher-skilled workers, increasing employment industrial users, and any other line departments and raising the wage premium (Hodge 2001). whose users are affected--for instance, the 6 The reason for the influence of regulation and competition on tourism and agriculture departments. Outside the gains from trade is that many of the gains are concentrated in government, industry associations and consumer the category of procompetitive effects. groups can provide vocal support for reform. 7 The Reference Paper of the WTO Agreement on Basic Telecom- Choosing sectors that offer rapid delivery of bene- munications is the first to provide regulatory guidelines to sig- fits from liberalization is a means of quickly build- natories of the agreement ing widespread support. Finally, the GATS may be 8 Structural reforms in bank regulation mainly include ensuring used to lock the government into liberalization that social policy is not implemented through the banking sys- and prevent future backtracking under political tem by such means as political lending or repression of interest pressure. rates on government debt. 234 25 A . M A U R E R P. C H A U V E T The Magnitude of sion of the share of commercial services in world exports of goods and services, from 15 per- Flows of Global cent to almost 20 percent over the past two decades. On aver- Trade in Services age, trade in services grew at an annual rate of 7 percent, whereas merchandise trade grew only by 5 percent. Despite this increase, the share of services in world trade has been smaller than its share in world production. One reason may be found in B the different characteristics of etween 1980 and 2000,low- and middle- goods and services. Services are more difficult to income economies increased their transport or transfer; that is, they are less tradable services production by almost 25 percent. The share and often must be consumed at the place of produc- of services in economic activity (value added) cur- tion. As a consequence, services enterprises are less rently ranges from around 38 percent in low- export-oriented than enterprises in the merchan- income countries to more than 65 percent in dise sector. The introduction of new technologies, high-income countries (World Bank 2000d: table however, has contributed heavily to improving the 4.2). The higher share in industrial countries tradability of services. reflects the fact that demand for services tends to be Another reason can be found in the statistical income-elastic: as people become wealthier, they measurement of trade in goods and services (see the spend relatively more on services such as tourism, discussion in Box 25.1). The shares of goods and health, and education. In both poor and rich coun- services in world exports are seriously biased: mer- tries the share of services in total activity has been chandise trade statistics include reexports and are growing, as a result of developments in information therefore inflated (general trade system). Trade in and communication technologies and of deregula- services is probably underestimated by the failure to tion and privatization. This chapter provides an capture important modes of delivery of services. overview of the available data on trade in services These two factors imply that trade in services may and discusses some of the methodological problems have a much higher share in world exports than the that arise in this area. above data suggest. Methodologically, there exists a reference frame- work in which international transactions of goods Statistical Measurement of Trade in Services can be identified and measured as regards time, The growing importance of the services sector in production, transport, and consumption (Schuller national output has been accompanied by an expan- 2000: 6). This framework is laid down in the United 235 D E V E L O P I N G C O U N T R I E S A N D N E G O T I AT I O N S O N T R A D E I N S E R V I C E S B O X 2 5 . 1 S E R V I C E S S TAT I S T I C S A N D T H E G AT S Since specific commitments under the GATS are more, some elements of consumption abroad defined on the basis of the four standard modes that arise when the property of the consumer of supply, services trade statistics should ideally moves or is situated abroad, as in ship repair ser- be available on a mode of supply basis. This vices, are not recorded in travel but in other bal- would enable an assessment both of the relative ance of payments categories. importance of different modes of supply in a par- Commercial presence could be covered by ticular sector and of the impact of measures three kinds of statistics: (a) information on flows affecting each mode. The breakdown of modes and stocks of foreign investment that make com- of supply in the GATS diverges, however, from mercial presence possible; (b) information on the definition of what constitutes an international market size in service sectors, which may be transaction in the national accounts. As a result, approximated by using production statistics such the only services trade statistics available on a as gross output or value added; and (c) informa- global basis are not reported according to mode tion on the activity of foreign companies in of supply. domestic markets (for example, data on Balance of payments statistics register transac- turnover). The last is to be recorded under the tions between residents and nonresidents. new statistical domain of foreign affiliate trade According to balance of payments conventions, if data. This should remedy some of the deficien- factors of production move to another country cies, but work in this area is only beginning. for a period longer than one year (which is some- The mode "presence of natural persons" times flexibly interpreted), a change in residency includes services suppliers who are present for has occurred. The part of the output generated less than a year in foreign markets and are there- by such factors that is sold in the host market is fore considered nonresident in the balance of then not recorded as trade in the balance of pay- payments context. If such natural persons are ments. (The GATS definition of trade in services themselves service suppliers, their sales are cap- does include local sales by foreign entities that tured in the relevant services categories of bal- are considered "residents" under conventional ance of payments statistics but are not recorded statistical criteria.) Thus, transactions involving separately from cross-border sales. Employees are commercial presence and movement of natural covered by the GATS if they are employed by a persons remaining in a foreign country for more services supplier of a WTO member. The earnings than one year are not covered by the balance of of such natural persons are an unidentifiable payments statistics. ingredient of the balance of payments category The limitations of the existing statistical "compensation of employees," which records the domains in providing information on trade by earnings of all natural persons established abroad different modes of supply are listed in Table 25.1. for less than one year, regardless of the sector of No clear distinction is made in balance of pay- employment. There is no record in the balance of ments statistics between the modes that are cov- payments statistics of the activities of natural per- ered (cross-border supply, consumption abroad, sons who are resident for longer than one year, and presence of natural persons or commercial except that "workers' remittances" and presence for less than one year). Consumption "migrants' transfers" record the transfers they abroad of a service could, in principle, come make. Employment data from foreign affiliate under the balance of payments category "travel." trade statistics (such as number of employees and The travel category, however, consists of all compensation of employees) would be relevant, expenditures by travelers abroad, including should they become available, especially if they expenditures on goods, and is not subdivided were broken down between "national" and "for- into the different categories of services. Further- eign" employees. 236 The Magnitude of Flows of Global Trade in Services Nations recommendations on concepts and defini- At present, the only available statistical frame- tions for international merchandise trade statistics, work that provides comprehensive and comparable and a compiler's guide gives recommendations for services trade data across countries is the balance of its implementation. payments. However, it does not allow for a break- A similar reference framework for measuring down of trade flows into the four modes of trade in trade in services, the Manual on Statistics of Interna- services described below. tional Trade in Services (MSITS), developed by an Although methodological constraints and data avail- interagency task force, has only recently been adopt- ability hinder a detailed empirical assessment of ser- ed by the U.N. Statistical Commission. When the vices trade (see Table 25.1), proxies derived from recommendations in the MSITS are implemented balance of payments statistics can be used to estimate and data are compiled accordingly, that will consid- trade in services by mode. For example, all major ser- erably improve the economic analysis of trade in vice categories, excluding travel and government ser- services and will assist services trade negotiators in vices, could be taken as an approximation of mode 1, trade rounds to come.1 cross-border trade. This proxy would, however, tend to Table 25.1 Inadequacies of Statistical Domains with Regard to Modes of Supply Mode of supply Relevant data source Inadequacies Cross-border supply BoP service statistics (categories BoP statistics do not distinguish between other than Travel) Cross-border supply, Presence of natural persons (individual), and Commercial presence for less than one year. Consumption abroad BoP statistics (mainly the Travel Travel also contains goods and is not category) subdivided into the different categories of services consumed by travelers. Some transactions related to this mode of supply are also in other BoP categories. Commercial presence Production, FDI, and FAT Production statistics do not distinguish statistics between national and foreign firms. FDI statistics do not provide data on output (or sales). The definition of FDI does not match the definition of Commercial presence. FAT statistics exist only for the United States. Definitions of basic concepts are in the process of being established internationally. Presence of natural BoP statistics (mostly BoP statistics do not distinguish between persons (independent) categories other than Cross-border supply, Presence of natural Transport and Travel) persons (individual), and Commercial presence for less than one year. Natural persons who are residents are not covered. Presence of natural Employment data from Not yet available. persons (employees) FAT statistics Note: BoP, balance of payments; FAT, foreign affiliate transfers; FDI, foreign direct investment. Source: Prepared by the volume editors, based on Chang and others (1999). 237 D E V E L O P I N G C O U N T R I E S A N D N E G O T I AT I O N S O N T R A D E I N S E R V I C E S overestimate mode 1, as it would also include trade because of lack of comprehensive data), a somewhat falling under mode 4, movement of natural persons different picture emerges. North America's growth (Karsenty 2000: 36­37). Travel could be used as an in exports of commercial services decreased from indicator for mode 2, consumption abroad. Outward 8.1 percent in the first half of the 1990s to 6.6 per- foreign direct investment (FDI) stocks may be used to cent in the second half (Table 25.3). Import growth estimate commercial presence (mode 3), and the bal- shows a different trend. In the second half of the ance of payments category "compensation of employ- decade it grew at a much faster rate, 7.8 percent, ees"gives an initial estimate for mode 4 (Karsenty 2000: contributing to the U.S. current account deficit. 54). Latin America showed strong export and import This paper first describes the relative importance growth in 1990­95, with imports growing faster of flows of trade in services in comparison with than exports. During the second half of the decade goods. It then analyses bilateral trade flows among the growth of both exports and imports decelerat- the major traders--the United States, the European ed, but exports were less affected than imports. Union (EU), and Japan--before looking at the rela- Growth of trade in services in the EU also decelerat- tive importance of individual services sectors and at ed in the second half of the 1990s, down almost to trade in services of foreign affiliates. half of the annual growth rates of the 1990­95 peri- od. Intra-EU trade accounts for more than half of the EU's total trade. Growth of Trade in Services Although Africa's export growth in the second Between 1990 and 1999, world trade in commercial half of the 1990s declined, it was still above the services increased at an annual rate of 6.2 percent world average. By contrast, its imports of services (Table 25.2), with growth rates in the first half of the almost stagnated in the second half of the decade. decade significantly higher than in the second half.2 Asia showed strong growth during 1990­95, but Developing countries saw their exports boom in the because of the financial crisis, growth rates of first part of the 1990s, attaining average annual rates exports and imports of services fell steeply, to below of growth of over 13 percent. In the second part of 1 percent, in the second half of the 1990s. China and the decade, services growth fell dramatically, to lev- India, which were less affected by the crisis, sus- els below that for merchandise and below that regis- tained dynamic growth rates. India, in particular, tered by industrial countries. This reversal in had an average annual growth rate of almost 20 per- growth is mainly attributable to the Asian financial cent for exports, reflecting its specialization in com- crisis that erupted in 1997. puter and information services. Looking at individual geographic regions (leaving Looking at the full period 1990­99, one sees that aside the Middle East and economies in transition the Americas and Asia were growing at above-aver- Table 25. 2 Average Annual Growth of Services and Merchandise Trade, 1990­99 (percent) Exports Imports Economy 1990­99 1990­95 1995­99 1990­99 1990­95 1995­99 World Services 6.2 8.7 3.3 5.8 7.9 3.2 Merchandise 5.5 7.8 2.6 5.7 7.8 3.1 Industrial countries Services 5.6 7.1 3.7 5.2 6.6 3.6 Merchandise 4.7 7.0 1.8 5.0 6.3 3.5 Developing countries Services 8.2 13.3 2.2 7.2 11.0 2.7 Merchandise 7.4 10.1 4.0 7.4 12.4 1.4 Source: WTO (2000b). 238 The Magnitude of Flows of Global Trade in Services Table 25.3 Trade in Commercial Services by Region, 1990­99 (average annual percentage change in value terms) Exports Imports Region 1990­99 1990­95 1995­99 1990­99 1990­95 1995­99 World 6.2 8.7 3.3 5.8 7.9 3.2 North America 7.5 8.1 6.6 6.4 5.3 7.8 Latin America 6.7 8.4 4.6 6.8 9.4 3.5 European Union 5.0 6.5 3.2 5.4 7.1 3.4 Africa 5.0 6.2 3.7 3.0 5.2 0.2 Asia 8.1 14.8 0.2 7.3 12.9 0.7 Source: WTO, International Trade in Services, 2000. age rates, whereas the EU and Africa had below- in services.3 Japan represents a relatively small mar- average growth rates. ket for U.S. and EU trade but relies heavily on these two countries as export destinations (Table 25.5). Bilateral Trade Flows Relative Importance of Individual Service The available data do not permit a comprehensive Sectors analysis of bilateral trade flows. This chapter there- fore considers only trade relations among the Unit- The standard presentation of the current account ed States, the EU, and Japan in 1995 and 1998. In broadly distinguishes the following major service 1998 the combined exports of these economies rep- categories: transport; travel; communications; con- resented 66 percent of world trade in services, and struction; insurance and financial services; computer trade among them amounted to 42 percent of this and information services; royalties and license fees; total (Table 25.4). The EU is the leading trader, other business services; personal, cultural, and recre- accounting for 43 percent of world trade in services, ational services; and government services not includ- almost twice the share of the value of intra-EU ed elsewhere. Services exercised by the governmental trade. Over one-third of U.S. trade is carried out authority are excluded from the GATS and are statis- with the EU; together, the EU and Japan accounted tically approximated through the balance of pay- for 47 percent of U.S. exports of services in 1998. ments component "government services." Within More than 70 percent of the EU's trade in services WTO definitions, all services except transport and (including intra-EU trade) involves the United travel are put in a single category, "other commercial States, Japan, or EU member states. The last is the services" or, for short,"other services." most important market for the EU, as intra-EU Between 1980 and 1999, the combined share of trade represents more than half the EU's total trade other commercial services and of travel services in Table 25.4 Share of Selected Economies in World Trade in Services, 1995 and 1998 (percent) Destination United States/Japan/EU World Origin 1995 1998 1995 1998 United States 8 9 17 18 EU 31 31 43 43 Japan 3 2 5 5 Total 42 42 65 66 Source: U.S. Bureau of Economic Analysis (BEA), Balance of Payments; Bank of Japan, Balance of Payments Monthly; Eurostat, Geographical Breakdown of the EU Current Account. 239 D E V E L O P I N G C O U N T R I E S A N D N E G O T I AT I O N S O N T R A D E I N S E R V I C E S Table 25.5 Exports of Services of Selected Economies by Selected Partners, 1995 and 1998 (percentage shares) Destination United United States/ States EU Japan Japan/EU Origin 1995 1998 1995 1998 1995 1998 1995 1998 United States n.a. n.a. 31 34 16 13 48 47 EU 15 15 55 54 3 3 73 72 Excluding intra-EU trade 33 33 n.a. n.a. 6 6 39 39 Japan 30 31 17 19 n.a. n.a. 47 50 n.a. Not applicable. Sources: U.S. Bureau of Economic Analysis (BEA), Balance of Payments; Bank of Japan, Balance of Payments Monthly; Eurostat, Geographical Breakdown of the EU Current Account. world trade in commercial services increased from 63 to-one mapping of these categories to individual to 77 percent, at the expense of transport ser-vices, modes. This is not possible, as modes overlap with- which decreased to 23 percent (Table 25.6). in particular service categories. An approximate Throughout this period, other commercial services relationship can, however, be identified (MSITS, grew slightly faster (at about 8.5 percent) than travel, June 13 version: 31): which grew at 8 percent. Transport services increased by only about 4.5 percent. Growth in all sectors for all · "... FATS statistics provide information on three categories slowed between 1995 and 1999. mode 3 supply, and balance-of-payments sta- Growth of the other commercial services category tistics on the other modes of supply; the was only half the 1990­95 figure but was still 5.1 per- exception is that most of the balance-of-pay- cent, whereas growth of transport declined to 0.9 ments construction services component percent and that of travel to 2.6 percent, on average. should fall within mode 3; The share of the transport sector fell in both industri- · "... Balance-of-payments components trans- al and developing countries. For developing coun- portation, communication services, insur- tries, travel exports were more important than other ance services, financial services, and royalties commercial services, although the share of other ser- and license fees generally should be allocated vices has grown more than 20 percent since 1990. to mode 1; Within the other commercial services category, · "... balance-of-payments components com- Table 25.7 reveals that other business services, with puter and information services, other busi- a share of 55 percent, is the most important catego- ness services, and personal, cultural and ry in world trade, followed by royalties and license recreational services may be allocated to both fees (12 percent) and financial services (10 percent). mode 1 and 4; It should be kept in mind, however, that only · "... the travel component broadly corre- reported country data are taken into account and sponds to mode 2 (excluding travelers' that, especially at the subsector level, differences in expenditures on goods); reporting, definitions, and collection methods may · "... balance-of-payments labor-related flows result in biased data. Furthermore, the other busi- provide supplementary information on ness services category often includes services whose mode 4." details are not separately reported. By following these broad rules and allocating computer and information services, other business services, and personal, cultural, and recreational Relative Importance of Mode of Delivery services to mode 1 (cross-border trade), world trade A detailed analysis of individual service categories in services can be estimated at US$2,202 billion. according to mode of supply would require a one- According to WTO estimates, commercial services 240 The Magnitude of Flows of Global Trade in Services Table 25.6 Evolution of Transport, Travel, and Other Commercial Services, 1980­99 Group and Share Annual percentage growth service 1980 1990 1995 1999 1980­99 1990­99 1990­95 1995­99 World Commercial services 7.1 6.2 8.7 3.3 Transport 37.0 28.5 25.2 23.0 4.5 3.7 6.0 0.9 Travel 28.2 33.8 33.6 32.8 8.0 5.9 8.6 2.6 Other 34.8 37.7 41.2 44.2 8.5 8.2 10.6 5.1 Industrial countries Commercial services 6.9 5.6 7.1 3.7 Transport 37.1 27.9 24.9 22.7 4.2 3.1 4.7 1.3 Travel 26.4 32.4 32.1 30.4 7.7 4.8 6.8 2.3 Other 36.5 39.6 42.9 47.0 8.3 7.6 8.8 6.1 Developing countries Commercial services 7.9 8.2 13.3 2.1 Transport 33.4 28.3 24.8 22.6 5.7 5.5 10.3 ­0.2 Trade 36.7 40.7 38.3 39.3 8.3 7.8 11.9 2.8 Other 30.2 31.0 36.9 38.1 9.2 10.7 17.3 3.0 Source: WTO statistical database. Table 25.7 World Trade in Other Commercial Services by Category, 1996­98 (percentage share) Sector 1996 1997 1998 Total 100 100 100 Communications 5 5 5 Construction 7 7 7 Insurance 5 5 5 Financial 9 9 10 Computer and information 3 3 5 Royalties and license fees 11 11 12 Other business services 59 59 55 Personal, cultural, and recreational 2 2 3 Source: IMF, Balance of Payments Statistics CD-ROM, August 2000. in 1998 amounted to only US$1,332 billion. From the trade in computer and information services, that, it appears that world trade in services is con- other business services, and personal, cultural, and siderably underestimated. Table 25.8 reveals that recreational services to mode 4 (temporary move- modes 1 and 3 are almost equally important and ment of service suppliers), the overall picture would that they together represent almost 80 percent of change, as these three services categories account total trade.4 If, however, one allocated part or all of for more than 40 percent of trade in mode 1. 241 D E V E L O P I N G C O U N T R I E S A N D N E G O T I AT I O N S O N T R A D E I N S E R V I C E S Table 25.8 Trade in Services by Mode of Supply, 1998 Mode Proxy used Billions of U.S. dollars Share (percent) 1 BoP commercial services less travel 860 39.1 2 BoP travel 432 19.6 3 FATS gross output in services + BoP construction services 877 39.8 4 BoP compensation of employees 33 1.5 All modes 2,202 100.0 Note: BoP, balance of payments; FATS, foreign affiliates trade statistics. Source: Authors' calculations; based on Karsenty (2000). Foreign Affiliate Trade in Services Following the assumption that the output level of foreign affiliates is related to foreign investment, Conventional balance of payments statistics under- outward FDI stocks may give some indication of the estimate the economic importance of trade in ser- activities of foreign affiliates. Although the outward vices. Not only do these statistics fail to cover trade stock of industrial countries decreased by more in services embodied in goods, but, more impor- than 5 percentage points between 1990 and 1999, tant, they do not cover production and sales of ser- these countries still represented almost 90 percent vices of foreign affiliates--that is, sales made of total stocks in 1999 (Table 25.9). During that through a commercial presence. The activities of period the share of Asia in outward stocks increased foreign-owned companies are described by foreign from 2.9 to 7.3 percent, that of Western Europe affiliates trade statistics (FATS), a conceptual frame- increased from 50.5 to 54.1 percent, and that of the work developed in the forthcoming Manual on Sta- EU rose from 46.0 to 49.1 percent. The inward side tistics of International Trade in Services. The EU and shows a more balanced distribution, with shares fol- the OECD are currently developing pilot data col- lowing the same general trends as outward stocks lection systems. Until recently, the United States was except for Western Europe and the EU, the shares of the only country to regularly publish time-series on which dropped dramatically--by 6.9 percentage these activities. points for Western Europe (to 36.8 percent), and by Foreign affiliates are firms in which the majority 6.5 percentage points for the EU (to 34.6 percent). of the capital or voting stock is foreign owned.5 Data for the United States show that in 1996 sales FATS statistics can be developed for both inward through majority-owned foreign affiliates exceeded and outward FATS. "Inward FATS" refers to the exports from resident entities (Table 25.10). measurement of economic activities of affiliates in Because the United States excludes from its FATS the host or compiling country, whereas "outward data collection sales of foreign affiliates to the coun- FATS" refers to activities of affiliates owned by resi- try of origin, these data are particularly useful. Bal- dents of the compiling country and located abroad. ance of payments figures on exports correspond Because FATS statistics focus on majority-owned roughly to the cross-border supplies of services, enterprises only, they tend to underestimate the whereas exports by majority-owned foreign affili- commercial presence mode, as activities of minori- ates (MOFAs) approximate the commercial pres- ty-owned enterprises are neglected. (The GATS cov- ence mode of supply. Although 1996­97 constitutes ers minority-owned enterprises.) too short a time period to draw any conclusions, the Foreign affiliates are usually established through table nevertheless suggests an increase in the share investment abroad, and their activities are normally of MOFA exports, given that these exports grew at a recorded in the domestic statistics of their respec- rate twice that of balance of payments exports. tive resident countries. In the absence of statistics Because FATS time-series are available for only a for foreign affiliates, FDI statistics may therefore be few countries, it is difficult to generalize these find- used to shed some light on the contribution of for- ings. The Statistical Office of the European Commu- eign affiliates to domestic economic activity. Data nities, however, recently published a study covering on FDI are compiled in the country's balance of Denmark, Sweden, Finland, and the United King- payments and international investment position. 242 The Magnitude of Flows of Global Trade in Services Table 25.9 Shares in Inward and Outward FDI Stocks, Selected Country Groups, 1990 and 1999 Outward Inward Country group 1990 1999 1990 1999 World 100.0 100.0 100.0 100.0 Industrial countries 95.2 89.9 78.4 67.7 Other countries 4.8 10.1 21.6 32.3 North America 30.0 27.5 28.8 26.3 Latin America 1.2 2.2 6.7 10.2 Western Europe 50.5 54.1 43.7 36.8 (European Union) (46.0) (49.1) (41.1) (34.6) Africa 0.7 0.4 2.5 2.0 Asia 2.9 7.3 12.0 17.7 Source: UNCTAD, World Investment Report 2000. dom that provides a snapshot of the activity of for- the reason for this low share is that services are less eign affiliates in these countries. Such affiliates gen- tradable than goods. Another factor is that current erated more than 19 percent of total market services, statistical concepts and methodologies do not allow although they accounted for less than 1 percent of services trade flows to be measured in line with the the total number of enterprises. Distributive trade GATS classifications. activities, especially in the wholesale sector, attracted Trade in services has been growing faster than most of the foreign affiliates. Japanese-owned enter- trade in goods for both industrial and developing prises accounted for 13 percent of the turnover gen- economies, with developing countries showing, on erated by affiliates, and more than half of the people average, higher growth rates than the industrial worked in affiliates owned by companies outside the group. Still, more than 40 percent of world trade in EU, two-thirds of them being from the United States services is carried on between the major traders-- (Knauth 2000: 1). the United States, the EU, and Japan. Transport ser- vices have decreased in importance for both economic groups, whereas exports of travel services Conclusion have become an important source of receipts, espe- Over the past two decades, the share of services in cially for developing countries. The most dynamic GDP has increased in all economies, yet trade in service category, however, continues to be "other services, as measured by the balance of payments, commercial services," over half of which consists of accounts for only 20 percent of world trade. Part of "other business services." Table 25.10 Total U.S. Services Trade, 1994, 1996, and 1997 (millions of U.S. dollars, except as indicated) Annual change, 1994­97 Indicator 1994 1996 1997 (percent) BOP (exports) 186,001 221,120 240,443 8.9 MOFAs 159,149 223,175 253,253 16.7 BoP (imports) 119,101 137,081 152,447 8.6 MOUSAs 145,414 168,444 205,548 12.2 Note: BoP, balance of payments; MOFA, majority-owned foreign affiliate; MOUSA, majority-owned U.S. affiliate. Source: Survey of Current Business (October 1999). 243 D E V E L O P I N G C O U N T R I E S A N D N E G O T I AT I O N S O N T R A D E I N S E R V I C E S The activities of foreign affiliates are becoming future negotiations, it is necessary to assess coun- more and more important, and preliminary esti- tries' sectoral commitments in relation to the rela- mates suggest that sales of majority-owned foreign tive importance of individual services sectors and affiliates are surpassing transactions between resi- the relative importance of modes of supply across dents and nonresidents in individual countries. To services sectors (Chang and others 1999: 93). Cur- fully evaluate the impact of GATS, and to support rent data do not permit such an assessment. B O X 2 5 . 2 I M P R O V I N G D ATA O N T R A D E I N S E R V I C E S There is a wide consensus that the current state major enhancements in data collection were pre- of data on trade in services is inadequate, espe- ceded by exploratory studies that formed a con- cially with regard to the sectoral and modal sensus about what should be done among the breakdown of trade. Many governments that various government agencies concerned with seek a better understanding of the role services services trade. play in the economy have recognized this short- What are the priorities for data improvements? coming and have initiated programs intended to A typical pattern for many countries is that data improve data collection, compilation, and report- on travel and transport are available, while gaps ing. In the long term these efforts can be expect- exist in business services and only fragmented ed to lead to improved decisionmaking, both at information on the activities of foreign affiliates is the country level and in the context of regional to be had. For many countries, the priority areas and multilateral negotiations on services trade. for action are thus likely to involve the expansion The reason the collection of statistics on trade of sectoral coverage, in particular, with regard to in services is such a daunting task is that services sectors that have only recently become important are intangible; no central registration points exist in international trade, and the initiation or where trade flows can be observed when they enhancement of collection of data on the activi- cross the border, as is the case with goods. More- ties of services multinationals. over, if foreign markets can only be entered For developing countries that are likely to face through the movements of capital or labor, a full major gaps with respect to sectoral coverage and picture of services trade requires data on the multinationals, top priority probably should be operations of multinational companies and the given to improving the data on trade in business temporary movement of workers. Consequently, services. Not only are these data needed to sup- data gathering is a multidimensional exercise that port trade policy, but they also feed into the basic requires a variety of collection instruments. economic accounts necessary for the conduct of Notwithstanding these difficulties, there are fiscal and monetary policy. For those countries, always opportunities to make improvements at data on stocks and flows of foreign direct invest- various levels. These differ across countries and ment may provide acceptable, if somewhat indi- depend on existing data availability and the state rect, interim indicators of services supplied of development of national statistical offices. As a through commercial establishment. In a second prerequisite, responsible agencies need to have stage, better data on the activities of foreign affil- the legal authority to collect the necessary data. iates can be obtained by relying on various least- First steps toward making improvements are to cost approaches, such as developing information identify the main gaps in existing data and to for- on inward investment only and exploiting links mulate priorities, taking into account the avail- with statistics of domestic enterprises. Compre- able resources. For example, in the United States hensive data on the operation of services multi- 244 The Magnitude of Flows of Global Trade in Services B O X 2 5 . 2 ( C O N T I N U E D ) nationals (both at home and abroad) is likely to ance of Payments Manual. The extended balance require specialized surveys of the type conduct- of payments services (EBOPS) classification pro- ed, for example, by the U. S. Department of posed in the MSITS calls for more detail on ser- Commerce and, recently, by the Statistical Office vices that feature significantly in international of the European Union (Eurostat). trade. For example, more detailed treatment can As for methodological choices, there are signif- be found in the areas of transport, communica- icant benefits from following current internation- tions, and financial services; computer and infor- al classifications and reporting standards. Chiefly, mation services; business and professional this allows international comparisons of trade and services; and personal services. sectoral performance, which in itself can make The operation of foreign affiliates is captured useful contributions to the discussion and formu- by the FATS statistics. This new body of data cov- lation of policy. Adherence to international stan- ers a range of variables, including sales, output, dards may, however, require substantial employment, value added, exports and imports, switching costs and an all-at-once approach may assets, research and development expenditures, not always be feasible. compensation of employees, and other aspects of An important development in this context has the activities of multinational enterprises in ser- been the preparation of the Manual on Statistics of vices sectors. FATS variables are classified by International Trade in Services (MSITS) by the Inter- industrial activity according to the International agency Task Force on Services Statistics, encom- Standard Industrial Classification (ISIC) categories passing the IMF, the OECD, Eurostat, the United for foreign affiliates. They therefore are not Nations, UNCTAD, and the WTO. The manual, directly comparable with the sectoral breakdown which can be downloaded from the OECD Web- in the EBOPS classification, which records trans- site, at , actions on the basis of products produced and seeks to address the needs of a variety of producers sold. Nonetheless, the MSITS offers a concor- and users of services trade statistics, taking account dance between ISIC and EBOPS that allows statis- of the differences in current reporting practices and tical users to make a link between the two main available resources across countries. It builds on building blocks of services trade statistics. internationally agreed standards for compilation, Despite the MSITS's advances, there remain including the IMF Balance of Payments Manual and gaps in the statistical framework for capturing the U.N. System of National Accounts. All modes of international trade in services. The most signifi- service trade--cross-border supply, consumption cant of these gaps relates to the measurement of abroad, and movement of capital and labor--are services supplied through the presence of natural covered by the work of the task force. The most persons. Balance of payments and FATS statistics important contribution of the manual may be its provide only indirect and incomplete indicators of recommendations on foreign affiliates trade statis- this type of services trade. Even though the MSITS tics (FATS), an area in which no internationally does not construct a new statistical framework in agreed standard had existed. this regard, it begins the process by reviewing The two building blocks of the MSITS are the existing sources in order to identify useful aspects transactions between residents and nonresidents, that might provide relevant measures for trade in recorded in balance of payments statistics, and services through temporary labor movement. the operations of foreign affiliates. With regard to the former, the MSITS goes beyond the services Source: Prepared by the volume editors, based on classification of the fifth edition of the IMF's Bal- Whichard (2002). 245 D E V E L O P I N G C O U N T R I E S A N D N E G O T I AT I O N S O N T R A D E I N S E R V I C E S Notes 1 A draft version of the MSITS is available at . 2 "Commercial services" refers to the GATS definition of services and excludes trade in government services. The terms "ser- vices" and "commercial services" are used as equivalent expressions in this chapter. 3 It should be noted that the bilateral trade figures reported by each of the two countries may show somewhat different flows. These discrepancies may be largely attributable to the fact that imports of services are often more difficult to record than exports and to methodological differences in the way individ- ual countries compile the data. 4 For methodological explanations concerning the compilation of these data, see Karsenty (2000); the data presented therein were updated from IMF balance of payments statistics, and estimates of foreign affiliate trade statistics (FATS) gross output were extrapolated using value added in services growth rates implicitly derived from GDP aggregates. 5 WTO, "Recent Developments in Services Trade: Overview and Assessment," 1999: p. 4. 246 26 R O B E R T M . S T E R N Quantifying consumption. Thus, FDI may be necessary in order to establish a foreign commercial presence, Barriers to and temporary cross-border movement of labor may also be Trade in Services required to serve foreign con- sumers. Types of Barriers to Trade and FDI in Services As noted by Hoekman and Braga (1997: 288), as a consequence of G the simultaneity of the produc- iven the significant growth in recent tion and consumption of services, border measures decades of international trade in such as tariffs will generally be difficult to apply services (now equal to about 20 percent of global because customs agents cannot readily observe the merchandise trade), it is obviously important to service as it crosses the border. Typically, therefore, consider the barriers that affect this trade and the the restrictive policies followed will be designed to issues concerning measurement of the barriers. The limit the access of foreign services and services sup- amount of foreign direct investment (FDI) in both pliers to domestic markets. Hoekman and Braga dis- goods and services sectors in and between industri- tinguish the following types of barriers: (a) al and developing countries is also increasing. FDI, quantitative restrictions such as quotas, local con- like services, is subject to a variety of barriers and tent, and prohibitions; (b) price-based instruments; merits attention in its own right. (c) standards, licensing, and procurement; and (d) The most distinguishing characteristic of services discriminatory access to distribution networks. transactions is that their production and consump- tion occur simultaneously, often requiring direct · Quantitative-restriction (QR)­type policies are contact between producers and consumers. commonly applied to services providers. Two Although some services (for example, "separated" prominent examples are bilateral agreements reg- services such as telecommunications) are traded ulating international air transport services, which internationally across borders in a manner similar are usually reciprocal and company-specific, and to cross-border trade in goods, other services may ocean-cargo-sharing arrangements, which also require the consumer to move to the location of the often rely on reciprocity in providing shipping producer, as in the case of tourism. Furthermore, services in mutual trade. Many countries have because of the necessary proximity of consumers outright prohibitions directed against foreign and producers, factors of production may have to providers of such services as domestic transport, move across national boundaries to the place of basic telecommunications, and legal, insurance, 247 D E V E L O P I N G C O U N T R I E S A N D N E G O T I AT I O N S O N T R A D E I N S E R V I C E S education, surveying, and investment-advising then the categories should reflect the key ele- services. Restrictions on transborder data flows ment of the commitment. . . . If the primary are also prevalent and may impede market access interest is instead the resource allocation impli- by foreign providers. cations of the barriers, some additional or dif- · Price-based instruments may take the form of visa ferent information may be useful. fees and entry or exit taxes, discriminatory airline Barriers to FDI may distort international landing fees, and port taxes. Tariffs can be signif- patterns and modes of . . . trade. They may also icant barriers to trade in goods that embody ser- distort allocation of capital between different vices (such as films, television programs, and economies, between foreign and domestic computer software) or in goods that are used in investment, between different sectors, and producing services (for example, computers, between portfolio and direct investment. . . . telecommunications equipment, and advertising [T]he classification system . . . should highlight materials). Furthermore, many services sectors the key characteristics of the barriers that will are subject to government-sanctioned or govern- determine their size and impact. Market access ment-monitored price controls; examples and national treatment are . . . relevant cate- include air transport, financial services, and gories from a resource allocation perspective. telecommunications. Government subsidies are . . . national treatment is generally taken to refer commonly used in such services sectors as con- to measures affecting firms after establishment. struction, communications, and road and rail A . . . way to classify barriers is therefore . . . transport. according to what aspect of the investment they · Licensing or certification requirements may be most affect: establishment, ownership and con- imposed on foreign providers of professional and trol; or operations. In addition . . . some further business services. Environmental standards may information may be useful . . . on distinctions also affect services providers, particularly in . . . between direct versus indirect restrictions transport and tourism. Government procurement on foreign controlled firms; and rules versus policies are often designed to favor domestic over case-by-case decisions." (Hardin and Holmes foreign providers of services, as well as goods, by 1997: 33­34)1 means of preference margins and outright prohi- bitions. The main types of existing FDI barriers are iden- · Discriminatory access to distribution and commu- tified in UNCTAD (1996). Further information on nications systems prevails in many countries in the barriers most commonly used to restrict FDI, such sectors as telecommunications, air trans- especially in the Asia-Pacific Economic Coopera- port, advertising, insurance, and dealer networks. tion (APEC) economies, is provided in Hardin and Holmes (1997, esp. 37­40, 45­55). As these authors Hardin and Holmes (1997) have focused specifi- note, some common characteristics appear to be: cally on barriers to FDI. They define an FDI barrier as "any government policy measure which distorts application of some form of screening or regis- decisions about where to invest and in what form. . . . tration process involving various degrees of [P]olicy measures such as limits on the level of for- burden for the foreign investor; restrictions on eign investment, or the need to go through costly and the level or share of foreign ownership, particu- time-consuming screening processes to convince larly in some service sectors, and often in the authorities that FDI in a project will be in the nation- context of privatization; widespread use of al interest, are considered barriers" (Hardin and case-by-case judgments, often based on nation- Holmes 1997: 24). al interest criteria; widespread use of restric- In considering ways of classifying FDI barriers, tions on ownership and control (e.g., the authors note: restrictions on board membership), particular- ly in sectors such as telecommunications, The appropriate classification system may vary, broadcasting, banking; and relatively limited depending on the purpose of the exercise. For use of performance requirements on input con- example, if the purpose is to check and monitor trols in services sectors. (Hardin and Holmes compliance with some policy commitment, 1997: 40)2 248 Quantifying Barriers to Trade in Services Measurement of Barriers to Trade and Using the value of output by sector for a representa- FDI in Services tive industrial country, it is then possible to con- struct weighted-average measures by sector and The measurement of barriers that affect services country. The weighted-average tariff equivalents for parallels to an extent the measurement of nontariff one-digit International Standard Industrial Classifi- barriers (NTBs) that limit trade in goods. Services cation (ISIC) sectors for selected countries are indi- barriers, however, involve greater complexities cated in Table 26.1. when account is taken of the different modes of It should be emphasized that Hoekman's mea- supply of services, which include not only cross- sures are designed to indicate the relative degree of border trade but also the movement of consumers restriction and are not to be taken literally as indica- to the location of providers, FDI, and temporary tors of absolute ad valorem tariff equivalents. But international movement of labor. even granting this, there are some important limita- tions worth mentioning. Thus, as Hardin and Holmes (1997: 72) note, Hoekman's method may be Frequency Measures misleading or biased because it assumes that the Frequency measures of services barriers have been absence of positive country commitments in the constructed by PECC (1995) and Hoekman (1995, GATS schedules can be interpreted as indicating the 1996). They are based on the information contained presence of restrictions--which may in fact not be in the country schedules of the General Agreement the case. Furthermore, the different types of restric- on Trade in Services (GATS) and refer to all four tions are given equal weight and are not distin- modes of services supply. Frequency ratios are con- guished according to their economic impact. structed on the basis of the number of commitments Finally, market access restrictions are the only ones scheduled in the GATS by individual countries that taken into account. designate sectors or subsectors as unrestricted or Hardin and Holmes (1997) attempted to build on partially restricted in relation to the maximum pos- and improve Hoekman's methodology for con- sible number of unrestricted commitments. The 155 structing "guesstimates" of tariff equivalents for ser- sectors and subsectors, 4 modes of supply, and 2 vices. In particular, they use information on actual commitment categories yield 1,240 total commit- FDI restrictions and incorporate information on ments on market access and national treatment for the types of barriers and their economic impact. each of 97 countries (Hoekman 1996: 101).3 Their objective is to construct an index of FDI Hoekman (1996) used frequency ratios to restrictions that can be translated into a tariff equiv- approximate the relative degree of restrictiveness of alent or tax equivalent. They specify (p. 76) the market access barriers to services trade across coun- components of their proposed index and the tries and sectors. The author established a judgmen- weights assigned to the different subcategories; the tal set of benchmark tariff equivalents for weights are set to reflect the relative degree of individual sectors to reflect the degree to which restriction of different barriers. Their calculations market access to these sectors was restricted. The of the indexes of FDI restrictions are confined to the benchmark tariff equivalents chosen ranged from a services sectors in the APEC countries.4 The results high of 200 percent for sectors, such as maritime are summarized in Table 26.2. It is evident that cabotage (coastal domestic shipping), air transport, communications and financial services are most postal services, voice telecommunications, and life subject to FDI restrictions, while business, distribu- insurance, in which market access was essentially tion, environmental, and recreational services are prohibited in most countries, to 20­50 percent for the least restricted. sectors in which market access was less constrained. Warren and Findlay (2000) reviewed a number of He then assigned a value to each country and sector studies of services sectors based on the use of fre- using the benchmarks multiplied by the calculated quency-type indices of services impediments.5 These frequency ratio. Thus, for example, assuming a studies include analyses of market access commit- benchmark tariff equivalent of 200 for postal ser- ments in financial services, covering direct insurance vices and a frequency ratio of 50 percent to reflect and banking (Mattoo 1999); a study of impediments the scheduled market access commitments, the tar- in financial services, including banking, securities, iff equivalent for that sector is set at 100 percent. and insurance, in Australia and in selected Asian 249 D E V E L O P I N G C O U N T R I E S A N D N E G O T I AT I O N S O N T R A D E I N S E R V I C E S Table 26.1 Constructed Ad Valorem Tariff Equivalent "Guesstimates" by One-Digit ISIC Services Sector, Selected Countries (percentage) ISIC 6 ISIC 7 ISIC 8 ISIC 9 (wholesale (transport, (business (social and retail storage, and and ISIC 5 distribu- and com- financial personal Economy (construction) tion) munications) services) services) Australia 12.0 7.4 183.4 24.8 25.4 Austria 5.0 4.6 98.7 20.1 13.9 Canada 6.0 9.0 117.7 25.9 40.2 Chile 40.0 34.4 182.2 45.2 42.9 European Union 10.0 10.0 182.0 27.2 23.6 Finland 19.0 14.6 181.0 23.8 31.7 Hong Kong (China) 32.0 31.5 149.8 39.0 42.9 Japan 5.0 4.6 142.0 28.9 32.3 Korea, Rep. of 16.0 21.4 164.9 36.3 40.7 Mexico 24.0 21.3 152.3 40.9 29.8 New Zealand 5.0 13.4 181.5 30.5 36.1 Norway 5.0 13.4 122.2 25.7 24.0 Singapore 12.0 34.4 138.8 35.9 33.7 Sweden 12.0 13.4 184.2 22.5 26.9 Switzerland 5.0 8.0 178.1 27.7 32.3 Turkey 5.0 34.4 31.6 35.4 35.9 United States 5.0 4.6 111.4 21.7 31.7 Note: ISIC, International Standard Industrial Classification. Source: Hoekman (1995): 355­56. Table 26.2 FDI Restrictiveness Indices, Selected APEC Economies and Selected Services Sectors (percent) Economy Business Communication Distribution Education Financial Transport Australia 18 44 18 18 45 20 Canada 23 51 20 20 38 24 China 36 82 28 53 45 46 Hong Kong (China) 2 35 5 0 23 9 Indonesia 56 64 53 53 55 53 Japan 6 35 5 20 36 11 Korea, Rep. of 57 69 63 55 88 57 Malaysia 32 42 8 8 61 12 Mexico 29 74 33 45 55 28 New Zealand 9 43 8 8 20 13 Papua New Guinea 30 48 30 30 30 30 Philippines 48 76 48 48 95 98 Singapore 26 52 25 25 38 25 Thailand 78 84 78 78 88 78 United States 1 35 0 0 20 3 Note: APEC, Asia-Pacific Economic Cooperation. The higher the score, the greater the degree to which an industry is restricted. The maxi- mum score is 100 percent. Source: Adapted from Hardin and Holmes (1997): 112­37. 250 Quantifying Barriers to Trade in Services economies (McGuire 1998); an analysis of impedi- the Philippines, Singapore, and Thailand. Argenti- ments to trade in banking services for 23 countries na, Australia, Canada, the EU, Hong Kong (China), plus the European Union (EU) that distinguished Switzerland, and the United States appear to have impediments to commercial presence and operations relatively low nonprudential restrictions on foreign and impediments affecting foreign banks and all banks. Warren and Findlay also discuss (p. 73) work banks (McGuire and Schuele 2001); a pilot study of in progress on price-based measures of policy vari- barriers affecting accounting services in Australia, ables for maritime services. Calculations of price- France, the United Kingdom, and the United States based measures are especially useful because they (OECD 1997a); an analysis of the commitments of are derived from observed data. They are, accord- 69 signatories of the February 1997 Agreement on ingly, well suited for use in economic models Basic Telecommunications (Marko 1998); indices for designed to assess the effects of banking restrictions measuring the restrictiveness of telecommunications on resource allocation and economic welfare. A policies in 136 countries (Warren 2001a); and restric- recent example is the multicountry model of Dee tions on foreign maritime services suppliers and all and Hanslow (2001), who use the estimates by Kali- maritime service suppliers, covering 35 economies rajan and others. (McGuire, Schuele, and Smith 2001). The various frequency-type measures that have Quantity-Based Measures of Services Barriers been constructed are useful in identifying the types of barrier and the relative degree of protec- Warren (2001b) has assessed the quantitative tion afforded to particular sectors across coun- impact of barriers in telecommunications services, tries. But these measures have only limited chiefly mobile telephony and fixed network ser- economic content with regard to assessing the size vices, for 136 countries. Combining the quantitative of service barriers and the consequences of main- estimates of the effects of removing existing barriers taining or eliminating these barriers. As indicated with an estimate of the price elasticity of demand in Table 26.5, below, Hoekman's guesstimates of for the telecommunications services involved, he services tariff equivalents, in particular, have been calculated tariff equivalents in the form of price used in a number of economic modeling studies wedges. He shows that the tariff equivalents for to represent actual barriers. The results of these domestic and foreign providers of telecommunica- studies are problematic, for the reason mentioned tion services for the advanced industrial countries above. The question, then, is whether it is possible are relatively low in comparison with the much to construct price-based or quantity-based meas- higher estimates for the newly industrializing coun- ures of services barriers that can be used for quan- tries and developing countries. His estimates have titative assessment of the costs of and benefits been used in the economic modeling work by Dee from the reduction or removal of the barriers. I and Hanslow already cited. next discuss a number of such recent measure- Francois (1999) fitted a gravity model to bilateral ment efforts. services trade for the United States and its major trading partners, taking Hong Kong (China) and Singapore as free trade benchmarks. He interprets Price-Based Measures of Services Barriers the differences between actual and predicted Warren and Findlay (2000) review ongoing efforts imports as indicative of NTBs and then normalizes to construct price-based measures of services them relative to the free trade benchmarks for Hong impediments, using estimates of price-cost mar- Kong and Singapore, which presumably have few gins. As they note, Kalirajan and others (2001) have barriers. The results for business and financial ser- calculated the "net interest margins" for 694 nation- vices and for construction are shown in Table 26.3. al and state commercial banks in selected As noted in Deardorff and Stern (1998: 24), gravity- economies. The latter authors' "price-wedge" calcu- model measures of this kind are useful mainly in lations distinguish barriers to establishment and to identifying relative levels of protection across sec- ongoing operations for foreign and domestic firms. tors and countries. They have, however, some Kalirajan and others show that the price impacts of important drawbacks: the attribution to NTBs of all restrictions on foreign banks are the highest for departures of trade from what the included vari- Chile, Indonesia, the Republic of Korea, Malaysia, ables can explain places a great burden on the 251 D E V E L O P I N G C O U N T R I E S A N D N E G O T I AT I O N S O N T R A D E I N S E R V I C E S Table 26.3 Estimated Tariff Equivalents in Traded Services: Gravity Model­Based Regression Method (percent) Business and Economy or region financial services Construction North Americaa 8.2 9.8 Western Europe 8.5 18.3 Australia and New Zealand 6.9 24.4 Japan 19.7 29.7 China 18.8 40.9 Taiwan (China) 2.6 5.3 Other newly industrialized countries 2.1 10.3 Indonesia 6.8 9.6 Other Southeast Asia 5.0 17.7 India 13.1 61.6 Other South Asia b 20.4 46.3 Brazil 35.7 57.2 Other Latin America 4.7 26.0 Turkey b 20.4 46.3 Other Middle East and North Africa 4.0 9.5 Central and Eastern Europe and Russia 18.4 51.9 South Africa 15.7 42.1 Other Sub-Saharan Africa 0.3 11.1 Rest of world 20.4 46.3 a. North American values are derived by assigning numbers for Canada and Mexico to the United States. b. Turkey and Other South Asia are not separately available in the U.S. data and have been assigned estimated rest-of-world values. Source: Francois (1999). model being used. The more imperfect the model, In general, a large number of factors will deter- the more likely it is that NTB estimates will have an mine the ability of firms to generate high mar- upward bias. Moreover, since trade cannot be pre- gins, including market size (number of firms), dicted accurately for particular industries and the business cycle, the state of competition pol- countries, it is not clear how the deviations should icy enforcement, the substitutability of prod- be interpreted or to what extent existing trade pat- ucts, fixed costs, etc. Notwithstanding the terns depart from free trade. As a consequence, the impossibility of inferring that high margins are results from modeling studies such as that by Hertel due to high barriers, there should be a correla- (2000), who used Francois's estimates of services tion between the two across countries for any barriers, pose problems of interpretation. given sector. Data on operating margins pro- vide some sense of the relative profitability of activities, and therefore, the relative magnitude Financial-Based Measures of Services Barriers (restrictiveness) of barriers to entry/exit that Hoekman (2000) has suggested that financial data may exist. (Hoekman 2000: 37) on gross operating margins calculated by sector and country may provide information about the effects Table 26.4 shows the results for 1994­96 by econo- of government policies on firm entry and condi- my and region, averaged over firms and sectors, for tions of competition. He points out, agriculture, manufacturing, and services. Services 252 Quantifying Barriers to Trade in Services Table 26.4 Average Gross Operating Margins of Firms Listed on National Stock Exchanges, 1994­96, by Economy or Region (percent) Economy or region Agriculture Manufacturing Services Australia 8.4 15.5 16.6 Canada 32.1 22.6 32.9 Chile 39.1 40.8 44.0 China 30.6 28.1 49.5 European Union 22.9 23.8 31.6 Hong Kong (China) 25.9 12.8 18.1 Indonesia 41.8 34.3 41.3 Japan 38.4 26.4 28.7 Korea, Rep. of 11.2 25.7 25.8 Malaysia 22.6 6.0 21.6 Mexico 38.4 39.3 37.2 New Zealand 33.3 16.6 26.8 Philippines 18.1 28.6 42.3 Singapore 0.0 11.1 22.0 Taiwan (China) 19.6 25.1 41.3 Thailand 38.2 27.3 52.6 United States 36.6 21.2 42.3 Rest of Cairns Groupa 36.3 31.1 39.0 a. Includes Argentina, Brazil, and Colombia. Source: Hoekman (2000), based on calculations using Worldscope (1998) data. margins are generally higher than manufacturing need to be interpreted with care, however, because margins by 10­15 percentage points, and they vary they are indirect measures and do not make considerably across economies. Australia, Hong allowance for intercountry differences in the quality Kong (China), and Singapore have the lowest ser- and variety of services. Nonetheless, these measures vices margins, around 20 percent, while Chile, China, are useful as a first approximation of the cost-rais- Indonesia, the Philippines, Taiwan (China), Thai- ing effects of services, and they can be incorporated land, and the United States have services margins of into economic models, as has been done, for exam- more than 40 percent. The sectoral results (Hoekman ple, in Brown and Stern (2001). 2000: 38) indicate that the margins for hotels and financial services are relatively high and the margins Modeling the Economic Effects of Services for wholesale and retail trade are lower. The margins Barriers for several developing countries appear to be relative- ly high in a number of services sectors. Overall, While the various measures of services barriers noted Hoekman suggests that"business services, consultan- are of interest, they need to be incorporated into an cy, and distribution do not appear to be among the explicit economic modeling framework in order to most protected sectors. . . . barriers to competition determine how the existence or removal of the barri- are higher in transportation, finance, and telecom- ers will affect conditions for competition, production munications. These are also basic `backbone' imports costs, economic welfare, and the intersectoral move- that are crucial for the ability of enterprises to com- ment of capital and labor. Most research to date on pete internationally" (Hoekman 2000: 39). the modeling of barriers has been focused on inter- Financial-based measures of services barriers are national trade in goods rather than on trade in ser- especially promising because they can be construct- vices and FDI. The reasons stem in large part from ed for a large number of sectors and countries. They the lack of comprehensive data on cross-border ser- 253 D E V E L O P I N G C O U N T R I E S A N D N E G O T I AT I O N S O N T R A D E I N S E R V I C E S vices trade and FDI and on the associated barriers, · Assumption that FDI responds to trade liberaliza- together with the difficult conceptual problems of tion or other exogenous changes that generate modeling that are encountered. Some indication of international capital flows in response to changes pertinent modeling work is provided in Table 26.5. in rates of return (Martin and Yagashima 1993; Following Hardin and Holmes (1997: 85), the Dee, Geisler, and Watts (1996); Donovan and Mai approaches to modeling can be divided as follows: 1996; McKibbin and Wilcoxen 1996; Bora and Guisinger 1997; Adams 1998; Dee, Hardin, and · Analysis of services trade liberalization in response Schuele 1998)6 to reductions in services barriers (Brown, Dear- · Modeling of links between parents and affiliates dorff, and Stern 1996; Brown and others 1996; and distinctions between foreign and domestic Francois and others 1996; Hertel 2000; Tamms firms in a given economy or region (Markusen, 2001; Robinson, Wang, and Martin 2002) Rutherford, and Hunter 1995; Petri 1997; Ben- Table 26.5 Alternative Approaches to Modeling the Impact of Barriers to Trade and Investment 1. Reduction in services barriers Brown, Deardorff, and Stern (1996); Brown and others (1996) Based on 8-region, 29-sector, 1990 reference year version of Michigan CGE model, with all goods and ser- vices tradable. Uses Hoekman's (1995) "guesstimates" of tariff equivalents covering all modes of providing services, including FDI. Factors involved in FDI assumed to be part of factor markets in country of origin. Francois and others (1996) CGE analysis using 1989 reference year and calculation of the effects of price wedges attributable to the Jones Act, which restricts trade in U.S. domestic water transport (cabotage) services. Australia, Department of Foreign Affairs and Trade (1999) Based on GTAP model, version 4 (1995) database covering 45 regions and 50 sectors in each region and on the Asia-Pacific G-cubed model with 18 regions and 6 sectors, with inclusion of a financial sector and full (dynamic) macroeconomic closure. Uses modifications of Hoekman's (1995) "guesstimates" of ser- vices tariff equivalents. Hertel (2000) Based on 19-region, 22-sector CGE model with GTAP 1995 reference year data projected to 2005. Post­Uruguay Round tariff rates are used for agriculture and manufactures. Barriers for business services and construction based on gravity-model estimates in Francois (1999). Tamms (2000) Constructs cost functions using data for 50 airlines from 27 countries for 1982­95 and estimates a frontier function to determine the extent to which an airline lies off its frontier. Robinson, Wang, and Martin (2002) Based on 10-region, 11-sector, 1995 reference year CGE model, with all goods and services tradable. Uses Hoekman's (1995) "guesstimates" of services tariff equivalents, with allowance for growth in total factor productivity (TFP) stimulated by imports of services by developing countries. 2. Flows of FDI in response to changes in rates of return Martin and Yagashima (1993) Analysis of trade liberalization in Asia and Pacific region coupled with assumed changes in inward FDI. Dee, Geisler, and Watts (1996) Based on 13-region, 4-sector, 1992 reference year CGE model, with all goods and services tradable; monopolistic competition in the resources, food processing, and manufacturing sectors; allowance for capital accumulation and international factor mobility. Uses Hoekman's (1995) "guesstimates" of services tariff equivalents. 254 Quantifying Barriers to Trade in Services Table 26.5 (continued) Donovan and Mai (1996) Uses MEGABARE model to estimate effects of trade liberalization with varying degrees of international capital mobility in response to differential rates of return on investment. McKibbin and Wilcoxen (1996) Uses G-cubed model with international capital mobility responding to changes in differential sectoral rates of return to capital. Bora and Guisinger (1997) Analysis of investment liberalization in APEC, with allowance for international capital mobility. Adams (1998) Based on GTAP model, with 14 regions and 37 perfectly competitive sectors, and on post-NAFTA data- base. Each region contributes a fixed proportion of its income to a global savings pool. Investment alloca- tion depends on relative rates of return. Focus is on effects of trade liberalization in APEC. FDI is not modeled explicitly. Dee, Hardin, and Schuele (1998) Based on same model and data as Dee, Geisler, and Watts (1996), with analysis of APEC sectors selected for "early voluntary sectoral liberalization." 3. Links between parents and foreign affiliates and distinctions between foreign and domestic firms Markusen, Rutherford, and Hunter (1995) Analysis of trade liberalization in the automobile industry in the NAFTA countries, using a model with multinational firms or national firms responding to changes in their market shares. Petri (1997) Based on 6-region, 3-sector CGE model, using 1992 GTAP dataset, with FDI separated into activities of domestic and foreign-owned firms. Products differentiated by both country of ownership and place of production. Capital allocation between sectors and between domestic and foreign investments responds to changes in rates of return and to investor preferences. Barriers to FDI modeled as a tax on FDI profits. Markusen, Rutherford, and Tarr (2000) Conceptual static and dynamic CGE model used to analyze how inward FDI in producer services may complement domestic skilled labor, affect the pattern of trade in goods, and determine the characteristics of the dynamic adjustment path. Benjamin and Diao (2000) Based on 10-region, 11-sector CGE model, using data for the early 1990s, with the focus on liberalization of cross-border trade of other private services in APEC. Services providers in the single services sector are imperfectly competitive, have fixed costs, and are able to price-discriminate across countries. Liberaliza- tion is modeled as both reducing fixed costs and removing the market segmentation that permits price discrimination. FDI is not modeled explicitly, but it could be. Dee and Hanslow (2001) Based on 19-region, 3-sector, 1995 reference year CGE model, with modifications of Petri's (1997) framework and updating of data on FDI stocks, output, and rates of return. Uses averages of services barriers for banking and telecommunications services contained in Kalirajan and others (2001) and in Warren (2001b). Brown and Stern (2001) Based on 18-region, 3-sector, 1995 reference year CGE model with features of structure and FDI similar to Dee and Hanslow (2001). Uses estimates of price-cost margins from Hoekman (2000) to estimate services barriers. Note: APEC, Asia-Pacific Economic Cooperation; CGE, computable general equilibrium (model); FDI, foreign direct investment; GTAP, Global Trade and Analysis Project; NAFTA, North American Free Trade Agreement. Source: Author's compilation. 255 D E V E L O P I N G C O U N T R I E S A N D N E G O T I AT I O N S O N T R A D E I N S E R V I C E S jamin and Diao 2000; Markusen, Rutherford, and declines as their resources are shifted out of agricul- Tarr 2000; Brown and Stern 2001; Dee and ture with the decrease in protection. The reduction Hanslow 2001). in barriers on mining and manufactured goods in Scenario B provides the greatest source of the wel- The third group of studies listed in Table 26.5 fare gain of US$141.2 billion, with Japan and the EU comes closest to capturing the important role played benefiting especially. There are also sizable welfare especially by multinational corporations (MNCs) gains for the Asian developing economies, particu- and their foreign affiliates in providing services. larly Taiwan (China), Korea, and China. For services Thus, for example, in the Brown and Stern (2001) liberalization (Scenario C), global welfare increases modeling study, each MNC is assumed to produce a by US$42.4 billion. Except for Japan, all the indus- differentiated product and to allocate production to trial countries show welfare increases, as do most of its various host-country locations. Firms employ the Asian developing countries. The welfare declines capital, labor, and intermediate inputs in produc- noted for services liberalization are associated pri- tion, and they set prices as an optimal markup of marily with whether a country attracts or loses cap- price over marginal cost. Consumers are assumed to ital as a result of liberalization. allocate their expenditure between goods and ser- Scenario D, which combines the welfare effects for vices that are produced by firms domestically and Scenarios A, B, and C, yields an increase in global those that are imported from each national source. welfare of US$193.2 billion. In this overall scenario, Labor is taken to be freely mobile between domestic the return to capital increases, making it likely that, sectors but not across borders. Capital, however, is over time, there will be an increase in world capital mobile internationally, although not perfectly so stock. To take this into account, Brown and Stern because there is a risk premium that varies depend- allow for an increase of 2 percent in world capital ing on the size of a country's capital stock. Barriers stock, which is the amount necessary to hold the real to FDI are assumed to take the form of an increased return to capital equal to its level in the base period. fixed cost of locating investment in a host country. The results, shown in Scenario E, indicate that the For this purpose, Brown and Stern use the cost-price welfare effects are now positive for all countries in margins estimated by Hoekman (2000) that are list- the model. For the world as a whole, welfare rises by ed in Table 26.4 as indicative of barriers to FDI. Since US$612.4 billion. The welfare increases are substan- the cost-price gap is smallest in most services sectors tial both in absolute terms and as a percentage of in Hong Kong (China), which is thought to be freely GNP in both industrial and developing countries. It open to foreign firms, the excess over the Hong Kong appears, therefore, that capital formation can play a figure in any other economy in the Brown and Stern far more important and substantive role than the model is taken to be attributable to barriers to the removal of the more traditional efficiency distor- establishment of foreign firms. tions caused by trade barriers in determining the Using a modeling structure with three sectors welfare effects of trade liberalization. (agriculture, manufactures, and services) and 18 economies or regions, Brown and Stern calculate Implications for Research and Policy the economic effects of an assumed 33 percent reduction in tariff barriers on agriculture and man- It should be evident from the preceding discussion ufactures and in barriers to establishment of foreign that most existing measures of services barriers are firms providing services.7 When barriers are low- subject to some uncertainty. The reason is that they ered, international capital in the form of FDI will be are indirect measures of the price-quantity dimen- attracted to the countries with the relatively highest sions involved, since services barriers do not readily rates of return and away from other countries. lend themselves to direct measurement in the same The welfare effects of the assumed 33 percent ways that tariffs do. The most promising measures of reduction in barriers are indicated in Table 26.6. For services barriers are price and quantity measures agricultural liberalization (Scenario A), global wel- based on detailed and careful analysis, especially at the fare rises by US$20.9 billion, with the biggest gain- sectoral level, such as the work being done under the ers being the United States, Australia, and Canada, auspices of the Australian Productivity Commission. as resources are shifted into their agricultural sec- Further refinement of financial-based measures of tors. Japan, in particular, and the EU show welfare cost-price margins also merits continued attention. 256 E US$ 6.9 8.5 4.4 4.2 7.7 1.4 3.4 Billion 20.8 33.8 80.2 28.3 10.3 14.4 18.5 28.8 11.5 including increase stock) 151.0 178.4 612.4 cent Scenario capital sectors, per in GNP 6.1 5.9 1.9 1.6 2.5 4.0 7.3 4.2 9.1 6.3 7.3 4.8 2.2 1.2 1.1 2 centage 12.0 10.2 10.5 of (all Per 3.1 5.7 5.0 9.8 7.2 6.6 1.4 3.0 1.0 US$ 12.8 26.8 41.6 45.8 15.4 27.0 ­0.5 ­2.9 D Billion 15.7­ 193.2 sectors) vices Scenario (all GNP 3.7 4.7 0.5 0.1 0.6 2.2 5.0 5.0 1.6 7.1 2.1 5.2 9.8 0.6 centage 10.0 ­0.7 ­1.0 ­1.5 Ser of Per and C US$ 2.2 1.5 9.6 9.0 1.9 8.0 0.7 0.5 0.4 6.2 Billion 23.6 14.1 ­2.9 ­0.9 ­0.3 ­2.7 42.4 ­28.5 ­11.2 vices) Manufactures, Scenario (ser GNP centage 0.6 4.2 0.2 2.6 0.1 1.3 1.9 4.1 0.7 0.7 0.7 2.2 ­0.6 ­0.6 ­0.5 ­0.4 ­1.0 ­1.1 of Per Agriculture, in US$ 6.0 0.1 4.1 6.1 6.0 3.2 1.6 5.7 0.7 2.4 1.4 Billion 28.3 53.3 10.9 21.2 ­0.2 ­1.7 ­7.9 B 141.2 radeT to Scenario manufactures)( GNP centage 1.8 0.0 0.4 1.0 7.1 0.1 0.8 3.2 0.8 2.4 6.1 1.0 4.2 7.7 0.9 ­0.3 ­0.6 ­0.8 Barriers of Per in US$ 4.7 3.4 0.2 1.1 0.0 0.3 0.3 0.3 0.2 0.4 0.4 0.0 1.3 2.9 Billion ­1.7 A 22.2­ 30.1 ­0.8 20.9 Reduction cent Per Scenario (agriculture) 33 GNP centage 1.4 0.6 0.0 0.4 0.4 0.2 0.1 0.1 0.3 0.5 0.4 0.1 0.3 0.0 0.5 0.3 ­0.4 ­0.2 of of Per Colombia. a Effects and (2001). Group Brazil, elfare (China) of Stern W economies and economies Union Kong Rep. (China) Cairns Argentina, 26.6 States of Zealand Brown China Hong Indonesia Korea, Malaysia Philippines Singapore aiwanT Thailand Chile Mexico Rest ce: ableT Economy Industrial Australia Canada European Japan New United Developing Asia Other otalT Includes a. Sour 257 D E V E L O P I N G C O U N T R I E S A N D N E G O T I AT I O N S O N T R A D E I N S E R V I C E S Because research on services barriers is so time-inten- transparency as against the exercise of administrative discre- sive, international organizations and government tion. agencies are in the best position to undertake it and 2 Hardin and Holmes (1997: 40­43) also provide information on underwrite the costs involved. The private sector can investment incentives, which are widely used and for the most also be helpful in providing specialized information part are not subject to multilateral disciplines. and knowledge about different barriers. 3 As noted in Hardin and Holmes (1997: 70), the GATS commit- Even though measurement of services barriers is ments are based on a "positive list" approach and therefore do imprecise, it does appear that these barriers have not take into account sectors and restrictions that are unsched- uled. In PECC (1995) it is assumed that all unscheduled sectors significant costs. This is attested by the potential and commitments are unrestricted, which would then signifi- gains in economic welfare that modeling studies cantly lower the calculated frequency ratios. It would be useful suggest would be realized if the barriers were to be to determine the accuracy of the PECC assumption. reduced or removed. Continuing research, with the 4 Details on the construction of the indexes and their sensitivity use of models incorporating both cross-border ser- to variations in the restrictive weights are discussed in Hardin vices trade and services-related FDI, by members of and Holmes (1997, esp. 103­11). the academic community and by international and 5 More recent information on studies completed and in progress governmental organizations should be encouraged. in association with the Australian Government Productivity From what we know to date, services liberalization Commission can be accessed at should remain a central objective in the ongoing . WTO services negotiations and as part of a broader 6 For more recent computable general equilibrium (CGE) model- WTO negotiating round. ing studies that focus on issues of international capital mobili- ty, see Ianchovichina, McDougall, and Hertel (1999); Verikos and Hanslow (1999); Walmsley (1999), and selected papers Notes presented at the Third Annual Conference on Global Economic Analysis, Sydney, Australia, June 2000; available at 1 Direct restrictions include limitations on the total size or share . of investment in a sector and requirements concernng inputs used (e.g., local content). Indirect restrictions include net ben- 7 See also Dee and Hanslow (2001) for computational results efit or national interest criteria and limitations on membership based on a related modeling framework and using estimates of of company boards. The distinction between rules and case- services barriers taken from Kalirajan and others (2001) and by-case decisions relates to issues of clarity in specification and Warren and Findlay (2000). 258 27 R U D O L F A D L U N G A N T O N I A C A R Z E N I G A B E R N A R D H O E K M A N M A S A M I C H I K O N O A A D I T YA M AT T O O L E E T U T H I L L The GATS: developing a framework of rules for policies related to trade in services. Key Features and Sectors General Principles The General Agreement on Trade in Services (GATS) that emerged consists of two main elements: (a) a set of general concepts, principles, and rules that apply to all measures affect- ing trade in services, and (b) T specific commitments that apply he first two sections of this chap- to the services sectors and subsectors listed in each ter present a broad overview of the member's schedule. The GATS covers all measures General Agreement on Trade in Services (GATS) imposed by members that affect the consumption and the status of negotiations on services. Subse- of services originating in other members (Art. I). quent sections focus on financial services, telecom- The agreement applies to four modes of supply munications, maritime services, and air transport through which services may be exchanged: services. The sections are the work of individual authors, as listed in a note at the end of the chapter. · Mode 1: cross-border supply not requiring the physical movement of supplier or consumer · Mode 2: movement of the consumer to the coun- The Agreement try of the supplier The nonexistence of tariffs on services and the · Mode 3: services sold in the territory of a member importance of regulation greatly complicate the by foreign entities that have established a com- lives of negotiators seeking to agree on the incre- mercial presence mental reduction of barriers to services trade. · Mode 4: provision of services requiring the tem- Negotiators require a focal point--some tangible porary movement of natural persons. variable enabling parties to set objectives and assess negotiating progress. In merchandise trade negotia- Trade in services in the GATS context therefore tions, the focus is on the value of bilateral trade covers both trade in the balance of payments (or flows and the associated tariff revenues. Lack of data national accounts) sense and local sales by foreign on trade and the complexities associated with iden- affiliates. The GATS does not apply to services sup- tifying and quantifying barriers to trade made this plied in the exercise of government functions. The approach impossible for services. In the Uruguay main provisions of the GATS are summarized in Round negotiators therefore focused primarily on Table 27.1. 259 D E V E L O P I N G C O U N T R I E S A N D N E G O T I AT I O N S O N T R A D E I N S E R V I C E S Table 27.1 Main Provisions of the GATS Article Subject matter I Definition. Trade in services covers all four modes of supply. II Most-favored-nation (MFN) obligation. Option to invoke exemptions on a one-time basis. III Notification and publication. Obligation to create an enquiry point. IV Increasing participation of developing countries. High-income countries to take measures to facilitate trade of developing nations. V Economic integration. Allows for free trade and similar agreements. VI Allows for domestic regulation. Requirements concerning the design and implementation of services sector regulation, including, in particular, qualification requirements. VII Recognition of qualifications, standards, and certification of suppliers. VIII Monopolies and exclusive suppliers. Requires that such entities abide by MFN and specific commitments (Arts. XVI and XVII) and do not abuse their dominant position. IX Business practices. Recognition that business practices may restrict trade. Calls for consul- tations between members on request. XIV General exceptions. Allows measures to achieve noneconomic objectives. XVI Market access. Defines a set of policies that may only be used to restrict market access for a scheduled sector if they are listed in a member's specific commitments. XVII National treatment. Applies in a sector if a commitment to that effect is made and if no limitations or exceptions are listed in a member's schedule. XVIII Allows members to make additional commitments, e.g., regarding qualifications, stan- dards, and licenses XIX Calls for successive negotiations to expand coverage of specific commitments (Arts. XVI and XVII). XXIX States that annexes are an integral part of the GATS. Annexes to the GATS allow for one-time most- ment no less favorable than that it accords to its favored-nation (MFN) exemptions, address the own like services and service suppliers."1 National movement of natural persons, exclude air transport treatment therefore applies only to those services services, define commitments on financial and inscribed in a member's schedule, and then only to telecommunications services, and clarify the poten- the extent that no qualifications or conditions are tial coverage of maritime transport commitments. listed in the schedule. As in the GATT, the core principle of the GATS is A second specific commitment is market access. MFN (Art. II). Members are, however, allowed to Article XVI stipulates a range of measures restrictive list MFN exemptions on entry into force of the of market access (mostly quotas) that a WTO mem- agreement. MFN exemptions are, in principle, to ber cannot maintain or adopt unless specified in its last no longer than 10 years and are subject to nego- schedule. These measures include restrictions on (a) tiation in future trade rounds. number of service suppliers allowed, (b) value of transactions or assets, (c) total quantity of services output, (d) number of natural persons that may be Specific Commitments employed, (e) type of legal entity through which a Article XVII.1 contains the basic national treatment services supplier is permitted to supply a service obligation, which is a so-called specific commit- (for example, in banking, branches versus sub- ment: "In the sectors inscribed in its Schedule, and sidiaries), and (e) participation of foreign capital in subject to any conditions and qualifications set out terms of limits on foreign equity or on the absolute therein, each Member shall accord to services and value of foreign investment. With the exception of service suppliers of any other Member, in respect of (e), the measures covered by Article XVI all take the all measures affecting the supply of services, treat- form of quantitative restrictions. 260 The GATS: Key Features and Sectors Article XVI has been interpreted, in "Scheduling of example is an "economic needs test." Finally, mem- Initial Commitments in Trade in Services: Explana- bers have the option of making additional commit- tory Note," as applying to both discriminatory and ments by listing actions to be taken that do not fall nondiscriminatory measures.2 It thus covers both under national treatment or market access. Article measures of the type "only five new foreign banks XVIII of the GATS provides for such additional will be granted licenses" and measures such as "only commitments, stating, "Members may negotiate 10 new foreign and domestic banks will be granted commitments with respect to measures affecting licenses." Although the six types of measure listed trade in services not subject to scheduling under above are in principle prohibited, if a member desires Articles XVI or XVII, including those regarding to maintain one or more of them for a scheduled sec- qualifications, standards or licensing matters. Such tor, it may do so as long as it lists them in its schedule. commitments shall be inscribed in a Member's To a degree, Article XVI is the equivalent of GATT Schedule." An example of the use to which Article Article XI, which prohibits the use of quotas. Specific XVIII has been put is the case of a member making commitments apply only to services sectors listed by an additional commitment in a particular sector to members, subject to whatever qualifications, condi- subscribe to international standards. tions, and limitations are maintained. Since commit- The specific commitments made by members can ments are scheduled by mode of supply as well as by be seen as the outcome of a two-step decision. Each sector, these exceptions may apply either across all member first decides which services sectors will be modes of supply or for a specific mode. subject to the GATS market access and national Members also make horizontal commitments treatment disciplines. It then decides what measures applicable to modes of supply, rather than sector, will be kept in place for that sector which violate that are often restrictive in nature. A common market access or national treatment. Table 27.2 Table 27.2 Format and Example of a Schedule of GATS Commitments Conditions and Conditions and Commitment type and limitations on qualifications on Additional mode of supply market access national treatment commitments Horizontal commitments (across all sectors) 1. Cross-border None. None. 2. Consumption abroad Unbound. Unbound. 3. Commercial presence Maximum foreign Unbound for subsidies. (foreign direct investment) equity stake is 49 Approval required for percent. equity stakes over 25 percent. 4. Temporary entry of Unbound except for Unbound except for natural persons intracorporate transfers categories listed in the of senior managers. market access column. Specific commitments (sectoral) 1. Cross-border Commercial presence Unbound. required. 2. Consumption abroad None. None. 3. Commercial presence 25 percent of manage- Unbound. Establishment of an (foreign direct investment) ment to be nationals. independent regulator. 4. Temporary entry of Unbound, except as Unbound, except natural persons indicated under Hori- as indicated under zontal commitments. Horizontal commit- ments. Note: "None" implies that no exceptions are maintained (a bound commitment). 261 D E V E L O P I N G C O U N T R I E S A N D N E G O T I AT I O N S O N T R A D E I N S E R V I C E S illustrates the rather complicated format of sched- Economic Expectations ules of commitments. A consequence of the deci- sions to distinguish between general and specific From an economic perspective, the rationale under- obligations, to schedule specific commitments by lying such provisions may not be immediately evi- mode of supply, and to allow for MFN exemptions dent. Liberal policy bindings under the GATS, not is that much depends on the content of the sched- least under mode 3 (commercial presence), could be ules. The GATS is not a particularly transparent or viewed as a boon rather than a liability for the devel- user-friendly instrument. opment process, as they provide an opportunity to enhance, through multilateral access guarantees, a country's attraction for international investment The Current Pattern of Commitments and the associated gains in skills and expertise. In Pursuant to Article XX.1 of the GATS, each WTO turn, such expectations may have prompted a few member is required to spell out in a schedule the developing countries--in particular, transition specific commitments on market access and nation- economies--not to rely on the agreement's flexibili- al treatment it undertakes in services. The scope and ty but to undertake commitments comparable in substance of schedules are not further specified; for breadth and depth to those undertaken by industrial example, the agreement does not prescribe any country members, if not even more ambitious. It minimum number of sectors to be included or would be inappropriate, nevertheless, to consider modes of supply to be liberalized. Member govern- more hesitant governments insensitive to develop- ments thus have wide discretion in selecting ser- mental needs; rather, they may have preferred a vices from a classification list, which was developed more prudent stance because of lack of experience in the Uruguay Round, and in specifying, by way of with the agreement or a perceived need for internal limitations, trading conditions under any of the legislation to accompany a process of external liber- four modes of supply of services. The range of alization. Although adequate competition rules, lia- scheduling options also includes the possibilities of bility laws, licensing and qualification procedures, departing from the common classification list and the like may be considered prerequisites for (compliance with which is not mandatory), restrict- ensuring effective market opening, the ability to ing access to subregions within the national territo- develop such legislation may depend in turn on a ry, or phasing in commitments at specified later country's level of social and economic development dates ("precommitments"). or the availability of competent technical assistance. Given the leeway provided under the agreement, it may prove difficult, if not impossible, to find two Current Scheduling Patterns identical schedules among the current 140-odd WTO members. Differences in national policy ori- Commitments, by Member. The WTO's current entation, negotiating strength, and sectoral inter- members can be roughly classified into three ests have translated into wide differences in groups, depending on the number of sectors they commitments across members, sectors, and modes. have included in their services schedules (see Table Although it might be tempting to use the term 27.3). About one-third of the membership has "imbalance" in this context, member governments scheduled 20 or fewer sectors of the 160 or so indus- with low levels of commitments would possibly tries specified in the GATS classification list, one- insist that their schedules are a balanced reflection third has committed between 21 and 60 sectors; and of the Uruguay Round process and of domestic the remaining members, about 50 in number, have policy constraints that might preclude liberaliza- included between 61 and about 130 sectors. The last tion of individual areas. Moreover, developing group not only encompasses virtually all industrial countries, which account for some four-fifths of countries but also includes some developing and the WTO's membership, are covered by various least-developed economies (The Gambia, Lesotho, flexibility provisions in the agreement allowing and Sierra Leone). Among the recently acceding them, for example, to open fewer sectors, liberalize members are such countries as the Kyrgyz Republic fewer types of transactions, and progressively and Georgia that have undertaken broader commit- extend access in line with their development situa- ments, in terms of sector coverage, than any tion (Article XIX.2). Uruguay Round participant. 262 The GATS: Key Features and Sectors Table 27.3 Number of Committed Services Sectors by Member, July 2000 Number of committed Number sectors of members WTO members <20 44 Angola, Bahrain, Bangladesh, Belize, Benin, Bolivia, Botswana, Burki- na Faso, Cameroon, Central African Republic, Chad, Congo (Dem. Rep), Congo (Rep.), Djibouti, Fiji, Gabon, Grenada, Guatemala, Guinea, Guinea-Bissau, Guyana, Haiti, Honduras, Madagascar, Mal- dives, Mali, Malta, Mauritania, Mozambique, Myanmar, Namibia, Niger, Paraguay, Rwanda, St. Kitts and Nevis, St. Lucia, St. Vincent and Grenadines, Suriname, Swaziland, Tanzania, Togo, Tunisia, Ugan- da, Zambia 21­60 47 Antigua and Barbuda, Argentina, Barbados, Brazil, Brunei Darussalam, Burundi, Chile, Colombia, Costa Rica, Côte d'Ivoire, Cuba, Cyprus, Dominica, Dominican Republic, Ecuador, Egypt, El Salvador, Ghana, India, Indonesia, Israel, Jamaica, Kenya, Kuwait, Macau (China), Malawi, Mauritius, Mongolia, Morocco, Nicaragua, Nigeria, Pakistan, Papua New Guinea, Peru, Philippines, Poland, Qatar, Romania, Sene- gal, Singapore, Solomon Islands, Sri Lanka, Trinidad and Tobago, Unit- ed Arab Emirates, Uruguay, R. B. de Venezuela, Zimbabwe >61 45 Australia, Bulgaria, Canada, Czech Republic, EU (15), Estonia, The Gambia, Georgia, Hong Kong (China), Hungary, Iceland, Japan, Jor- dan, Korea (Rep.), Kyrgyz Republic, Latvia, Lesotho, Liechtenstein, Malaysia, Mexico, New Zealand, Norway, Panama, Sierra Leone, Slo- vak Republic, Slovenia, South Africa, Switzerland, Thailand, Turkey, United States Source: WTO secretariat. Commitments, by Sector. The positive develop- and communications services, including basic mental expectations that may be associated with telecommunications, are largely influenced by the commitments under the agreement are reflected, to extended negotiations in these areas, which were some degree, in the sector structure of current conducted beyond the time frame of the Uruguay schedules. Among the services most frequently Round and were successfully concluded in Febru- included are not only areas traditionally considered ary 1997 (basic telecommunications) and Decem- to carry low levels of restrictions, such as tourism, ber 1997 (financial services). but also core infrastructural services such as At the bottom of the sector hierarchy are the finance and communication. If the sectors con- health and education sectors, each with fewer than tained in the classification list are aggregated to a 50 inclusions in schedules. Governments may have few large clusters, the pattern illustrated in Figure wanted to retain policy discretion (subject to the 27.1 emerges: tourism ranks first, having drawn MFN principle) in these areas, which may be viewed commitments from all but 10 WTO members in at as core public sector responsibilities. Some mem- least one subsector, followed by financial services bers may also have felt that, given the organization and business services. Commitments in the latter of their countries' health and education systems, two areas have been scheduled, with varying these sectors were beyond the sectoral scope of the breadth and depth, by over 100 members. Commu- agreement. Article I.3 of the GATS provides a gener- nications services (with commitments from slightly al exception for services provided in the exercise of fewer than 100 members), transport, and construc- government authority that are not supplied on a tion services rank next. The results for financial commercial basis or in competition. 263 D E V E L O P I N G C O U N T R I E S A N D N E G O T I AT I O N S O N T R A D E I N S E R V I C E S 27.1 Number of WTO Members Committing to at Least Part of an Aggregate Services Sector, July 2000 Number of countries 140 120 100 80 60 40 20 0 Health Tourism Financial Business Transport Recreation Education Construction Environment Distribution Communications Source: WTO secretariat. Not surprisingly, given the WTO's membership access or on national treatment) and full discretion structure, the above sectoral pattern largely reflects (no market access or national treatment obliga- the scheduling preferences of developing countries. tions). To provide a rough indication of the market In contrast, industrial countries have made com- access commitments bound for individual modes, mitments in nearly all major sectors except for the WTO secretariat has examined the entries for a health and education. (Health services have been sample of about 40 industries deemed representa- omitted by Liechtenstein, New Zealand, Norway, tive of the full services spectrum. (The focus of the and Switzerland, while Canada, Finland, Iceland, study was on market access limitations, as these are and Sweden have committed neither on health nor likely to have a more direct impact on trade than on education services.) At the subsectoral level, national treatment limitations.) Three types of there are further notable exceptions, including the commitments--full, partial, and none--were dis- omission of maritime transport services by the tinguished in this context. United States and the European Union (EU) and of The ensuing picture (Figure 27.2) is very clear in at audiovisual services by Canada and many European least two regards: trade conditions are most liberal countries. Core segments of the air transport sec- for mode 2 (consumption abroad), where close to 50 tor--traffic rights and directly related services--are percent of all commitments are full bindings, and excluded from the application of the agreement, they are most restrictive for mode 4, where virtually pursuant to a special annex that is currently subject all entries carry significant limitations. This basic to review, and thus have not drawn any commit- pattern may be attributed, on the one hand, to gov- ments. ernments' perception that it is not possible in many cases to influence their nationals' behavior once they Commitments, by Mode. For any listed sector, the have left the country and, on the other hand, to the scheduling country specifies the market access and political sensitivities associated with the physical national treatment conditions with regard to each presence of foreigners as services suppliers in domes- of the four modes of supply. A sector commitment tic markets. The situation is less clear for modes 1 is thus made up of eight entries that could range (cross-border trade) and 3 (commercial presence). between full binding (no limitations on market While mode 1 has drawn almost twice as many full 264 The GATS: Key Features and Sectors 27.2 Market Access Commitments by WTO Members on Movement of Natural Persons (Services Mode 4) Percent 100 90 80 70 60 50 40 30 20 10 0 I DT A I DT A I DT A I DT A Mode 1 Mode 2 Mode 3 Mode 4 Full commitments Partial commitments Note: I, industrial economies; DT, developing and transition economies; A, acceding economies. Calculated on the basis of a sample of 37 sectors deemed representative of various services sectors. (See WTO, S/C/W99, March 2, 1999.) commitments as mode 3 (about 30 percent), it also unfettered access (Table 27.3); this is about 20 per- has by far the highest share (about one-third) of non- centage points above the corresponding averages for bindings among all modes. These nonbindings are all members. not necessarily indicative of restrictive policy inten- tions; they might be attributed, in particular in "tra- Financial Services: Past Negotiations and ditional"sectors such as tourism, to the nonfeasibility Issues for the Next Round of cross-border trade. It is nevertheless conceivable in various cases that governments have preferred mode Financial services in the GATS context is divided 3 to mode 1 commitments. Two factors are potential- into two broad categories: insurance and insurance- ly relevant in this respect: the positive employment related services, and banking and other financial and technology effects associated with increased for- services. The first category includes life and nonlife eign commercial presence, and a preference by insurance, reinsurance, insurance intermediation domestic regulators in sensitive areas for the supplier such as brokerage and agency services, and services to be locally established in order to ensure adequate auxiliary to insurance such as consultancy and actu- control. arial services. Banking includes all the traditional In general, the quality of commitments scheduled services provided by banks, such as acceptance of for individual modes does not differ significantly deposits, lending of all types, and payment and between industrial and developing countries. While money transmission services. Other financial ser- developing countries display slightly higher shares vices include trading in foreign exchange, deriva- of full commitments for modes 1 and 3, they have tives, and all kinds of securities; securities listed more nonbindings under mode 2 than indus- underwriting; money brokering; asset management; trial country members. One group of countries settlement and clearing services; provision and stands out, however: the nine developing and tran- transfer of financial information; and advisory and sition economies that joined the WTO between other auxiliary financial services. 1995 and July 2000 not only undertook far broader At the close of the Uruguay Round negotiations in commitments than many Uruguay Round partici- 1993, negotiations on financial services, along with pants but also listed fewer limitations. For example, those on basic telecommunications and maritime over 50 percent of the newly acceding countries' transport, remained unfinished. Specific commit- commitments on market access under mode 1, and ments to provide market access and national treat- more than 70 percent under mode 2, guarantee ment were made for the sector by participants, but 265 D E V E L O P I N G C O U N T R I E S A N D N E G O T I AT I O N S O N T R A D E I N S E R V I C E S they were not considered enough to conclude the member governments and 16 lists of MFN exemp- negotiations. As a result, broad MFN exemptions tions (or amendments thereof) were annexed to the based on reciprocity remained. The Second Annex Fifth Protocol to the GATS, which was open for rat- on Financial Services to the GATS and the Decision ification and acceptance by members until January on Financial Services adopted at the end of the 29, 1999. Fifty-two member governments accepted Uruguay Round provided for extended negotiations the protocol by the due date, and those members to be held during a six-month period following the decided to put the protocol into force on March 1, entry into force of the GATS--that is, up to the end 1999, in accordance with the terms of the protocol. of June 1995. At the conclusion of this period, WTO It was also decided by the Council for Trade in Ser- members had the possibility of improving, modify- vices that the protocol would be kept open for ing, or withdrawing all or part of their commitments, acceptance until June 15, 1999, for the remaining 18 and they were also able to introduce additional MFN members. After this deadline expired, each accept- exemptions. Broad MFN exemptions would not be ing member had to request the council to reopen applied until the end of the period. the protocol for acceptance. As of November 1, 2000, nine Members still had not accepted the Fifth Protocol. The Interim Agreement of 1995 With five countries making commitments in The 1995 negotiations, concluded on July 28, 1995, financial services for the first time, the total number led to an "interim" agreement, since negotiators of WTO members with commitments in financial again decided that the results of the negotiations services (including the newly acceding countries) were not satisfactory. It was agreed that further will increase to around 105 on the entry into force negotiations would commence after two years, in of the Fifth Protocol. As a result of the negotiations, 1997. As a result of the 1995 negotiations, 29 WTO India, Thailand, and the United States decided to members (counting the European Union as one withdraw their broad MFN exemptions based on member) improved their schedules of specific com- reciprocity; only a small number of countries sub- mitments or removed, suspended, or reduced the mitted limited MFN exemptions or maintained scope of their MFN exemptions in financial ser- existing broad MFN exemptions. Several countries, vices, or did both. Those improved commitments including Hungary, Mauritius, the Philippines, and were annexed to the Second Protocol to the GATS. Venezuela, reduced the scope of their MFN exemp- Three other countries, Colombia, Mauritius, and tions. The United States submitted a limited MFN the United States, decided not to improve their exemption in insurance, applicable in a circum- commitments and took broad MFN exemptions stance of forced divestiture of U.S. ownership in based on reciprocity. As a result of the extended insurance services providers operating in WTO negotiations, and with new accessions to the WTO, member countries. 97 members (counting the 15 EU members individ- In financial services, WTO members have the ually) had made commitments in financial services option of adopting the Understanding on Commit- by mid-1997, compared with some 76 countries at ments in Financial Services (a formula for making the end of the Uruguay Round. advanced commitments under the GATS), and 31 WTO members have adopted it. Among other pro- visions, the new commitments introduce significant The 1997 Negotiations improvements concerning commercial presence of The negotiations were reopened in April 1997. foreign financial services suppliers by eliminating or Between November 1 and December 12, 1997, relaxing limitations on foreign ownership of local members again had an opportunity to improve, financial institutions, on the juridical form of com- modify, or withdraw their commitments in finan- mercial presence (branches, subsidiaries, agencies, cial services and to take MFN exemptions in the sec- representative offices, and so on), and on the expan- tor (see the discussion in Box 27.1). As a result of sion of existing operations. Important progress was the negotiations, a new and improved set of com- also made on "grandfathering" existing branches mitments in financial services under the GATS was and subsidiaries of foreign financial institutions agreed to on December 12, 1997. A total of 56 that are wholly owned or majority-owned by for- schedules of commitments representing 70 WTO eigners. Improvements were made in all three major 266 The GATS: Key Features and Sectors B O X 2 7 . 1 F I N A N C I A L S E R V I C E S : M A R K E T A C C E S S C O M M I T M E N T S O F D E V E L O P I N G A N D T R A N S I T I O N E C O N O M I E S The approach of many developing and transition The GATS provoked concern that its rules countries to the negotiations on financial services might compromise the ability of governments to was tentative, for perhaps two reasons. First, the pursue sound regulatory and macroeconomic negotiations were concluded during the East Asian policies or might limit their freedom to achieve financial crisis, in a climate of uncertainty and other domestic policy objectives. Most of these increased awareness of widespread regulatory inad- concerns seem to have been addressed. First, equacies. Second, financial services were being none of the GATS provisions prevent a member negotiated separately from other goods and ser- from taking measures for prudential reasons--for vices, and countries with export interests in other example, to protect investors or depositors or to areas were reluctant to give up negotiating curren- ensure the integrity and stability of the financial cy by making significant commitments. Therefore, system. Second, services supplied in the exercise even though the number of countries that partici- of governmental authority, including activities pated in the eventual agreement was impressive conducted by a central bank or monetary author- (all industrial countries and more than 100 devel- ity or by any other public entity in pursuit of oping and transition economies took part), the lib- monetary or exchange rate policies, are excluded eralizing content of commitments was in many from the scope of the GATS. Third, even though a cases quite limited (see Tables 27.4 and 27.5). member's market access commitments oblige it Interestingly, the few African and Eastern European to allow a certain degree of capital mobility-- participants made much more liberal commit- specifically, when the cross-border movement of ments than the many Asian and Latin American capital is an essential part of the service itself and participants. The Asian countries were more forth- inflows of capital are related to commercial pres- coming in insurance than the Latin American coun- ence--the agreement allows a member to tries, but the converse was true in banking services. impose restrictions on current or capital transac- Two other aspects of the commitments were tions in the event of serious balance of payments somewhat disappointing. First, in many cases or external financial difficulties or the threat there was less emphasis on the introduction of thereof. Finally, the agreement allows a member competition through allowing new entry than on to pursue other domestic policy objectives, allowing (or maintaining) foreign ownership and through, for instance, directed lending programs, protecting the position of foreign incumbents. provided that the measures are neither discrimi- Second, even where it was deemed not feasible natory nor intended to restrict the access of sup- to introduce competition immediately, partici- pliers to a market. pants took little advantage of the GATS to lend credibility to reform programs by precommitting Source: Prepared by the volume editors, based on Mat- to future liberalization. too (2000). financial service sectors--banking, securities, and tiations is to achieve progressively higher levels of insurance--as well as in other services such as asset liberalization of trade in services across all sectors, management and provision and the transfer of including financial services. In these negotiations, financial information. there will be a need for developing countries to rec- ognize the benefits of liberalization of financial ser- vices in helping to develop efficient and robust The Next Round of Negotiations financial markets with foreign capital and expertise. In accord with the mandate contained in Article Building an efficient financial system is a key to eco- XIX of the GATS, negotiations on trade in services nomic growth and development, and liberalization were launched in early 2001. The aim of these nego- of financial services under the GATS will help coun- 267 of and 3) equity Lanka, Number Egypt Brunei Darussalam, India, Malaysia, Pakistan, Philippines, Singapore, Sri Thailand Slovenia (mode suppliers foreign presence cial Foreign equity only Ghana, Kenya Korea Bulgaria Limitations of commer ,y a Rep., Rep. on only Number suppliers Gabon, Mauritius, Morocco Qatar Czech Hungar Slovak the or form Leone, Africa Kong on Gambia, Commitments Full, limitations only legal The Lesotho, Nigeria, Senegal, Sierra South Bahrain, Hong (China), Indonesia, Israel, Macau (China), Solomon Islands, urkeyT Poland, Romania (China), TS Kong Rep., ,y Rep., , GA on 2) Lanka, Limited Ghana, Kenya, unisiaT Brunei Darussalam, Hong (China), Macau Malaysia, Qatar Sri urkeyT Czech Hungar Slovak Slovenia the (mode under Africa Commitments consumption abroad Gambia, Full Nonlife) Egypt, Gabon, The Lesotho, South Bahrain, Solomon Islands, Thailand and (Life (Rep), Rep., ,y Rep., on , 1) nisia Lanka, Limited Egypt, Ghana, Kenya, Nigeria, Tu India, Korea Malaysia, Philippines, Qatar Sri Thailand, urkeyT Bulgaria, Czech Hungar Slovak Slovenia Insurance border (mode on Commitments cross supply Gambia Full Gabon, The Bahrain, Solomon Islands Commitments first modes Full Gambia on The Bahrain, Solomon Islands Access commitments three Market Africa Pacific Region Europe 27.4 and Asia Eastern ableT 268 Dominican Rep., Honduras Cuba, Mexico , , de B. Argentina, Bolivia, Brazil, Chile, Colombia, Ecuador Jamaica, Nicaragua, Peru, Uruguay R. enezuelaV Guyana, Panama, Paraguay Argentina Guyana Argentina, Brazil, Colombia limitations. other or Guyana condition, reciprocity Guyana licensing, y (2000). America discretionar Caribbean Mattoo Latin the ce: Unbound, and a. Sour 269 of and 3) equity (Rep), Gambia, Lanka, Number The Morocco India, Indonesia, Korea Kuwait, Malaysia, Pakistan, Philippines, Singapore, Sri Thailand (mode suppliers foreign presence cial Foreign equity only Zimbabwe Bahrain Limitations TS of commer Arab a Kong GA , on only the Number suppliers Angola, Benin, Gabon, Mauritius, Morocco, nisiauT Hong (China), Macau (China), Qatar United Emirates under the or form Leone, Africa New on ypes)T Commitments Full, limitations only legal Egypt, Ghana, Kenya, Lesotho, Mozambique, Nigeria, Senegal, Sierra South Israel, Papua Guinea, Solomon Islands All of Lending on 2) Limited Benin, Gabon Israel, Malaysia and (mode Arab consumption Leone, Kong New , Deposits Commitments abroad Gambia, of Full The Ghana, Kenya, Malawi, Mozambique, Sierra nisia,uT Zimbabwe Bahrain, Hong (China), Indonesia, Kuwait, Macau (China), Papua Guinea, Philippines, Qatar Solomon Islands, United Emirates (Acceptance on 1) Limited Angola, Benin, Gabon, Morocco Israel, Kuwait, Malaysia (mode Banking border on Leone, New Arab Commitments cross supply , Gambia, Full The Ghana, Kenya, Malawi, Mozambique, Sierra unisia,T Zimbabwe Bahrain, Indonesia, Papua Guinea, Qatar Solomon Islands, United Emirates Commitments first modes Leone New Full on Ghana, Kenya, Malawi, Mozambique, Sierra Papua Guinea, Solomon Islands Access commitments three Market Africa Pacific Region 27.5 and Asia ableT 270 Brazil, Dominican Rep. Mexico , ,y , , de Salvador B. Hungar Slovenia Chile, Colombia, Ecuador El Honduras, Nicaragua, Peru, Uruguay R. enezuelaV Rep., Rep. Rica, Bulgaria, Czech Poland, Romania, Slovak Argentina, Bolivia, Costa Guyana, Haiti, Jamaica, Panama, Paraguay Rep., Rep., Czech Romania, Slovak Slovenia , Argentina, Ecuador Guyana, Haiti, Jamaica, Panama, Paraguay Rep., Rep., Czech Slovak Slovenia limitations. , other or Romania Ecuador Guyana, Haiti, Panama condition, reciprocity Guyana, Haiti, Panama licensing, y Europe (2000). America discretionar Caribbean Mattoo Latin the Eastern ce: Unbound, and a. Sour 271 D E V E L O P I N G C O U N T R I E S A N D N E G O T I AT I O N S O N T R A D E I N S E R V I C E S tries achieve this goal. At the same time, the benefits privatization tenders and new license offerings out- of liberalization will need to be underpinned by side their home market and around the world. adequate regulation and supervision of financial These trends mean that not only do emerging institutions and markets. Maintaining a sound economies see their own telecommunications com- financial system is essential to all countries, and the mitments as a means of attracting much-needed key international codes and standards for building foreign investment, but they have also developed sound financial systems will assist them in their important "export" interests in telecommunica- efforts. Market discipline will need to be applied, tions. In addition, developing countries are general- through appropriate disclosure and improved cor- ly more proactive in the WTO today than in the porate governance mechanisms. past--for example, in tabling negotiating proposals and requesting commitments of others--and they can be expected to do likewise in telecommunica- Telecommunications in the New Round tions. With the globalization of telecommunica- As was the case with financial services, negotiations tions, no national market will be considered "too on telecommunications were extended beyond the small" to be of interest in the negotiations. Uruguay Round. Talks were concluded in 1997, with As with any other sector in a trade round, some 80 countries scheduling commitments. (See telecommunications negotiations will have two Table 27.6, which includes later commitments.) The main objectives: (a) securing new telecommunica- outcome of these negotiations is discussed in Chapter tions commitments from the governments that cur- 28, by Mattoo. Negotiations on telecommunications rently have none (60, in this case), and (b) obtaining are expected to remain an important area of empha- improvements from those with commitments sis. The momentum for telecommunications reform already in their schedules (80 governments). around the world is such that much can be accom- Regarding new commitments, negotiations will plished in the new round of services negotiations. offer a blank slate for governments to commit on There are two reasons why new and continuing the telecommunications reforms they have put in reforms are especially common. First, there is a place since the mid-1990s, and governments will widespread realization that telecommunications has also be able to commit on blueprints for future become one of the most vital components of eco- reforms by specifying phase-in dates for various nomic infrastructure in support of growth and reforms in the GATS schedules. The main challenge development. Second, the advent of electronic com- for these governments will lie in drafting commit- merce and the digital economy is a catalyst for an ments that accurately reflect national decisions on even greater sense of urgency about modernizing the breadth and sequencing of reforms in view of the sector. These trends give participants in every differences in the domestic telecommunications economy--domestic and foreign, businesses and landscape (teledensity, population distribution, and consumers--cause for continued interest in nation- income distribution) and in economic development al telecommunications reforms and in the value of objectives. GATS commitments on those reforms. With respect to improving commitments, the As a result, in telecommunications there is a high challenge for governments is to convince one degree of synergy between national and multilateral another to reduce and eliminate the limitations that initiatives that will generate a positive atmosphere are inscribed in the existing schedules. Given the for the GATS negotiations. Major trading types of limitations currently listed in schedules, the economies hope to obtain additional market-open- following elements are likely to be considered. ing commitments, while most emerging economy governments consider telecommunications sector 1.Measures restricting foreign equity participation reform a national economic and social priority. The and the number of providers permitted to supply a negotiations, however, are not likely to pit North particular type of service. Developing countries against South. In many emerging economies where maintain most of these types of limitations. markets have already been opened, national opera- 2.Gaps in commitments created by sequencing on tors have become international players, perhaps national telecommunications liberalization. Some much to their own surprise. They find themselves governments committed on selected rather than participating, either alone or through consortia, in all market segments for voice telephone--for 272 The GATS: Key Features and Sectors Table 27.6 Current Status of GATS Commitments on Voice Telephone Services Fixed public voice telephone Additional Limits on commit- Additional foreign equity ments, commit- Phase-in or number Reference ments Participant Market segmenta dateb of suppliers Paper (RP)c (other)d Commitments in the Fourth Protocol of the GATS Antigua and Barbuda [-] [-] (I) [-] 2012 100% X [-] Argentina L LD I R 11/2000 100% X [-] Australia L LD I R 11.7% Telstra; X [-] 100% other Bangladesh L LD [-] [-] 100%; 3 suppliers; [-] To review Dhaka--L/LD monopoly and con- sider adding Belizee [-] [-] [-] [-] 25% single entity share- X [-] holding limit for BTL Bolivia L LD I R 12/2001 100% [-] X Brazilf,g** [-] [-] [-] [-] n.a. [-] Consid- ering introduc- tion of RP Brunei Darussalamh [-] [-] I [-] n.a.; I--duopoly X [-] Bulgaria (L) (LD) (I) [-] 2003­05 100% X [-] Canada L LD I R I­03/2000 46.7%--cumulative-- X [-] 20% direct, 33.3% indirect Chile [-] LD I R 100% X [-] Colombia L LD I [-] 70%; ENT only for LD X [-] and I voice Côte d'Ivoire (L) (LD) (I) (R) 2005 100% X [-] Czech Rep. L LD I R 2000 100% X [-] Dominicae [-] [-] [-] [-] 100% for committed X [-] services Dominican Rep. L LD I R 100% X [-] Ecuador [-] [-] [-] [-] n.a. [-] [-] El Salvador L LD I R 100% X [-] European Union L LD I R P, IR--2000 P: 25% F: 20% radio X [-] Gr 2003 licenses (direct only) Ghanah L LD I [-] Must have joint venture X [-] with nationals (no specific equity limit) Grenada (L) (LD) (I) (R) 2006 100% X [-] Guatemala L LD I R 100% [-] X Hong Kong (China) L LD [-] IR 100% X [-] Hungary (L) (LD) (I) (IR) LD--2003 75% for Matrav and X [-] L--2004 Antenna Hunagria; 100% other Iceland L LD I R 100% X [-] (continued) 273 D E V E L O P I N G C O U N T R I E S A N D N E G O T I AT I O N S O N T R A D E I N S E R V I C E S Table 27.6 (continued) Fixed public voice telephone Additional Limits on commit- Additional foreign equity ments, commit- Phase-in or number Reference ments Participant Market segmenta dateb of suppliers Paper (RP)c (other)d Indiah L LD [-] [-] Duopoly by service area; X X 25% Indonesiah L [-] I [-] Local PTT + 5 cooperators; X [-] I--duopoly; 35% Israelg,h [-] [-] I [-] 74% X [-] Jamaica (L) (LD) (I) 09/2013 100% X [-] Japan L LD I R 20% NTT/KDD; 100% other X [-] Korea, Rep. L LD I R For. eq.-- 33% KT; 49% other X [-] 2001 (single investor limit-- 10% wire-based voice) Malaysia L LD I [-] Limited to current [-] X suppliers; 30% Mauritius (L) (LD) (I) [-] 2004 100% [-] Consid- ering introduc- tion of RP Mexico L LD I R 49% X [-] Morocco L LD I [-] 2002 IAM: unbound; all others X [-] 100% at phase-in New Zealand L LD I R 49.9% NZT (single X [-] investor limit only), 100% other Norway L LD I R X [-] Pakistan [-] [-] [-] [-] Commercial presence: unbound X [-] Papua New Guineah ** [-] [-] [-] [-] X [-] Peru L LD I R 07/1999 100% X [-] Philippines** L LD I [-] 40% X Poland L (LD) (I) (R) 2003 I (facilities based)-- X [-] 49% at phase-in; I (resale) and LD--49%; L--100% Romania (L) (LD) (I) (R) 2003 100% X [-] Senegale,h [-] [-] [-] [-] n.a. X [-] Singapore L LD I [-] 04/2000 73.99% (49% direct; X [-] 24.99 indirect) Slovak Rep. (L) (LD) (I) (R) 2003 100% X [-] South Africah (L) (LD) (I) (R) 2003 Duopoly at phase-in; 30% X [-] Sri Lankae,h L LD I [-] I--01/ 35% for international; X [-] 2000, 40 % other committed subject to services; L and LD for tariff re- WLL only--3 suppliers; balancing I--duopoly 274 The GATS: Key Features and Sectors Switzerland L LD I R 100% X [-] Thailandf [-] [-] [-] [-] n.a. Consid- ering introduc- tion of RP Trinidad and Tobagof (L) (LD) (I) (R) 2010 100% X [-] Tunisia (L) [-] [-] [-] 10% TT (2002), other 49% [-] [-] Turkey (5) (L) (LD) (I) (R) 2006 49% at phase-in [-] X United States L LD I R 20% radio licenses (direct only); 100% other X R. B. de Venezuela L LD I [-] 12/2000 100% [-] X Commitments submitted after Fourth Protocol Albania L (LD) (I) (R) 2003 100% X [-] rural; (L) urban Barbados (L) (LD) (I) (R) 2012 100% X [-] China (L) (LD) (I) (R) By service Joint venture required. X [-] area, 25% as of 12/2004; 12/2004­ 35% as of 12/2006; 12/2007 49% as of 12/2007 Croatia (L) (LD) (I) (R) 2003 100% X [-] Cyprush [-] [-] [-] [-] n.a. [-] Decision taken in 1998 re: liber- alization Estonia L (LD) (I) (R) 2003 100% X [-] Georgia L LD I R 100% X [-] Jordan (L) (LD) (I) (R) 2005 100% X [-] Kenya L LD (I) (R) 2003; 30% X [-] (L) Nairobi Kyrgyz Rep. L (LD) (I) (R) 2003 100% X [-] Latvia (L) (LD) (I) (R) 2003 100% X [-] Lithuania (L) (LD) (I) (R) 2003 100% X [-] Moldova (L) (LD) (I) (R) 2004 100% X [-] Oman (L) (LD) (I) (R) 2004 70% as of 2001; X [-] 100% as of 2005 Surinameg,h L LD I 40% public services; X [-] 100% nonpublic fixed and wireless nonvoice services; duopoly for public voice Taiwan (China) L LD I R Chunghwa Telecom X [-] 20% aggregate dir. and indir.; facilities-based 20% dir. and 60% aggregrate dir. and indir.; 100% resale (continued) 275 D E V E L O P I N G C O U N T R I E S A N D N E G O T I AT I O N S O N T R A D E I N S E R V I C E S Table 27.6 (continued) Fixed public voice telephone Additional Limits on commit- Additional foreign equity ments, commit- Phase-in or number Reference ments Participant Market segmenta dateb of suppliers Paper (RP)c (other)d Uganda L LD I Duopoly X [-] Total governments, including other (86) 73 72 73 57 73 7 (Subject to phase-in) (21) (24) (26) (22) [-] No commitment. ( ) Phased-in commitment. ** Entry into force pending acceptance of Fourth Protocol. n.a. Not applicable. a. Listings for voice telephone are broken down by market segment: local (L), domestic long distance (LD), international (I), and R, resale of public voice. The listings also show whether the commitment to competition is, or was, to be phased in on a date subsequent to the entry into force of the schedule as a whole. A market segment is indicated as subject to competition on entry into force of the schedule if it may be provided, at that time, by two or more suppliers. The public voice commitments in this table relate to commercial presence (GATS mode 3). Many GATS schedules commit on other telecommunications services such as mobile voice, data transmission, mobile services, and value-added services, which are not included in this table. b. Italics denote a phase-in date that has passed. c. "X" denotes that member incorporated the Reference Paper on regulatory principles with few, if any, modifications. d. "X" denotes that member included some portions of the Reference Paper or independently drafted regulatory commitments, or as indi- cated. e. Fourth Protocol commitments do not include voice telephone services but do include other services. f. Commits to improve offer once pending national legislation has been adopted. g. Where no public voice telephone commitments are indicated, voice over closed user groups is nonetheless committed. h. Commits to review the possibility of allowing market access for additional suppliers. Source: WTO. example, on local services but not international, view that amending the commitments to include or the reverse. Another example is facilities-based the earlier dates will help secure investor confi- services versus simple resale of voice services. It is dence. not uncommon for schedules of developing 4.Modes of supply. In the telecommunications sec- countries to commit on opening fixed and mobile tor, as in many other service sectors, commercial telephony to competition while offering no com- presence is often more open than cross-border mitments to allow simple resale. In low-teledensi- supply. Unlike other sectors, however, telecom- ty countries, such commitments may reflect munications, by its very nature, has always been reforms that give priority to expansion of infra- an "on-line" service. In many respects, restrictions structure and the subscriber base. " In the course on cross-border supply (that is, on the ability of a of the new round, some governments may find foreign company to sign up customers within a that advances in their telecommunications national territory without a physical presence regimes will facilitate expansion of the scope of there) may represent a holdover from the era of commitments. monopoly market structures and fixed-line 3.Phase-in dates inscribed by governments in sched- telephony. Today, wireless technologies (both cel- ules regarding the staged implementation of reforms lular and satellite), IP telephony (the so-called (that is, legally bound commitments to offer mar- voice over Internet protocol), and international ket access on a specified date in the future). About simple resale techniques have important implica- 40 percent of existing GATS telecommunications tions. Among these are the technical feasibility of commitments use the phase-in approach. Some offering telecommunications services directly to a commitments of this kind have already gone into country's consumers from beyond its borders; the effect, but others stretch to 2005 or beyond. A fact that these technologies are increasingly number of governments have succeeded in bring- attractive to both consumers and suppliers ing the dates of planned telecommunications because they can be more cost-efficient than a full reforms significantly forward, and some take the commercial presence; and the considerable 276 The GATS: Key Features and Sectors potential that such technologies hold for expand- tiations were aimed at commitments on each of ing access to telecommunications services in three so-called "pillars": international shipping, developing countries.3 auxiliary maritime services, and access to and use of 5."Other services." It perhaps goes without saying port facilities. In the event, no agreement could be that negotiators will see more commitments on reached, and negotiations were suspended. telecommunications services that are often listed Thus, even though the maritime transport sector in schedules as "other services." These include is an integral part of the GATS, it is not subject to competitive supply of satellite services and leased the MFN rule, and existing market access and circuit services (or trading in bulk or wholesale national treatment commitments are limited to transport capacity). In addition, because of an those that certain members have been willing to anomaly resulting from the focus of the extended make unilaterally. The suspension of the MFN obli- negotiations on basic telecommunications ser- gation was prompted by the difficulty of eliminat- vices, a variety of computer-age telecommunica- ing MFN-inconsistent measures in the maritime tions services (often referred to as value-added or sector.4 An example of such measures is bilateral enhanced services), such as e-mail and database cargo-sharing arrangements such as those under access and retrieval, are subject to fewer commit- the United Nations Code of Conduct for Liner Con- ments than some basic services, even though they ferences, which allows up to 80 percent of bilateral have long been open to competition in many traffic flows to be reserved for national-flag carriers. regimes. Another example of MFN-inconsistent measures is unilateral retaliatory actions against trading part- Finally, many governments have found the so- ners who are perceived to resort to restrictive for- called Reference Paper on telecommunications reg- eign trade practices. ulatory principles (described in Box 29.2) a useful U.S. legislation in the maritime sector provides an blueprint for supporting the transition from example of the tension in the multilateral trading monopoly to competitive markets and ensuring system between unconditional MFN and reciproci- success in implementing desired telecommunica- ty.5 When a country accepts the full disciplines of tions reforms. Governments that have not included any WTO agreement, it forgoes the right to discrim- the regulatory principles in their schedules show inate between its trading partners. It is then obliged significant interest in taking them on, and govern- to extend no less favorable treatment to a trading ments that have scheduled them continue to be partner that has a relatively protected market than it interested in encouraging their trading partners to does to another trading partner with a relatively do so. Indeed, it is widely recognized that telecom- open market. If a country is concerned with gaining munications regulators are the front line in ensur- improved access for its exporters, reciprocity (the ing that telecommunications reforms succeed and policy of "doing unto others as they do unto you") that GATS commitments are honored. Many regu- can become, in the hands of a large country, an lators in developing countries, being new and inex- effective instrument for prizing open foreign mar- perienced, view the Reference Paper as an asset in kets. If the United States had participated in an meeting the day-to-day challenges of their work. MFN-based agreement, it would no longer have the freedom to resort to retaliatory legislation--some- thing it argued that it needed, given existing restric- Maritime Negotiations in the WTO tions on its access to foreign markets. Despite nearly 10 years of difficult negotiations, lit- What are these restrictions? International ship- tle success has so far been achieved on maritime ping is, for the most part, a relatively open sector. transport in the WTO. Liberalization in maritime Government-created barriers affect a small and transport services was a central concern in the shrinking proportion of liner shipping, and bulk Uruguay Round, but at the end of the process only a shipping is virtually free of restrictions. The most few countries were willing to offer commitments, protected segment in many countries is cabotage, most with significant limitations. As in other areas which, ironically, was excluded from the scope of (telecommunications, finance, and movement of the WTO negotiations, presumably because coun- natural persons), it was decided to extend negotia- tries did not feel ready to liberalize coastal domestic tions in this sector until the end of June 1996. Nego- shipping trade. Restrictions also exist in auxiliary 277 D E V E L O P I N G C O U N T R I E S A N D N E G O T I AT I O N S O N T R A D E I N S E R V I C E S services and port services, where public monopolies transport services are, to a large degree, excluded are only gradually being privatized and more com- from the GATS. During the Uruguay Round, WTO petition is being introduced. A third area, and one members only negotiated certain complementary of growing importance, is multimodal transport, services (aircraft repair and maintenance, selling whereby goods pass from one country to another by and marketing of air transport services, and com- various modes of transport in addition to ocean puter reservation services). A GATS Annex on Air transport. Even though foreign multimodal trans- Transport specifically excludes from GATS rules the porters usually receive nondiscriminatory access to complex network of bilateral agreements on air onward transport by road and rail, they frequently traffic rights, and the MFN obligation has been sus- encounter difficulties in establishing their own pended for the air transport sector. inland transport operations. This can put them at a Although precise estimates are hard to come by, competitive disadvantage in situations where verti- empirical research suggests that there may be sub- cal integration provides benefits. stantial gains from reforming the current interna- In 2000 maritime discussions were reinitiated as tional air transport regime in the direction of more part of the broader negotiations on all services open and competitive markets. Findlay and Nikom- called for in GATS Article XIX. There are several borirak (1999) report on work by Oum and Yu reasons for believing that negotiations may be more (1998) that analyzes differences in airlines' unit successful this time. First, it may be possible to costs, decomposed into labor costs and operating break the stalemate in maritime transport services efficiency. The latter might be affected by the extent by exploiting intersectoral negotiating tradeoffs. of competition in markets, the characteristics of the Second, it may be easier to negotiate commitments overall management and control of the firm, the use in the important area of multimodal transport of information technology systems, aircraft utiliza- when all transport sectors are being negotiated tion planning, levels of employee skills, and so on. rather than maritime alone. Finally, and perhaps The results of Oum and Yu's analysis illustrate the most important, unilateral liberalization in this sec- possibility for developing economies to host inter- tor is gathering steam as more and more countries nationally competitive carriers. Moreover, although appreciate that restrictions on maritime trade input prices have traditionally been the most impose a significant cost on the whole economy important determinant of competitiveness, efficien- (see WTO 1998c). It may therefore soon be possible cy improvements resulting from market liberaliza- to lock in the liberalization already achieved in an tion and openness to foreign investment can yield agreement under the GATS. significant reductions in airlines' unit costs. Findlay and Nikomborirak (1999) also report on modeling work undertaken by the Australian Pro- Liberalizing Trade in Air Transport Services ductivity Commission that examines the effects of International air transport services are governed by market liberalization policies on services offered on an elaborate set of more than 3,500 bilateral agree- routes to and from Australia. Among the estimated ments. These agreements typically specify the air- impacts of entry of another carrier were price lines of the parties that are allowed to fly on each decreases of 2.4­7.7 percent, increases in both Aus- international route between two countries, the tralian and foreign welfare, and an increase in total capacity that can be provided by those designated net passenger movements of nearly 4 percent. airlines, and the extent to which capacity can be Besides the beneficial effect on the sector's per- offered by airlines from third countries. In econom- formance, there are several channels through which ic terms, they define country- and route-specific air transport liberalization may provide a positive quotas. Most agreements only cover international stimulus to other sectors of the economy, notably traffic between two countries. Domestic air trans- tourism services. In the case of air cargo, more effi- port services are still generally protected from for- cient service provision may translate into greater eign competition. participation in international trade. The existing structure of bilateral agreements is What are the possible approaches to sector inherently discriminatory, in that the agreements reform? One route is from within the system of violate both the most-favored-nation (MFN) prin- bilateral agreements. The increased proliferation in ciple and the national treatment principle. Air the 1990s of open skies agreements, which remove 278 The GATS: Key Features and Sectors restrictions on access between and beyond the vices. Frequent flier programs and long-term negotiating countries for their airlines, have been arrangements with travel agents may pose substan- credited with leading to substantial market opening tial entry barriers to foreign firms. Finally, the emer- on selected routes, notably over the North Atlantic. gence of carriers' alliances, which often substitute This approach would, however, retain the discrimi- for multinational mergers where these are restrict- natory nature of the current system. Countries that ed, has raised concerns about market concentration do not enter into an agreement with a large trading and cartel-like practices. Enforcement of competi- partner risk diversion of traffic to third countries. tion policies in the international arena is necessarily Because most open skies agreements do not cover difficult. Here again, the GATS may offer a small domestic traffic, foreign carriers are at a competitive step forward--for example, by establishing a core disadvantage, since they are not able to draw on the set of procompetitive regulatory principles, such as extensive network maintained by domestic airlines. already exist in the telecommunications sector. A series of open skies agreements is therefore unlikely to lead to a progressively more open and Notes competitive system for international air transport. A second route to reform would be to rely on Contributions to this chapter are as follows: "The Agreement," regional agreements, which are likely to be associat- Bernard Hoekman; "The Current Pattern of Commitments," Rudolf Adlung and Antonia Carzeniga; "Financial Services: Past ed with broader trade agreements, to foster sector Negotiations and Issues for the Next Round," Masamichi Kono; reform. Examples of regional initiatives include the "Telecommunications in the New Round," Lee Tuthill; "Maritime European Union agreement to establish a single Negotiations in the WTO," Aaditya Mattoo; and "Liberalizing market by 1997 and the Andean Pact Open Skies Trade in Air Transport Services," Aaditya Mattoo, based on input Agreement. The danger in regional approaches lies from Carsten Fink and on Findlay and Nikomborirak (1999). in the formation of blocs that lead to greater com- 1 It should be noted that GATS Article I states that the agree- petition within the agreement area but raise barriers ment applies to measures affecting trade in services, whereas for airlines from outside the region. Article XVII refers to all measures affecting the supply of ser- The third approach, and, it can be argued, the vices. It is not clear whether any significance should be attached to the distinction. In any case, it would seem that the most desirable one in the long run, would be to set of measures affecting the supply of services cannot be nar- strengthen the GATS obligations with regard to air rower than that affecting trade in services. transport services and to negotiate market access 2 MTN.GNS/W/164, September 3, 1993. The document warns concessions in a multilateral trading round. A first that "the answers should not be considered as an authoritative step might be to expand the coverage of services to legal interpretation of the GATS." It is, however, the basis on include air freight and chartered services. A recent which many schedules of specific commitments have been UNCTAD Expert Meeting on Air Transport Ser- drafted. vices observed that a large number of countries 3 Often, limited forms of commercial presence are nonetheless might wish to exclude the coverage of mode 1 necessary. Such forms may include local intermediary services (cross-border supply) but might be more open to suppliers for purposes of marketing subscriber takeup, local partners for billing and customer service functions, and pres- negotiations on mode 3 (commercial presence). ence of some localized facilities such as switches or leased line GATS coverage of the latter mode only would at arrangements that may require on-site representatives or tech- least offer the benefits of the GATS in terms of nicians. MFN, market access, and national treatment for 4 Even though the GATS does permit members to seek tempo- foreign investors in this sector. rary exemptions from the MFN obligations, the dominant view As a final remark, it is important to stress that any was that the continued suspension of the MFN rule would reform route needs to be accompanied by appropri- avert the need for many countries to take MFN exemptions ate mechanisms that address anticompetitive that may be more difficult to negotiate away once explicitly listed. behavior of air transport services providers. Several characteristics of the sector highlight the potential 5 The three relevant pieces of legislation are Section 19 of the for anticompetitive business practices. Major hubs Merchant Marine Act of 1920, Section 13(b)(5) of the Ship- ping Act of 1984, and the Foreign Shipping Practices Act of are often dominated by the main national carriers, 1988. Even though specific action has rarely been taken, it has raising the possibility that foreign carriers might be been claimed that the credible threat of doing so has induced denied access to airports or essential airport ser- an opening of foreign markets in some instances. 279 28 A A D I T YA M AT T O O Negotiating ownership from public to pri- vate hands. In the GATS context, countries, under pressure from Improved trading partners, have often conceded increased market Market Access access in the form of allowing increased foreign ownership of Commitments existing domestic firms or limit- ed new entry rather than elimi- nating all barriers to entry. Considerable negotiating energy has also been devoted to protect- ing incumbent foreign owner- T ship (Mattoo 1999). This trend his chapter focuses on three issues was particularly visible in the negotiations on finan- that influence the quality of liber- cial services, where "grandfather provisions" guar- alization commitments. These issues are the rela- anteed the ownership and branching rights of tionship between foreign equity participation and incumbent foreign firms while far more limited the conditions of competition in the market; the use rights were afforded potential entrants, possibly of General Agreement on Trade in Services (GATS) placing them at a competitive disadvantage. schedules as a mechanism for precommitment to Foreign investment clearly brings benefits even future liberalization; and the link between unilater- in situations where it does not lead to enhanced al liberalization and reciprocity-based market access competition (that is, where there are entry restric- negotiations. tions). Foreign equity may relax a capital con- straint, help ensure that weak domestic firms are bolstered (for example, by recapitalizing financial Choosing the Pattern of Liberalization institutions), and serve as a vehicle for transferring Restrictions on foreign commercial presence technology and know-how, including improved assume particular significance in the case of services management. If, however, foreign direct investment for which cross-border delivery is not possible, so (FDI) is undertaken simply because the returns to that consumer prices depend completely on the investment are artificially raised by restrictions on domestic market structure. Under the GATS, competition, the net returns to the host country restrictions on new entry and on the participation may be negative; that is, the returns to the investor of foreign capital are most common, particularly in may exceed the true social productivity of the communications and financial services (Table 28.1). investment (Hindley and Smith 1984). To some A basic conclusion from the literature on privatiza- extent, the rent appropriation may be prevented by tion is that larger welfare gains arise from an taxing profits or by holding competitive auctions of increase in competition than from a mere change in licenses or equity, but the static and dynamic ineffi- 280 Negotiating Improved Market Access Commitments Table 28.1 Types of Market Access Restrictions on Commercial Presence in Services Sectors, All WTO Members, 2000 Restrictions per commitment Number of members On On value On On On partici- with number of trans- number number On type pation of commit- of actions of opera- of natural of legal foreign Sector ments suppliers or assets tions persons entity capital Business services 89 0.2 3.4 0.0 0.8 4.0 3.6 Communications services 85 3.9 2.1 0.2 0.5 4.3 3.8 Construction and related engineering services 60 0.1 0.9 0.0 0.2 0.9 0.9 Distribution services 38 0.2 0.7 0.0 0.1 0.4 0.4 Educational services 32 0.1 0.3 0.0 0.0 0.9 0.5 Environmental services 40 0.2 0.7 0.0 0.1 0.7 0.6 Financial services 91 4.6 4.3 1.4 0.9 8.6 4.4 Health and related social services 34 0.2 0.3 0.1 0.2 0.6 0.4 Tourism and travel-related services 114 0.2 0.4 0.1 0.1 0.5 0.4 Recreational, cultural, and sporting services 49 0.1 0.3 0.0 0.1 0.5 0.2 Transport services 70 0.3 0.9 0.0 0.1 1.9 1.4 Note: Restrictions per commitment are calculated by dividing the total number of restrictions in a sector by the number of members with commitments. Source: WTO Services Database. ciencies from lack of competition would remain.1 duplication. It is possible to think of other special Although much of the theoretical basis for these models of market or regulatory failure where entry assertions is in a static context, there is an ever- barriers enhance welfare (Laffont 1999). stronger presumption that competition also pro- Notwithstanding these considerations, entry duces significant dynamic benefits through its restrictions are becoming more difficult to justify in impact on the incentives to improve performance the face of technological change and the mounting and to innovate. evidence that competition works.2 Technological advances have significantly lowered network costs, and vertical separation ("network unbundling") has Are There Any Good Reasons to Limit the Number widened the scope for competitive entry (Smith of Suppliers? 1995). Furthermore, the inefficiencies introduced In some cases there is no choice because of technical by duplication of networks may be small compared limitations on competition, such as those imposed with the operational inefficiencies that can result by the scarcity of radio frequencies needed for the from lack of competitive pressure.3 provision of mobile telecommunications services or The observed restrictions on entry may well have by the limited space available for department stores more prosaic purposes. First, restrictions may be or airports in a city. In other segments entry restric- designed so that incumbent suppliers are only grad- tions might be justified by the existence of signifi- ually exposed to competition--for infant indus- cant economies of scale stemming, for example, try­type reasons, to facilitate "orderly exit," or from the substantial fixed costs of networks; that is, simply because of political-economy pressures. This competitive entry could lead to inefficient network explains, for example, why governments have gener- 281 D E V E L O P I N G C O U N T R I E S A N D N E G O T I AT I O N S O N T R A D E I N S E R V I C E S ally been more willing to liberalize mobile than 1.If there are rent-generating restrictions on com- fixed-line telecommunications services; mobile petition, the purpose of the observed limitations telephony has only recently been introduced, and on ownership may be to balance the efficiency- there is thus no incumbent to protect. The entry enhancing and rent-appropriation aspects of for- restrictions sometimes benefit not only national eign investment. This argument does raise the firms but also foreign incumbents, as was the case question of why rent appropriation cannot be with financial services in Malaysia. Other instru- prevented by ex ante auctions of equity or ex post ments, such as discriminatory subsidies or taxes, taxation of profits.4 (And, indeed, why do restric- could be more easily targeted to achieve protection tions on competition continue to exist?) of the national firm. 2.Under a sort of "infant entrepreneur" rationale, Monopoly or oligopoly rents are sometimes seen foreigners are induced to form equity joint ven- as a means of helping firms fulfill universal service tures so that local investors can learn by collabo- obligations through cross-subsidization. Govern- rating. As with all such arguments, it is difficult to ments, however, are increasingly devising means of judge whether the costs of protection are likely to achieving these objectives without sacrificing the be offset by the eventual benefits. benefits of competition. In some cases a form of 3.Probably most important is a purely political "investment pessimism" leads to the belief that reluctance to allow foreign control of an essential promises of oligopoly rents are necessary to finance service. These political concerns should be less new investment--although it is not clear why the strong if it is not one foreign monopolist but a market structure needs to be determined by policy, number of competing foreign firms that provide unless there are some initial investments the bene- the service. fits of which may be appropriated by rivals. Finally, governments may seek to raise revenue (or rents for In any case, there is so far no good analytical and politicians and bureaucrats) by auctioning monop- empirical basis for evaluating the benefits and costs oly or oligopoly rights. This usually explains the of ownership restrictions and how they interact promise of exclusive rights prior to privatization. with entry restrictions. Where competition would be feasible, granting exclusive rights amounts to indirect appropriation What Can the GATS Do? of consumers' surplus and may deny important dynamic efficiencies consequent on competition. The domestic political-economic forces that lead to Ideally, governments would not resort to trade protection may also dictate that it be obtained restrictions to pursue objectives that are better through inefficient instruments. Unlike the GATT, achieved through other means. In each of the cases the GATS has created no hierarchy of instruments listed above, entry restrictions are a second- or of protection, although the ranking of instruments third-best instrument for achieving the objective in in the cases of both goods and services is similar. question but are chosen because of constraints such Hence, quantitative restrictions, which have been as inability to raise revenue without economic or discredited and outlawed as regards trade in goods, political cost. flourish in trade in services. When the rents accrue to foreigners, these quotas resemble voluntary export restraints. For instance, in the last round of Are There Good Reasons to Limit Foreign negotiations countries sometimes conceded, and Ownership, and What Are the Implications? trading partners were content to receive, increased Most countries in the region maintain limits on for- market access in the form of increased foreign own- eign or private ownership, or both, but it is not easy ership of existing domestic firms, rather than to find a sound economic rationale for these restric- through allowing new entry. tions. Insofar as the incentives to transfer technolo- Although it may not yet be politically feasible, or gy, improve management, and so on are related to unambiguously desirable, to impose the same hier- an owner's share in the profits, ownership limita- archy of instruments in services as in goods, an tions are bound to dampen those incentives and attempt could nevertheless be made to create a legal adversely affect firm performance. Governments are presumption in favor of instruments such as fiscal willing to bear this for three types of reasons: measures that provide protection more efficiently. 282 Negotiating Improved Market Access Commitments The difficulty of switching to fiscal instruments of contestability of markets. Unfortunately, even in rela- protection in services has probably been exaggerat- tively open markets, commitments are sometimes ed. In the case of commercial presence, a number of couched in language that diminishes their value. For fiscal instruments are possible, including entry taxes instance, Chile's financial services schedule states that (or auctions of entry licenses), output taxes, and "a supplier of financial services operating through a profit taxes. An entry tax increases the fixed costs of commercial presence may be subject to evidence of firms and lessens their willingness to enter the mar- economic need," and Philippines' telecommunica- ket; the market structure is therefore likely to be less tions schedule notes that entry is subject to a "Fran- competitive than in the absence of the tax. An out- chise from the Congress of the Philippines" and a put tax on foreign suppliers increases their marginal "Certificate of Public Convenience." It is far from cost of providing a service and is similar in effect to clear whether such approval is contingent only on a specific tariff. A profit tax is least likely to affect the transparent and nondiscriminatory criteria such as economic decisions of firms, but if there are fixed technical or financial soundness or whether approval costs of entry that must be covered by future profits, is a euphemism either for a restriction on the num- a profit tax would reduce the number of firms that ber of firms or for discrimination against foreign could recover those costs. One or more of these fis- entrants. A priority in the next round would be to cal instruments could help achieve outcomes supe- purge the schedules of such language. rior to quotas, from a social welfare point of view. One reason governments may be reluctant to lib- Ironically, the legal systems of many countries allow eralize immediately is a perceived need to protect discrimination against foreigners through outright incumbent suppliers from competition, because of bans and entry quotas but make it difficult to infant industry­type arguments or to facilitate impose discriminatory taxes. For instance, in the "orderly exit." The failure of infant industry policies European Union a locally established foreign firm is in the past, and the innumerable examples of per- treated in all respects like a European firm and may petual infancy, are in part attributable to govern- not be subject to any form of discrimination. ments' inability to commit themselves credibly to liberalize at some future date. The GATS offers a valuable mechanism for overcoming the credibility Precommitment to Future Liberalization difficulty. Under the agreement, governments can Policies that are believed in are most likely to suc- make binding commitments to provide market ceed. The provision of many services, from finance access and national treatment at a future date. Fail- to transport, requires highly specific sunk invest- ure to honor these commitments would create an ments in assets that are not easily deployable for obligation to compensate those who are deprived of other uses. Investors' business plans are typically benefits, making the commitment more credible stretched out over long time periods, and many ser- than a mere announcement in the national context vices providers expect to incur substantial losses in of intent to liberalize. A precommitment to liberal- their first years of operation. It is therefore impor- ize can also instill a sense of urgency regarding tant that market liberalization programs be credi- domestic reform and efforts to develop the neces- ble. Otherwise two sorts of problems can arise: (a) If sary regulatory and supervision mechanisms. there is significant uncertainty about policy, fewer Several governments have taken advantage of the investments will be made, and services providers GATS to strike a balance between their reluctance to will demand a premium on their returns on capital. unleash competition immediately on protected (b) If it is possible to influence policy, services national suppliers and their desire not to be held providers may behave strategically to manipulate hostage in perpetuity either to the weakness of policy choices in their own favor. domestic industry or to pressure from vested inter- Credibility itself has two dimensions. One is to ests. The most striking examples are in basic convince agents that current reforms will not be telecommunications, where a number of develop- reversed. The other is to persuade them that future ing countries have bound themselves to introduce reforms will be carried out. competition at precise future dates (Table 28.2). Under the GATS, many countries have bound the The use of the GATS as a mechanism for lending status quo. In principle, a clear GATS commitment credibility to liberalization programs has been dis- not to restrict entry could add significantly to the appointing in other sectors. 283 D E V E L O P I N G C O U N T R I E S A N D N E G O T I AT I O N S O N T R A D E I N S E R V I C E S Table 28.2 Developing Country Precommitments to Liberalize in Basic Telecommunications Negotiations, 1998 Economy Precommitment to liberalization Latin America Antigua and Barbuda International voice telephony is reserved for an exclusive operator until 2012 but will be opened to competition as of 2012 (applies only to inter- national service). Domestic service is reserved to a government-owned exclusive operator. Argentina No restrictions after November 8, 2000. Bolivia Exclusive contract for six years to provide long-distance national and inter- national telecommunications services, but no restrictions on these services after November 27, 2001. Brazil Within one year after the enactment of the draft General Telecommunica- tions Law, Brazil will schedule commitments on the supply of public telecommunications services binding the relevant parts of the new law. Chile Limits on competition in national long-distance service for a four-year peri- od starting August 27, 1994. Grenada Reserved for exclusive supply until 2006; no restrictions thereafter. Jamaica Reserved for exclusive supply until September 2013; no restrictions there- after. Mexico International long-distance services that require use of satellite are obliged to use Mexican satellite infrastructure until 2002. Trinidad and Tobago Reserved for exclusive supply until 2010; no restrictions thereafter. Fixed satellite services: until 2000, only on satellite network capacity sup- plied by the exclusive public operator; no restrictions thereafter. Venezuela No restrictions after November 27, 2000. Africa Côte d'Ivoire Monopoly until 2005; no restrictions thereafter. Ghana Duopoly for five years followed by review of policy. Mauritius Monopoly until 2004; no restrictions thereafter. Morocco Monopoly until 2001; no restrictions on entry thereafter Senegal Monopoly until at least December 31, 2003, but not after December 31, 2006; review of policy in 2003. South Africa Monopoly until December 31, 2003; then duopoly, and authorities will consider the feasibility of more licenses. Tunisia No restrictions on supply of local calls after 2003. Asia Bangladesh Will review the possibility of adding regulatory principles in the future. Brunei Darussalam Will review policy and consider whether to allow additional suppliers 10 years after privatization (at an unspecified date) of public monopoly for local services and in 2010 for international services. Hong Kong (China) Will consider issuing more than the existing four licenses for local fixed network services in June 1998. India Will review the subject of opening up national long-distance service beyond the defined service area to competition in 1999 and opening up international services in 2004. Indonesia Will review policy to determine whether to admit additional suppliers on the expiry of exclusive rights (in 2011 for local service, in 2006 for long- distance service, and in 2005 for international service). 284 Negotiating Improved Market Access Commitments Korea, Rep. of Beginning in 2001, will raise foreign equity participation in facilities-based suppliers from 33 to 49 percent and in the national supplier (Korea Tele- com) from 20 to 33 percent. Market access for domestic voice resale to be allowed as of 1999, with foreign equity participation up to 49 percent, to be raised to 100 percent after 2001. Pakistan Will phase out exclusivity on cross-border supply of voice telephony as of 2004. Proposes to divest 26 percent of the Pakistan Telecommunication Company Ltd. stake through international competitive bidding to a strate- gic investor that will have an exclusive license for the operation of basic telephonic services for seven years. Singapore Will phase in competition of facilities-based telecommunications services in April 2000, when up to two additional operators will be licensed. Addi- tional licenses will be granted thereafter. Sri Lanka Proposes to issue an additional license in 2000, depending on satisfactory progress by the monopoly on tariff rebalancing. Number of operators licensed for local and domestic long-distance mobile cellular services to be reviewed in 2000. Thailand Will introduce revised commitments for voice telephone and several other services in 2006, conditional on the passage and entry into force of new communication acts. Using the GATS Negotiations to Enhance the existing levels of unilaterally determined policy Market Access rather than liberalization achieved through a recip- rocal exchange of "concessions." Reciprocity has been a central principle governing It might well be that reciprocity cannot and will GATT/WTO negotiations: one country reduces its not play a major role in services trade. Services level of protection in return for a reciprocal reduc- liberalization could for the most part be under- tion by its trading partner.5 Since reciprocity-based taken unilaterally, and the GATS would be impor- negotiations are widely credited with the substantial tant only in preventing the reversal of reduction in levels of protection achieved in goods liberalization--that is, in its credibility role (see trade, it is surprising that the limited application of Hoekman and Messerlin 2000). Indeed, for coun- the principle has not, conversely, been seen as the tries that are either determined to liberalize or reason for the disappointing results in services determined to protect, negotiations are not trade. important. For countries in the middle ground, The GATS was given a deliberately symmetric however, that are open to reform but whose abili- structure. In principle, there was scope for industri- ty to implement reform is constrained by domes- al and developing countries to exploit their modal tic opposition, multilateral negotiations can be comparative advantages: improved access for capital useful. Many developing countries today are in from industrial countries was exchanged for this situation. Furthermore, with severe shortages improved temporary access for individual services of skilled labor in the United States and Europe providers from developing countries. In practice, and with the powerful constituency of high-tech- there was little political will to improve access for nology companies lobbying for relaxation of visa foreign individuals (except for the limited class of limits, the prospects for serious intermodal trade- skilled intracorporate transferees), and a tradeoff off--such as obtaining labor movement in return between modes of delivery simply did not take for allowing greater commercial presence by for- place. Moreover, even the negotiating links across eign service providers--are more promising. And service sectors and between services and goods sec- a wider application of the principle of reciprocity tors do not seem to have been particularly fruitful. may deliver expanded liberalization and more bal- So, the GATS commitments reflect for the most part anced outcomes (see Box 28.1). 285 D E V E L O P I N G C O U N T R I E S A N D N E G O T I AT I O N S O N T R A D E I N S E R V I C E S B O X 2 8 . 1 S E R V I C E S E X P O R T S B Y D E V E L O P I N G C O U N T R I E S : P O T E N T I A L G A I N S A N D C U R R E N T B A R R I E R S There are likely to be significant gains worldwide times that of India (the cheapest country), and if restrictions on services exports from developing average salaries are more than 11 times higher in countries are eliminated. With greater liberaliza- Switzerland (Mattoo 1999). Even though differ- tion, particularly in mode 4--movement of natu- ences in labor productivity imply that a lower ral persons--many more developing countries average salary of programmers may not necessar- could "export" at least the significant labor com- ily translate into a lower average cost per line of ponent of construction, distribution, environ- software code, by outsourcing programming mental, and transport services, among others. activities, firms in industrial countries can save One of the most striking recent success stories significantly on development and support costs. in the areas of developing country services Against the background of a total market for soft- exports is the Indian software industry, which has ware services worth (in 1997) about US$58 bil- emerged as a significant supplier to industrial lion in the United States, US$42 billion in Europe, country markets. Indian software exports grew and US$10 billion in Japan, such cost savings from US$225 million in 1992­93 to US$1.75 bil- could well be substantial (computed from WTO lion in 1997­98 (a compound annual growth rate 1998b: table 3). Other gains from trade liberal- of approximately 50 percent).* This story has ization for importing countries include a more some noteworthy elements. competitive market structure for software ser- First, despite the growing importance of cross- vices, increased choice (because countries may border electronic delivery of software services, develop a special expertise in certain develop- the movement of natural persons remains a cru- ment or support services), and greater diffusion cial mode of delivery. Even though the share of of knowledge. onshore services (that is, at the client's site over- Health services is another area in which devel- seas) in total Indian software exports has been in oping countries could become major exporters, continuous decline (in 1988, the percentage of either by attracting foreign patients to domestic on-site development was almost 90 percent), hospitals and doctors or by temporarily sending about 60 percent of Indian exports is still sup- their health personnel abroad. The Cuban gov- plied through the temporary movement of pro- ernment's strategy is to convert that country into grammers. a world medical power. SERVIMED, a trading Second, it cannot be assumed that other coun- company created by the government, prepares tries' trade policies will become progressively health tourism packages. During 1995­96, more liberal, particularly with regard to move- 25,000 patients went to Cuba for treatment, and ment of persons. In the early 1990s the U.S. gov- 1,500 students went there for training; income ernment introduced rules that obliged foreign earned from sales of health services to foreigners workers to acquire temporary work visas (H1-B was US$25 million. Again, cost savings for visas) and limited the number of visas issued dur- patients and health insurers can be significant. ing a year to 65,000. This contributed to the rela- For instance, the cost of coronary bypass surgery tive decline of onshore services by Indian firms can be as low as 70,000 to 100,000 rupees in (Heeks 1998). In 1998, in response to mounting India, about 5 percent of the cost in industrial labor shortages experienced in the U.S. informa- countries. The cost of a liver transplant in India is tion technology sector, the annual visa cap was a tenth that in the United States (UNCTAD and raised to 115,000 for both 1999 and 2000. WHO 1998). Third, significant gains can be had from further A major barrier to consumption abroad of liberalization. There are wide differences in the medical services is lack of portability of health cost of software development and support: the insurance. For instance, U.S. federal or state gov- average cost per line of code in Switzerland (the ernment reimbursement of medical expenses is most expensive country) exceeds by more than 5 limited to licensed, certified facilities in the Unit- 286 Negotiating Improved Market Access Commitments B O X 2 8 . 1 ( C O N T I N U E D ) ed States or in a specific U.S. state. The lack of the filing of any work permit or visa application. long-term portability of health coverage for Qualification and licensing requirements and the retirees from OECD countries is also a major con- regulations of professional bodies are other signif- straint on trade in health services. In the United icant barriers. The entry of foreigners can be States, for instance, Medicare, the program for impeded by nonrecognition of their professional retirees, covers virtually no services delivered qualifications, burdensome licensing require- abroad. Other nations may extend coverage ments, or the imposition of discriminatory stan- abroad, but only for limited periods such as two dards. The requirement of registration with, or or three months. This constraint is significant membership of, professional organizations can because it tends to deter some elderly persons constitute an obstacle for a person wishing to pro- from traveling or retiring abroad, and those who vide a service on a temporary basis. do retire abroad are often forced to return home to obtain affordable medical care. The potential * See the National Association of Software and Service effect of permitting portability could be substan- Companies (NASSCOM) Website, . These exports consist mainly of standardized persons living in OECD countries retired to devel- coding and testing services. oping countries, they would bring with them See . The dominance of possibly US$30 billion to US$50 billion annually on-shore delivery is attributable to, among other rea- in personal consumption and US$10 billion to sons, a reduction in information asymmetries with US$15 billion in medical expenditures (UNCTAD regard to the performance of programmers, the need and WHO 1998). for continuous client-developer interaction, and Many different barriers constrain the movement demands by Indian programmers to be sent abroad, in of natural persons. The numerous formalities part to improve their skills and expose themselves to alone (for obtaining a visa, for example) make red international markets (see Heeks 1998). tape related to FDI seem trivial by comparison. Other barriers to movement of natural persons The most obvious barriers are explicit quotas or include double taxation, wage-matching requirements economic needs tests, such as requirements that (that is, requirements that wages paid to foreign work- employers take timely and significant steps to ers be similar to those paid to nationals in that profes- recruit and retain sufficient national workers in the sion, which eliminates the cost advantage for specialty occupation and that no worker be laid foreigners), and local training requirements (to replace off for a certain period preceding and following foreign with national labor within a certain time frame). Facilitating Reciprocity across Modes mandering: all countries would liberalize access in all modes, including the movement of individuals. A collective commitment to the use of appropriate- Environmentalists and environmental services ly designed formulas offers the best chance of link- exporters could then be relied on to counter the ing different modes of delivery.6 Such formulas can opposition of employees and individual suppliers in also help overcome concerns about free-riding that the domestic environmental industry. arise in an MFN-based system. But is it technically An alternative way of creating a link between feasible to link concessions across modes? (See Sapir modes is by requiring each country to provide 1998; Thompson 2000.) One simple option is to increased "foreign labor content entitlements" to its take advantage of the current political pressure for domestic firms in relation to the country's increased accelerated liberalization in selected sectors, such as exports of services (Mattoo and Olarreaga 2001). environmental services. This approach could be Entitlements would be global rather than bilateral, accepted on the condition that there be no gerry- and the extent and pattern of use would be deter- 287 D E V E L O P I N G C O U N T R I E S A N D N E G O T I AT I O N S O N T R A D E I N S E R V I C E S mined by sound economic considerations of modal already undertaken liberalization, and the granting comparative advantage. Some of the social and of such credit relies on the unlikely generosity of political difficulties could be overcome by clarifying those who have not liberalized. The proposed ex that the liberalization is only with respect to tempo- ante assurance of credit rule has three virtues: (a) it rary movement of service suppliers and does not would help induce or enhance liberalization in imply migration. Establishing clear links between some countries between negotiating rounds; (b) it increased exports and increased foreign labor con- could lead to deeper levels of multilateral liberaliza- tent entitlements may also help make the political tion and force other countries to go further than in case. The presence of foreign workers would be seen the absence of a rule; and (c) most important, it as a direct consequence of increased opportunities does not rely on altruism to be generally acceptable. for export abroad and as contributing to the (The alternative rules are discussed more fully in increased competitiveness that makes it possible to Mattoo and Olarreaga 2001.) exploit these opportunities. GATS Article XIX.3 requires that in each future round "modalities shall be established" for the treat- ment of liberalization undertaken autonomously by Reciprocity within Modes across Sectors members since previous negotiations. In principle, It would be wrong to suggest that reciprocity must this is precisely the type of ex ante assurance of necessarily take an intermodal form. There may, for credit that would be desirable. But the nebulousness instance, be scope for cross-sectoral reciprocity in of the provision and the postponement of the estab- the same mode. Trade in electronically delivered lishment of modalities suggest that in practice the products that falls within the scope of cross-border provision may provide little more than a basis for ex supply is of growing importance and offers an post demands for credit. One way of giving the rule increasingly viable alternative to the movement of operational content would be to establish that any individuals. If the United States can supply financial agreed liberalizing formula would be applied not to and audiovisual services to the Philippines electron- current actual levels of protection but to the levels ically, the Philippines in turn can supply software bound in the previous round of negotiations.8 development and data-processing services to the United States.7 A possible formula would be for all Conclusion WTO members to agree that no restrictions would be imposed on cross-border delivery, either of all Although the most important services policy services or of a bundle the composition of which reforms need to be made at the domestic level, there could be negotiated. is substantial scope for constructive use of the mul- tilateral trading system both in realizing credible domestic liberalization and in securing market Remedying the Hold-Back Problem through a access abroad. This chapter has discussed some of Credit Rule the main issues confronting developing countries; a An undesirable aspect of an emphasis on reciproci- more comprehensive treatment can be found in ty is that it creates the temptation to hold back from Mattoo (2000, forthcoming). unilateral liberalization. This is why most econo- Certain policy choices made by developing coun- mists view reciprocity with suspicion. The holdback tries, often under negotiating pressure, are not like- problem can be overcome, however, by rules that ly to maximize domestic welfare. Examples create an ex ante assurance (at the end of a round of emphasized in this paper were "market access" con- negotiations) that credit will be given in future cessions that allow increased foreign ownership of rounds of negotiations for unilateral liberalization existing firms rather than new entry and that guar- undertaken between rounds. The impulse to liberal- antee the privileged status of foreign incumbents. ize unilaterally then need not be inhibited by fear of Where the immediate introduction of competition losing negotiating coinage. The proposed rule is dif- has not been feasible, governments have taken too ferent from the demands for credit that are typically little advantage of the GATS to lend credibility to made at the beginning of a new round of negotia- future liberalization plans. A multilateral commit- tions. The acceptance of such demands has only a ment by a government to allow entry influences the distributional effect, favoring those who have degree to which markets are contestable. Regardless 288 Negotiating Improved Market Access Commitments of the existing market structure, established suppli- operator's cost of production 26 percent. Based on these find- ers in the market are likely to behave more compet- ings, Das concludes that India's market liberalization program, which started in the mid-1990s, is justified, but he argues that itively if all policy barriers to entry are eliminated there may be a need to regulate entry in order to reduce either immediately or at a fixed date in the future. unnecessary duplication of common costs. With continued Persistent barriers to services exports by develop- improvements in technology, the fixed costs of entrants are ing countries are depriving the world of substantial likely to fall, reducing losses of scale economies and increasing welfare gains. These barriers include explicit quotas the costs of entry restrictions. whose elimination or relaxation must be negotiated 4 The fear of creating a disincentive for investors might be a rea- directly and implicit regulatory hurdles that must son to refrain from taxation. be dealt with by strengthening GATS rules on 5 This emphasis on achieving a "balance of (liberalizing) conces- domestic regulations. In particular, efforts must be sions" has led to the perception of WTO negotiations as a mer- made to break the stalemate on the movement of cantilist process driven by political forces that nevertheless individual services providers; creating "foreign leads to the desirable outcome of reduced levels of protection. labor content entitlements" is one possibility. In an important recent paper Bagwell and Staiger (1999) show that reciprocity can be given a more positive economic inter- pretation: it acts to neutralize the adverse terms of trade Notes effects associated with unilateral reductions in protection and therefore leads to greater liberalization. 1 Neither taxation nor auction addresses appropriation by exist- ing foreign share owners. In this context, grandfathering com- 6 Developing countries have resisted this option, preferring a mitments assume particular significance. request-and-offer approach. Their reluctance stems from defensive considerations and a belief that they would be 2 In Latin America, for example, countries that granted monop- obliged to concede excessively high levels of openness if a for- oly privileges to telecommunications operators of the priva- mula approach were adopted. tized state enterprises for 6 to 10 years saw connections grow at 1.5 times the rate achieved under state monopolies, but 7 There is considerable scope for an improvement in commit- that was only half the rate in Chile, where the government ments. For instance, of the more than 130 WTO members, 56 retained the right to issue competing licenses at any time have made commitments in software implementation and 54 (Wellenius 1997). in data processing. Unrestricted market access is guaranteed in only about half of these commitments and unqualified nation- 3 Interesting evidence in this context is available from the Indian al treatment in a similar share. It is particularly striking that in telecommunications sector. Das (2000) estimates a frontier the core banking services, where around 75 WTO members multiproduct cost function of the incumbent fixed-line opera- have made commitments, about a third of the developing tor covering 25 years from 1969 to 1994. The study finds very countries guarantee unrestricted cross-border supply, whereas high economies of both scale and scope in the technology only 1 of the 10 industrial countries in this group has done so. used; the parameter estimates even suggest that telecommuni- cations in India is a natural monopoly. The incumbent opera- 8 This suggestion was in fact contained in a proposal from Brazil tor, however, displays great inefficiency that increases the submitted just before the Seattle WTO ministerial conference. 289 29 C A R L O G A M B E R A L E A A D I T YA M AT T O O Domestic approach. We argue, however, that even though services sectors differ greatly, the underlying Regulations and economic and social reasons for regulatory intervention do not, Liberalization of and that focusing on these rea- sons provides the basis for the Trade in Services creation of meaningful horizon- tal disciplines. Such a generic approach is to be preferred to a sectoral approach for at least three reasons: it economizes on negotiating effort; it leads to the T creation of disciplines for all ser- rade in services, far more than trade vices sectors rather than only the politically impor- in goods, is affected by a variety of tant ones; and it reduces the likelihood that domestic regulations. A central task in multilateral negotiations will be captured by sectoral interest trade negotiations has been to develop disciplines groups. which ensure that such regulations support rather than impede trade liberalization. One basic disci- An Overview of Regulation in Services pline, the national treatment obligation, requires that regulations not discriminate in any way against The economic case for regulation in services, as in foreigners. Trade, however, can be inhibited even by goods, arises essentially from market failure attrib- regulations that do not discriminate, such as certain utable to three kinds of problems: natural monop- standards and licensing requirements, and by the oly or oligopoly, asymmetric information, and absence of procompetitive regulations. Although externalities. The social case for regulation is based important initiatives have recently been taken to primarily on considerations of equity. Figure 29.1 remedy these problems in the areas of accountancy provides an overview of the main reasons for and and telecommunications, the overall disciplines on the forms of regulation in different services sectors. domestic regulations in the General Agreement on The existence of natural monopoly or oligopoly is Trade in Services (GATS) remain weak. As we a feature of the so-called locational services (UNC- approach the next round of services negotiations, TAD and World Bank 1994). Such services require, the question arises as to whether it is best to rely on first of all, specialized distribution networks: roads further sectoral initiatives or whether it is possible and rails for land transport, cables and satellites for to adopt a more general approach. communications, and pipes for water supply and The diversity of services sectors and the difficulty sanitation and for energy distribution. They may of making certain policy-relevant generalizations also require specialized equipment for transmitting have tended to encourage a sector-specific or receiving the service: railway stations and bus ter- 290 Domestic Regulations and Liberalization of Trade in Services 29.1 Approaches toward Domestic Regulation at the Multilateral and National Levels Market failure Services sector Multilateral approach Action required at national level Monopoly; Network services: Generalize key disciplines Develop procompetitive oligopoly transport (terminals in telecommunications regulation to protect consumer and infrastructure); Reference Paper to ensure interests where competitive environmental services cost-based access to market structures do not exist. (sewerage); energy essential facilities (roads, services (distribution railroad tracks, terminals, networks). sewers, pipelines, etc.). Strengthen Article IX to deal with international cartels. Asymmetric Intermediation and Exercise nondiscrimination Strengthen domestic regulation information knowledge-based and generalization in to remedy market failure in an services: financial applying the "necessity" economically efficient manner. services, professional test. Use the test to services, etc. create a presumption in favor of choice of economically efficient policies for remedying market failure. Externalities Transport, tourism, etc. Social objectives: Transport, telecom- Devise economically efficient universal service munications, financial, means of achieving social education, health. objectives in competitive markets. minals, seaports, airports, telephone exchanges. that all service providers, domestic and foreign, One reason for the tendency toward monopoly or have access to the essential facilities that continue to oligopoly is the difficulty of duplicating networks be monopolistically controlled. and terminals, given space constraints. A second The problem of asymmetric information occurs reason is the high barriers to entry associated with in a wide range of intermediation and knowledge- large initial investments. It must, however, be kept based services (UNCTAD and World Bank 1994). in mind that recent technological developments in Buyers are often inadequately informed about the areas such as telecommunications are leading to the true attributes of sellers. Thus, consumers cannot emergence of relatively small optimal scales of pro- easily assess the competence of professionals such duction and are overturning conventional wisdom as doctors and lawyers, the safety of transport ser- about the inevitability of monopoly. vices, or the soundness of banks and insurance Liberalization in locational services could imply companies. In principle, the adequate dissemina- two types of change. First, the monopoly itself tion of information could remedy the problem, needs to be delineated as narrowly as possible so but it may be too expensive to communicate the that competition is introduced where feasible. For necessary information to individual buyers. In instance, in railways a monopoly track owner could such situations, it may be easier to regulate suppli- sell track services to separate, competing operators ers than to educate consumers. The imposition of of trains. Second, competition could be introduced minimum regulatory conditions on suppliers for the right to provide the remaining monopoly reflects a certain uniformity of preferences among services. Thus, the right to provide track services consumers about the quality of services. Thus, reg- could be auctioned off to the firm that commits to ulators ensure that all banks meet a certain thresh- supply the services at the lowest price. In the current old of financial soundness and professionals a context, the interesting issue arises when services certain threshold of competence (see Box 29.1). that can be provided competitively must rely on ser- Regulating the output of a services industry, which vices provided monopolistically--either by govern- is often invisible and customized, is usually more ments or by private firms. The challenge is to ensure difficult than regulating inputs. Such regulation of 291 D E V E L O P I N G C O U N T R I E S A N D N E G O T I AT I O N S O N T R A D E I N S E R V I C E S B O X 2 9 . 1 W H O S E R E G U L AT I O N S , A N D F O R W H AT P U R P O S E ? S T I C K Y I S S U E S I N E L E C T R O N I C C O M M E R C E Domestic regulations that affect trade pose the household (Glassman 2000). Of course, reporting main challenge to ensuring open conditions for of personal credit histories is critical to consumer electronic delivery of services. Two examples illus- credit, and even in theory, excessively strict privacy trate the difficulty of distinguishing between reg- laws could create significant asymmetries of infor- ulations that incidentally impede trade in the mation and affect the efficiency of markets pursuit of legitimate objectives and regulations (Kitchenman 1999). This is not to suggest that that deliberately discriminate against foreign pro- there might not be good reasons to protect priva- vision for the sake of protection. cy. The desired level of such protection, however, may differ across countries, and if trade is made Privacy conditional on the existence of "comparable" laws, An issue that could have a profound effect on a socially costly "race to the top" might result. electronic commerce is privacy. In late 1998 the European Union (EU) issued a wide-ranging direc- Offshore Financial Services tive aimed at safeguarding the privacy of EU citi- A number of Caribbean countries have become zens' personal data and preventing its misuse offshore financial services centers and have bene- worldwide. The directive is backed by the power fited significantly as a result. In recent years con- to cut off data flows to countries that the EU cern has been expressed about certain aspects of judges not to have adequate data protection rules their tax and regulatory regimes, and there is sus- and enforcement. The directive caused friction picion that some centers facilitate money launder- with the United States, which accused the EU of ing. For instance, the Financial Stability Forum, trying to impose laws beyond its own frontiers. A which assesses conformity with the core Basle Prin- compromise was reached under which the United ciples for Effective Banking Supervision, placed States agreed to set up arrangements for the pro- most of the Caribbean offshore centers in the low- cessing by companies of personal data from the est category, and the Financial Action Task Force, EU, but the issue has not been fully resolved. which has established a list of 40 criteria, placed The issue could have an impact on developing most of the Caribbean centers on its blacklist. The countries' exports of data-processing services, and it countries have also attracted the attention of the poses a difficult choice for these countries. If they OECD for supposedly harmful tax practices. choose not to enact laws that are deemed adequate, Although the regulatory concerns may have they could be shut off from participation in this been legitimate, interviews with regulators both growing market. In the absence of such laws, and within and outside the Caribbean countries raise given the weakness of local legal systems, it might be some doubts about the fairness of the assessment difficult for private firms in developing countries to processes. First, the standards were unilaterally emulate U.S. firms like Microsoft and credibly com- determined, and while that may be defensible in mit to meet the required high standards. this sector, the process of establishing conformity If the countries do enact stringent laws, it is was not transparent. Second, no criteria were spec- unlikely that these laws could be formulated to be ified as to how a country that had been blacklisted specific to trade with particular jurisdictions, and so could improve its regulatory standards and be the result could be an economywide increase in the removed from the list. Finally, and perhaps of great- costs of doing business. For instance, according to est concern to the Caribbean countries, the claims private sector estimates generated in the United of harmful tax practices raised issues of sovereignty States, information sharing saves the customers of over tax policy and of discrimination in favor of 90 financial institutions (accounting for 30 percent European offshore financial centers that were not of industry revenues) US$17 billion a year (US$195 targeted despite the similarity of their tax regimes. per average customer household) and 320 million hours a year, or 4 hours per average customer Source: Prepared by the volume editors. 292 Domestic Regulations and Liberalization of Trade in Services inputs, however, often amounts to restrictions on regulatory intervention provides a basis for the cre- entry into the market. ation of meaningful horizontal disciplines. As is argued in UNCTAD and World Bank (1994: Market failure due to natural monopoly or oli- 42), "service providers are likely to prefer the higher gopoly may create trade problems because incum- incomes that result from control of entry into their bents can impede access to markets in the absence of occupation, or from restrictions on competition appropriate regulation. Because of its direct impact between those who are admitted to it . . . whenever on trade, this is the only form of market failure that regulation is judged necessary, a major concern needs to be addressed directly by multilateral disci- must be to ensure that regulatory powers are not plines. The relevant GATS provision, Article VIII captured by the existing providers of a service and (dealing with monopolies), is limited in scope. As a used to further their interests." The difficulty in consequence, in the context of the telecommunica- imposing border restrictions on services, which tions negotiations, the Reference Paper described in rarely take a visible form, makes regulation a partic- Box 29.2, with its competition principles, was devel- ularly attractive means of protecting domestic sup- oped to ensure that monopolistic suppliers would pliers from foreign competition. not undermine market access commitments. The The problem of externalities arises when market first element of our proposal is that these principles prices do not fully capture the costs and benefits of should be generalized to a variety of other network the associated transactions. Typical examples are the services, including transport (terminals and infra- negative environmental externalities generated by structure), environmental services (sewerage), and providers of transport services or tourism services. energy services (distribution networks). This is to be done by ensuring that any major supplier of essential facilities provides access to all suppliers, A Possible Way Forward under the GATS national and foreign, at cost-based rates. Even though services sectors differ greatly, focusing Anticompetitive practices that fall outside the on the underlying economic and social reasons for jurisdiction of national competition law may be B O X 2 9 . 2 A S S E S S I N G T H E W T O T E L E C O M M U N I C AT I O N S R E F E R E N C E PA P E R Carsten Fink competition from new entrants in otherwise open services markets. This fear was compound- A key feature of the WTO Agreement on Basic ed by the fact that telecommunications markets Telecommunications--a post­Uruguay Round in many member countries were still dominated agreement concluded in 1997--is the Reference by fully or partially state-owned entities, while Paper that sets out regulatory principles in the the separation of regulatory, policy, and opera- telecommunications sector. Sixty of the 69 mem- tional responsibilities had not progressed to the bers that made specific market access and same point. national treatment commitments under the The Reference Paper covers principles in six agreement subscribed to the Reference Paper in areas: competitive safeguards, interconnection, its "blueprint" form or adopted a modified ver- universal service, public availability of licensing sion of it. The acceptance of regulatory disci- criteria, independent regulators, and allocation plines on telecommunications under the GATS and use of scarce resources. The provisions on agreement was unprecedented. WTO members competitive safeguards require members to pre- felt there was a need for regulatory principles to vent major suppliers from engaging in anticom- ensure effective market access. Specifically, there petitive cross-subsidization and from abusing was concern that the lack of procompetitive reg- control of information. Arguably, some of the ulation on key matters such as network intercon- most significant obligations concern network nection and number portability could forestall interconnection, which (among other require- (continued) 293 D E V E L O P I N G C O U N T R I E S A N D N E G O T I AT I O N S O N T R A D E I N S E R V I C E S B O X 2 9 . 2 ( C O N T I N U E D ) ments) must take place on nondiscriminatory, the agreement did not adopt the Reference Paper transparent, and reasonable terms and at cost- in full but excluded certain provisions. Arguably, oriented rates. The provisions regarding inde- these countries were concerned that their exist- pendent regulators require the regulatory body ing regimes would be inconsistent with the obli- to be impartial, separate from, and not account- gations set out in the paper. Second, the able to any services supplier. experience from two disputes pertaining to inter- To what degree do the Reference Paper's obli- connection prices charged by the dominant gations bite? Can the paper live up to the expec- incumbents in Mexico and Japan suggests that tation of ensuring effective market access? On the prospect of WTO arbitration can contribute the one hand, the regulatory principles lack pre- to the implementation of procompetitive regula- cision in many respects and appear to leave room tion. In the case of Mexico, the United States ini- for discretionary decisionmaking by national reg- tiated WTO dispute settlement proceedings and ulators. It seems difficult, for example, to derive reserved its right to press its case even after the far-reaching obligations from vague language dominant Mexican carrier agreed to a reduction such as "reasonable" terms of interconnection or in interconnection fees. In the case of Japan, the "appropriate measures" to prevent abusive busi- United States threatened the initiation of a WTO ness practices. Similarly, the obligation to provide complaint, and Japan subsequently agreed to interconnection at cost-oriented rates can only be lower interconnection charges substantially. limited in scope, as no reference is made to a spe- Although these two examples undoubtedly cific methodology or definition of network cost. point to the relevance of the principles set out in Thus, taken at its face value, it would seem that the Reference Paper, one has to keep in mind that the Reference Paper prevents only the most egre- in both cases the lowering of interconnection gious departures from procompetitive regulation rates was the immediate result of bilateral pres- (Bronckers and Larouche 1997). sure applied by the United States. Moreover, the On the other hand, two factors suggest that regulatory principles have not been critically test- the Reference Paper's obligations have teeth. ed, in the sense that so far no WTO panel has First, as already pointed out, several signatories of arbitrated on the basis of the Reference Paper. important in sectors such as maritime, air trans- The second obligation is necessary to deal with the port, and communications services.1 The current possibility of inadequate enforcement by public GATS provision in this area (Article IX) provides agencies and already has a precedent in the WTO only for information exchange and consultation. rules on intellectual property and government pro- Strengthened multilateral rules are needed to reas- curement. sure small countries with weak enforcement capaci- In all other cases of market failure, multilateral ty that the gains from liberalization will not be disciplines do not need to address the problem per appropriated by international cartels. For instance, se; rather, they should ensure that domestic mea- two obligations could be created. The first would sures for dealing with the problem do not unduly require an end to the exemption from national restrict trade. (The same is true for measures competition law of collusive agreements that affect designed to achieve social objectives.) Such trade- other countries. (Both the United States and the restrictive effects can arise from a variety of technical European Union currently exempt maritime con- standards, prudential regulations, and qualification ferences from the scope of their competition law.) requirements in professional, financial, and numer- The second would create a right of foreign con- ous other services (see Box 29.3), as well as from sumers to challenge anticompetitive practices by grants of monopoly rights to complement universal shipping lines in the national courts of countries service obligations in such services as transport and whose citizens own or control these shipping lines. telecommunications. 294 Domestic Regulations and Liberalization of Trade in Services B O X 2 9 . 3 R U L E S F O R D O M E S T I C R E G U L AT I O N S : T H E E X P E R I E N C E W I T H A C C O U N TA N C Y John Hegarty The necessity rule is without doubt the most substantive provision in the accountancy disci- In the course of the Uruguay Round negotiations, plines. The rest of the disciplines either add some accountancy services received significant atten- limited value to existing GATS articles (such as tion. Accountancy makes a useful case study of Art. III, on transparency) or specify the applica- how liberalizing trade in services requires atten- tion of the necessity test to types of measure. In tion to "behind-the-border" issues, in that the some key cases, however, the specific rules con- main barriers to trade derive from domestic regu- tain language weaker than the necessity test. lations governing the sector. For example, the provision on residency After the Uruguay Round, in March 1995, the requirements says that members "shall consider" Council for Trade in Services established the whether means less restrictive of trade could be Working Party on Professional Services (WPPS) to employed to achieve the same policy objective, develop the disciplines necessary to ensure that and the provision on qualification requirements specific regulatory measures did not constitute states, "A Member shall ensure that its compe- unnecessary barriers to trade. As a matter of pri- tent authorities take account of qualifications ority, the WPPS was requested to make recom- acquired in the territory of another Member, on mendations for multilateral disciplines in the the basis of equivalency of education, experience accountancy sector. In December 1998 the coun- and/or examination requirements" (emphasis cil adopted the "Disciplines on Domestic Regula- added). A proposal to create a presumption in tion in the Accountancy Sector" that had been favor of a test of competence as the least trade submitted to it by the WPPS. restrictive measure received little support. The disciplines on accountancy represent an The disciplines are also subject to three con- important step forward in the definition of rules straints that limit their impact. First, they apply on domestic regulation under the GATS because only to those WTO members that have made they contain a binding "necessity test." Members commitments in the accountancy sector, not to are required to all members. Second, they have no immediate application but enter into force only when for- ensure that measures not subject to sched- mally integrated into the GATS, which requires uling under Articles XVI or XVII of the waiting until the current negotiations have been GATS, relating to licensing requirements concluded. Third, and most important, they do and procedures, technical standards and not apply to measures that are subject to sched- qualification requirements and procedures uling under Articles XVI and XVII. In other words, are not prepared, adopted or applied with a domestic regulation that has the effect of deny- a view to or with the effect of creating ing market access or national treatment is not unnecessary barriers to trade in accountan- automatically negated when the disciplines enter cy services. For this purpose, Members into force but must be addressed through future shall ensure that such measures are not negotiations of specific commitments. This came more trade-restrictive than necessary to as a disappointment to those who had hoped to fulfil a legitimate objective. Legitimate see at least some "behind-the-border" issues objectives are, inter alia, the protection of dealt with by way of generally applicable rules. consumers (which includes all users of There are several reasons for the relatively mod- accounting services and the public gener- est result in accountancy. The interest in these ally), the quality of the service, professional negotiations of those who had a real stake in lib- competence, and the integrity of the pro- eralization was somewhat diminished when it fession. ("Disciplines on Domestic Regula- became apparent that the more explicit barriers tion in the Accounting Sector") could not be addressed. To a significant extent, (continued) 295 D E V E L O P I N G C O U N T R I E S A N D N E G O T I AT I O N S O N T R A D E I N S E R V I C E S B O X 2 9 . 3 ( C O N T I N U E D ) there was also a lack of involvement and commit- strengthening. The World Bank, for example, has ment on the part of the ministries and agencies supported moves to improve the functioning of responsible for the domestic regulatory regimes the relevant international standard-setters, the governing accountancy, which were not con- IASC and the International Federation of Accoun- vinced of the need to make changes in response tants (IFAC), and the quality of the standards they to trade policy considerations. Most systems of produce. In May 2000 IOSCO endorsed interna- national regulation have evolved to address tional accounting standards (IASs), and in June national concerns about protecting the interests 2000 the European Commission announced its of national users and providers of accountancy intention to propose legislation to make the use services, and the need to take the international of IASs mandatory for all listed companies in the dimension into account was not widely enough EU by 2005, at latest. This should dispel the hesi- recognized to enable positive progress during the tations with respect to IASs that may have existed Uruguay Round. To the extent that accountancy during the Uruguay Round. regulators did wish to address international Given that much of the progress in accountancy- issues, they chose other fora for doing so, and related fields is taking place outside the context of there was a reluctance to have such matters cov- the WTO, why should the current negotiations ered as part of trade talks as well. Specifically, address trade in accountancy services rather than there was hesitation about submitting to a leave it to other organizations and fora? There are process that could result in binding commit- several reasons. ments to reform, given a strong preference for voluntary and unilateral (albeit coordinated) 1. When domestic reforms are anchored in changes in this field. This was the case with binding commitments under the GATS, this regard to accounting standards, for example, provides legal certainty and enhances exter- where negotiations between the International nal confidence in the durability and irre- Accounting Standards Committee (IASC) and the versibility of the changes. International Organization of Securities Commis- 2. Given that the negotiations touch on many sions (IOSCO) had not yet reached the stage related regulatory issues, such as the rules where IOSCO felt able to endorse international governing financial services, coverage of accounting standards. accountancy services allows negotiators to In recent years a stronger understanding has ensure consistency of approach and to avoid emerged of the contribution of accountancy ser- unintended conflicts. vices to creating a sound investment climate. 3. While it is recognized that other organiza- There is also increased recognition of the interna- tions and fora have mandates of their own, tional and systemic externalities of weaknesses in the WTO does provide a useful platform for domestic regimes for the regulation of accoun- discussing enhanced cooperation between tancy. This has led to several international initia- countries on regulatory issues and for ensur- tives to improve global financial stability and the ing that the trade dimension is taken appro- global financial architecture, and accounting and priately into account when national reforms auditing are included in the areas identified for are being considered. The second element of our proposal is that the leaves governments free to deal with economic and trade-inhibiting effect of this entire class of regula- social problems, provided that any measures taken are tions is best disciplined by complementing the not more restrictive of trade than is necessary to national treatment obligation with a generalization of achieve the relevant objective. The test is already part the so-called "necessity" test. This test essentially of the recently established disciplines in the accoun- 296 Domestic Regulations and Liberalization of Trade in Services tancy sector. It would seem desirable to use the test to asymmetric information, and tradeoffs between create a presumption in favor of the choice of eco- efficiency and equity. nomically efficient policies in remedying market fail- ure and in pursuing noneconomic objectives (see Dealing with Monopolies Mattoo and Subramanian 1998). For instance, in the case of doctors and similar professionals, a require- The telecommunications Reference Paper illustrates ment to requalify would be judged unnecessary, since both the strengths and the limitations of the multi- the basic problem--inadequate information about lateral approach. The primary concern of the paper, whether the persons possess the required skills-- as of WTO rules in general, is to ensure effective could be remedied by a less burdensome test of com- market access; hence the focus on the terms of petence. interconnection. The paper does not address wider In sum, the telecommunications and accountancy concerns about consumer interests and how they models, suitably developed and generalized, can may be affected by monopolistic behavior. together ensure that domestic regulations achieve Although there can be little doubt that price deter- their objectives without sacrificing economic effi- mination is ideally left to competitive markets and ciency. This is not to say that there is no need for sec- that regulatory price-setting is fraught with difficul- tor-specific disciplines. For instance, there is valuable ties, regulatory authorities in developing countries work that could be done to establish how best to deal where competition is slow to develop need to equip with asymmetric information and differences in themselves, legally and technically, with the ability standards between countries. But a useful beginning to regulate prices.2 This would seem particularly can be made by taking a cross-sectoral approach. desirable in countries like some of those in the Caribbean area that have locked themselves into exclusive supply contracts with a single telecommu- The Regulatory Challenge at the National nications provider well into this century. It is Level important to note that although nothing in the The development of multilateral disciplines is in no GATS prevents a country from exercising any form way a substitute for strengthening domestic regula- of procompetitive regulation provided that it is not tory mechanisms and institutions. At least three discriminatory, the capacity of most developing areas are of considerable importance: monopolies, countries in this respect is limited (see Box 29.4). B O X 2 9 . 4 C H A L L E N G E S I N I M P L E M E N T I N G P R O C O M P E T I T I V E R E G U L AT I O N It is now widely recognized that in basic telecom- ed licenses for both basic and cellular services munications, procompetitive regulation is need- while remaining the main telecommunications ed to deliver effective competition and gains services provider. In the absence of an indepen- from liberalization. But the experience of differ- dent regulator empowered to rebalance tariffs, ent countries reveals a number of political and enforce fair interconnection agreements, and economic difficulties that are only gradually ensure rapid, equitable allocation of the radio being overcome. spectrum, the benefits of opening the sector to In India a conflict between the Department of allow private participation and foreign invest- Telecommunications (DOT) and the regulatory ment were significantly limited. agency, the Telecommunications Regulatory On March 26, 1999, the Indian government Authority of India (TRAI), as initially constituted, announced a new telecommunications policy hampered progress toward an efficient telecom- that addressed several of these outstanding munications infrastructure. Underlying a number issues. DOT's policymaking and service provision of these problems was DOT's dual role: it award- functions were separated, and its operations arm (continued) 297 D E V E L O P I N G C O U N T R I E S A N D N E G O T I AT I O N S O N T R A D E I N S E R V I C E S B O X 2 9 . 4 ( C O N T I N U E D ) was corporatized. TRAI was reconstituted in In response to these problems, in May 2000 St. 2000, and its dispute resolution powers are now Lucia, Dominica, Grenada, St. Vincent and the vested in a new quasi-judicial agency. The author- Grenadines, and St. Kitts and Nevis set up, with ity announced a decision on telephone tariffs that World Bank support, the Eastern Caribbean will substantially restructure telephone service Telecommunications Authority (ECTEL), the first prices over a three-year period, significantly regional telecommunications authority in the improving incentives for investment in local net- world. ECTEL is in the process of developing from works. The regulator has also programmed an a legal entity into a functioning institution. agenda of activities to address several other Although the member countries will retain their important regulatory matters, such as intercon- sovereign power over licensing and regulation, nection arrangements, the numbering plan, qual- ECTEL will provide technical expertise, advice, and ity of service, business rules, and customer support for national regulations. In addition to the satisfaction. economies of scale in establishing a common reg- For smaller countries, a different problem arises: ulator, there are at least three other advantages: the creation and operation of an efficient regulato- the arrangement will promote the development ry agency involves substantial fixed costs that could of harmonized and transparent regulation in the place a significant burden on resources. Apart from region, allow for a greater degree of indepen- spectrum-monitoring equipment, computers, and dence (and hence credibility) in regulatory advice, programs, there is the cost of professional assis- and enhance bargaining power in negotiations tance for activities such as interconnection, cost with incumbents and potential entrants. In fact, estimation, and spectrum management. For exam- there is evidence that the creation of ECTEL, along ple, the total cost of government in Dominica is with other reforms, has already prompted a US$41 million a year, whereas the budget of the decline in the prices of telecommunication ser- U.S. telecommunications regulator, the Federal vices in the region. For example, the per-minute Communications Commission, runs to US$210 cost of a daytime call to the United States has fall- million a year. It is estimated that even a bare-bones en between 24 and 42 percent in these countries. regulatory authority is likely to cost about US$2 million each year, or 5 percent of Dominica's gov- Source: Prepared by the volume editors, based on ernment budget. DeFreitas, Kenny, and Schware (2001). Dealing with Asymmetric Information in domestic regulation can legitimize external barri- ers to trade. A further twist is that domestic con- The need for effective regulation of financial ser- sumers may actually prefer cheap, low-quality vices needs no elaboration, particularly in light of products. The question of how best to meet the the recent experiences of many countries. Again it is needs of export markets given domestic preferences incumbent on the countries themselves to create regarding quality is clearly an area in which much adequate mechanisms for such regulation--some- more research is needed. thing that is clearly necessary if a country is to ben- efit fully from liberalization (see Box 29.5). Managing Conflicts between Efficiency and Equity Other areas in which the inadequacy of regulatory mechanisms for dealing with asymmetric informa- Trade policy affects the poor through three main tion is a problem have received relatively less atten- channels: through the price of their consumption tion. For instance, in professional services, low bundle; through demand for their labor and other standards and disparities in domestic training and assets they own, which has consequences for their examinations can become a major impediment to income; and by bringing about changes in govern- obtaining foreign recognition. Thus, inadequacies ment revenue and therefore in the government's 298 Domestic Regulations and Liberalization of Trade in Services B O X 2 9 . 5 F I N A N C I A L S E C T O R L I B E R A L I Z AT I O N : T H E N E E D F O R P O L I C Y C O H E R E N C E Stijn Claessens sector framework by creating a constituency for improved regulation and supervision, better dis- Financial reform is an especially complicated sub- closure rules, and improvements in the legal and ject. It is useful, to begin with, to distinguish regulatory framework for the provision of finan- three types of financial liberalization and the cial services. It also added to the credibility of scope of each. rules. These benefits of opening up to foreign entry followed from both top-down actions on · Domestic financial liberalization allows market the part of government and bottom-up pressures forces to work by eliminating controls on from the market, as best international practices lending and deposit rates and on credit allo- and experiences were introduced. cation and, more generally, by reducing the Although the two reform processes--interna- role of the state in the domestic financial sys- tionalization and domestic financial deregula- tem. tion--are mutually reinforcing, they are not · Capital account liberalization removes con- sufficient in themselves. More than in other sec- trols on the movement of capital in and out tors, the gains and costs of financial reform of the country and does away with restric- depend on the regulatory and supervisory frame- tions on the convertibility of currency. work (Barth, Caprio, and Levine 2001). Experi- · Internationalization of financial services elimi- ence shows that it is vital to strengthen the nates discrimination in treatment between supporting institutional framework in parallel foreign and domestic financial services with domestic deregulation and internationaliza- providers and removes barriers to the cross- tion. In the absence of such strengthening, for- border provision of financial services. eign entry may entail risks. Foreign bank entry can destabilize local banks by drawing away the Internationalization has given rise to fears con- lowest-risk business--including large exporting cerning the survival of local banks and financial firms--leaving local banks to venture farther out companies, the loss of monetary autonomy, and on the risk frontier. And several countries, espe- the increased volatility of capital flows. Many of cially in Africa, discovered with the failure of these concerns relate not just to internationaliza- banks like BCCI and Meridien that a foreign name tion of financial services but also to the processes did not necessarily guarantee safety and sound- of financial deregulation and capital account lib- ness even when these foreign banks were operat- eralization. The magnitude of the benefits and ing in industrial economies or had some the costs of internationalization depend, to a ownership links with reputable foreign sources. great extent, on how internationalization is The need for a supportive institutional frame- phased in with these other two types of financial work is even more obvious when it comes to cap- reform and, in particular, with the strengthening ital account liberalization. Experience in recent of prudential regulation and supervision. years, most recently in Asia, has shown that being Many countries that have had successful expe- able to achieve the potential gains, while avoid- riences with opening up to foreign financial firms, ing the risks, of capital account liberalization including Brazil, Chile, Hungary, Ireland, Poland, depends to a great extent on whether domestic Portugal, and Spain, also engaged in a process of institutions and prudential authorities have devel- domestic deregulation and reaped substantial oped sufficiently to ensure that foreign finance gains (World Bank 2001a). The experience of the will be channeled in productive directions countries acceding to the European Union sug- (Eichengreen 2001). Recent experience also gests that internationalization and domestic shows the potential benefits of foreign financial deregulation can be mutually reinforcing. institutions in stabilizing capital flows. Several Increased foreign entry bolstered the financial countries with significant foreign presence, such (continued) 299 D E V E L O P I N G C O U N T R I E S A N D N E G O T I AT I O N S O N T R A D E I N S E R V I C E S B O X 2 9 . 5 ( C O N T I N U E D ) as Argentina and Mexico, benefited from the Caprio, and Levine 2000; La Porta, Lopez de access of these institutions to foreign capital dur- Silanes, and Shleifer 2000). Insofar as foreign ing periods of economic difficulty (Dages, Gold- presence leads to a stronger regulatory and berg, and Kinney 2000). More generally, studies supervisory framework, it contributes to making show that diversity in ownership contributes to capital account liberalization and international- greater stability of credit in times of crisis (Barth, ization mutually reinforcing. ability to finance programs for the poor. The first prices decline, prices for certain key end-users may step in the assessment of the impact of a change in actually increase. trade policy on the well-being of the poor is to iden- Despite these considerations, there is evidence of tify what the poor consume and what they own-- a positive relationship between competitive market type of labor, skills, and other assets. structures and services expansion in sectors such as If a country is a relatively inefficient producer of a basic telecommunications. This is especially true in service, liberalization and the resultant foreign com- countries where initial conditions are feeble, as petition are likely to lead to a decline in domestic exemplified by low teledensity or by service prices and to improvements in quality. Insofar as the rationing (long waiting lists for obtaining connec- poor are consumers of these services, they are likely tions). Simply letting the market work can substan- to benefit. But there is a twist: frequently, before lib- tially improve access in an environment where eralization, prices are not determined by the market services have been traditionally provided by ineffi- but are set administratively and are kept artificially cient public monopolies--even in the poorest low for certain categories of end-users or certain countries and among low-income consumers. For types of services product. Thus, rural borrowers may example, women entrepreneurs in Bangladesh vil- pay lower interest rates than urban borrowers, and lages provide pay phone services at a profit, using prices of local telephone calls and of public trans- mobile cellular technology. Even though rural vil- port may be kept lower than the cost of provision. lagers cannot afford a phone individually, they can (Box 29.6 summarizes a study of the effect of finan- afford one collectively (Lawson and Meyenn 2000). cial liberalization on poor people's access to credit.) These examples suggest that the important ques- Sometimes the object is to ensure access for all con- tions are these: In which sectors is there a conflict sumers at the same price, irrespective of the cost of between efficiency and equity? How can the conflict provision (for example, in transport and postal ser- be best addressed? What is the current situation? (If vices). Or, the object may be to ensure cheaper access the status quo reflects redistribution in favor of to, say, financial services for certain categories of vested interests--as, for example, the structure of user. This price structure is often sustained through telecommunications and transport prices often cross-subsidization within public monopolies or favors the urban middle class, and the pattern of through government financial support. directed lending programs favors the politically Unless special regulatory measures are taken, lib- powerful--a move to the market can provide both eralization threatens these arrangements. Elimina- efficiency and equity benefits. Precommitting to the tion of restrictions on entry implies an end to market can serve as a device to preclude private cap- cross-subsidization because it is no longer possible ture.) A more general question is, how can distribu- for firms to make extranormal profits in certain tive mechanisms achieve their goals in a world of market segments. New entrants may focus on the unequal economic and political power? Which most profitable market segments, such as urban redistributive instruments lend themselves to cap- areas, where network costs are lower and incomes ture by the politically powerful? are higher. And privatization could mean the end of government support. The result is that even though Ensuring Universal Service. Achieving the desir- the sector becomes more efficient and average able social objective of universal service in an eco- 300 Domestic Regulations and Liberalization of Trade in Services B O X 2 9 . 6 F I N A N C I A L L I B E R A L I Z AT I O N A N D A C C E S S T O C R E D I T B Y T H E P O O R A study by Paul Mosley empirically estimates the digestible by commercial banks." Informal sector impact of financial liberalization on access to rural and rural lending is still carried out by traditional credit in four African countries: Kenya, Lesotho, moneylenders, nongovernmental organizations, Malawi, and Uganda. Domestic financial liberaliza- and government agencies. tion via decontrol of interest rates and removal of Mosley's study also shows that financial reform credit subsidies has led to higher real interest rates in the forms of financial innovation in rural areas but, except in Uganda, has not translated into a and development of financial institutions catering higher level of savings or increased access to rural to the poor has had strong and significant effects credit. In Uganda average annual estimated credit in improving access to rural credit and lowering disbursements to the agricultural sector showed a poverty. Some examples of successful microfi- marginal increase, from US$116 million in the nance institutions in Africa are PCEA Chogoria three years before liberalization to US$123 million and the Rural Enterprise Program in Kenya and in the three years after liberalization. In Malawi, by the CCEI/Gatsby Trust Scheme in Cameroon. contrast, agricultural credit disbursements Non-African examples include BancoSol in Bolivia declined from US$121 million in the three years and Bank Rakyat Indonesia. These microfinance before liberalization to US$109 million in the three institutions offer savings and credit services on years after liberalization. Using sample survey data, commercial terms to marginal households and Mosley found that between 1992 and 1997 the use peer pressure as a substitute for collateral in percentage of sampled households with access to loan repayments and recovery. They have man- rural credit rose from 13.1 to 25 percent in Kenya aged to sustain high loan recovery rates, cover and from 9.2 to 21 percent in Uganda but that in costs, and make profits. Their lending rates lie in Malawi the share declined from 12 to 8 percent. between those of commercial banks and informal It is also worth noting that access to credit by moneylenders. Formal institutions can also mobi- the poorest income decile remained unchanged lize deposits and allocate credit to small borrow- in Kenya and Uganda but declined after liberal- ers by forging links with informal and ization from 1.9 to 0.9 percent in Malawi and microfinance agents, thereby lowering informa- from 2 to 1.9 percent in Lesotho. Commercial tion costs and developing community-based con- and foreign banks have been reluctant to move tract enforcement mechanisms. It should be into informal lending despite the higher rates in noted that simply privatizing state microfinance that sector. Mosley observes that this is probably agencies has proved disastrous, as illustrated by "due to high levels of subjective risk, supplement- the collapse of Malawi's MRFC. ed with ignorance and a shortage of individuals able to act as go-betweens and present the finan- Source: Prepared by the volume editors, based on cial results of microfinance institutions in a form Mosley (1999). nomically efficient way is a major challenge for Historically, governments have frequently relied national policymakers. The manner in which they on public monopolies to pursue (often unsuccess- pursue this objective is likely to have a profound fully) the goal of universal service, either through effect on trade in a variety of areas, including cross-subsidization across segments of the market financial, transport, telecommunications, health, or through transfers from the government or from and education services. The government can government-controlled banks. In addition to the choose from a spectrum of policies that range inefficiencies created by monopolistic market struc- from completely replacing the market to installing tures, the burdens these obligations imposed on supportive policies that affect the market outcome existing national suppliers are even now a signifi- (see Box 29.7). cant impediment to liberalization in many coun- 301 D E V E L O P I N G C O U N T R I E S A N D N E G O T I AT I O N S O N T R A D E I N S E R V I C E S B O X 2 9 . 7 P U R S U I N G S O C I A L O B J E C T I V E S I N C O M P E T I T I V E T E L E C O M M U N I C AT I O N S M A R K E T S Charles Kenny and the establishment of an independent regula- tor. The mobile telephone market was partially There is growing evidence that private competi- liberalized at once and is now fully competitive. tive markets can rapidly enhance the level of The telecommunications privatization con- access to basic services. For example, in 1993 the tracts included substantial obligations to install Peruvian government embarked on a major public telephones in rural areas, supporting an reform of its telecommunications sector, support- increase in the number of public phones country- ed by the World Bank. New laws enacted in 1993 wide from 8,000 to 50,000 over the 1993­98 and 1994 provided for the privatization of the period. As the table shows, improvements were two state-owned telecommunications utilities seen in many other areas, as well. 1993 1998 Sector investments (millions of U.S. dollars) 28 2,099 Fixed lines Number 660,000 1,850,000 Penetration rate per 100 population (percent) 2.9 7.5 Mobile telephone lines Number 50,000 600,000 Penetration rate per 100 population (percent) 0.2 2.4 Towns with phone service (number) 1,450 3,000 Poor households in Lima with a telephone (percent) 1 21 Average waiting time for connection 118 months 45 days Connection fee (U.S. dollars) 1,500 150 In 1999 a universal service levy of 1 percent The scheme in Peru was modeled on a similar charged on telecommunications operators' gross universal access program in Chile, where just over revenues was introduced to finance a telecommuni- US$2 million in public funds leveraged US$40 cations fund (FITEL) dedicated to meeting universal million in private investment to install telephones access objectives in areas that remained out of the in 1,000 localities, at about 10 percent of the cost reach of the network, as defined through socioeco- of direct public provision. Through a combina- nomic studies and consultation with the local tion of sector liberalization and competitive pro- authorities and population. Funds were allocated vision of subsidized public access to areas not through a competitive bidding process. The process previously served, household ownership of a tele- encouraged operators to look for the best technolo- phone in Chile increased from 16 to 74 percent gy and other cost-savings practices, minimizing the between 1988 and 2000, and all but 1 percent of need for subsidies. The choice of funding strategy the remaining households were provided with can also support a level playing field among opera- public access to telephony. tors so that no operator is overcompensated or unfairly burdened by the funding mechanism. Source: Wellenius (2001); World Bank (2001d). tries. For instance, domestic banks saddled with bad as well, in a nondiscriminatory way. Thus, such debts because of past directed-lending programs are obligations were part of the license conditions for ill equipped to deal with foreign competition. new entrants into fixed-network telephony and The current handicap of universal service obliga- transport in several countries. Recourse to fiscal tions can in principle be imposed on new entrants, instruments, however, has proved more successful 302 Domestic Regulations and Liberalization of Trade in Services than direct regulation. For instance, in Chile gov- evidence that pessimism may not always be justi- ernment subsidies equivalent to less than 0.5 per- fied. For example, a number of developing coun- cent of total telecommunications revenue, allocated tries have managed to maintain or even increase through competitive bidding in 1995, mobilized 20 employment in their liberalized telecommunica- times that amount of private investment for extend- tions sectors. Since many developing countries have ing basic telephone services to rural areas (Welle- low teledensities (in the vicinity of 5 lines per 100 nius 1997). population), roughly 70 percent of telecommunica- A third method is to fund the consumer rather tions investment in developing countries is directed than the provider (Cowhey and Klimenko 2001). toward building wire line and mobile networks, Governments have experimented with various which is labor-intensive and hence helps maintain forms of vouchers in such areas as education and or raise employment levels. Petrazzini and Lovelock energy services. This last instrument has at least (1996) found in a study of 26 Latin American and three advantages: it can be targeted more directly at Asian economies that telecommunications markets those who need the service and cannot afford it; it with competition were the only ones that consis- avoids the distortions that arise from pricing ser- tently increased their employment levels, while two- vices artificially low to ensure access; and it does not thirds of the countries with monopolies saw discriminate among providers. considerable declines in the telecommunications work force.4 Nevertheless, there can be little doubt Employment Effects of Trade Liberalization. Dif- that many reform programs will require comple- ferent modes of supply have different effects on fac- mentary policies to alleviate the social and econom- tor markets. Cross-border trade and consumption ic costs of adjustment in factor markets. abroad resemble goods trade in their implications. The effects of the movement of factors of produc- Notes tion depend critically on whether the factors are substitutes for or complements to domestic factor 1 For instance, recent World Bank research has shown that while services. Given the structure of factor prices in poor maritime trade liberalization would lead to an average reduc- tion in transport prices by 9 percent and to cost savings of up countries, liberalization can typically be expected to to US$850 million, the breakup of private carrier agreements lead to an inflow of capital and skilled workers. would cause prices to decline by another 25 percent and Such inflows would tend to be to the advantage of would yield additional cost savings of up to US$2 billion on the unskilled poor, increasing employment oppor- goods carried to the United States alone. tunities and wages.3 Interestingly, it has been shown 2 In many industrial country markets where fully competitive that even when foreigners compete with local conditions have not been established, such as the telecommu- skilled workers in a services sector, the productivity nications sector in the United Kingdom, the final price itself boost to the sector from allowing access to foreign- has been regulated. ers can lead to an increase in the demand for 3 Because the poor are likely to be unskilled, the question arises domestic skilled workers: the scale effect can out- as to which services sectors they are likely to be employed in. weigh the substitution effect (Markusen, Ruther- Unfortunately, data on the skills composition of the work force ford, and Tarr 2000). in services sectors are available only for some OECD countries, Given these predictions, why are workers in devel- and even there, at a rather aggregate level. Still. a certain pat- tern can be inferred: construction, distribution, and personal oping countries sometimes skeptical about the ben- services tend to be unskilled labor­intensive, whereas commu- efits of liberalization? One concern is the possible nications, financial, and business services tend to be skilled reduction in employment in formerly public labor­intensive. monopolies, which have frequently employed sur- 4 In India the incumbent operator, the Department of Telecom- plus labor. Alexander and Estache (1999) found, for munications, expanded its work force over the 1996­2000 example, that the privatization of electricity distri- period. In the face of competition, it was forced to improve its bution in Argentina led to a 40 percent reduction in marketing strategy and expand its network, and it opened up the work force after privatization. But there is also thousands of public call offices all over the country. 303 30 R U PA C H A N D A Movement of countries have a comparative advantage in exporting labor- intensive services and manpow- Natural Persons er, the failure to liberalize mode 4 limits the overall value of and the GATS GATS for these countries. In order to effectively promote ser- vices trade and address the interests of developing coun- Major Trade Policy Impediments tries, the ongoing round of GATS negotiations must aim at liberalizing that mode. This chapter is prompted by G the need to generate informa- iven the growing importance of ser- tion and ideas for discussion of services trade via vices in the world economy, this sec- the movement of natural persons. The next section tor was brought under the purview of the summarizes the main constraints on trade by this multilateral trading system during the Uruguay mode. The extent of liberalization that occurred in Round of negotiations. The resulting General mode 4 under the first round of GATS negotiations Agreement on Trade in Services (GATS) seeks to is then assessed. On the basis of this analysis, the progressively liberalize trade in services. final section suggests ways to promote further liber- The GATS addresses trade in services according to alization in mode 4 by improving the structure of four modes of supply: (1) cross-border supply; (2) commitments and by introducing and strengthen- consumption overseas (movement of consumers to ing GATS disciplines in this regard. the home country of the service supplier); (3) com- mercial presence (movement of capital); and (4) Constraints on the Movement of Natural movement of natural persons (temporary cross- border movement of labor).1 Under the first round Persons of GATS negotiations, countries scheduled market The barriers to trade in services through temporary access and national treatment commitments for cross-border movement of labor can be broadly each of these four modes of supply, in individual grouped into four categories: services sectors that are of interest to them, and hor- izontally across all services sectors. However, these · Immigration-related regulations governing entry commitments have been highly uneven across dif- and stay of services providers ferent services sectors and also across the different · Regulations concerning recognition of qualifica- modes of supply. In particular, liberalization has tions, work experience, and training been strikingly limited in the case of movement of · Differential treatment of domestic and foreign natural persons (mode 4). Since many developing services personnel 304 Movement of Natural Persons and the GATS: Major Trade Policy Impediments · Regulations on other modes of supply, particular- providers. In addition to more stringent qualifica- ly on commercial presence. tion requirements for foreign services providers, and citizenship and residency conditions, there is It is important to note that some of these regula- differential treatment in the form of taxes, benefits, tions, particularly those concerning recognition, subsidies, and government procurement policies. stem from public policy concerns such as consumer Services providers may be required to make social protection, public interest, and national security. security contributions in the host country even The following discussion elaborates on the four cat- though they are on deputation abroad for a period egories of regulations. less than that required to be eligible for social secu- rity benefits in the future. Economic, local market, and management needs tests are used to ascertain Restrictions on the Entry and Stay of the need for foreign services providers and to Natural Persons impose quantitative limits on their entry. These The main restriction on movement of natural per- tests do not usually have clearly defined objectives, sons originates in the immigration and labor mar- nor are they transparent in their criteria or adminis- ket policies of individual countries. Temporary tration. Government procurement policies may give movement of labor is not separated from perma- preference in procurement and price to domestic nent movement of labor and comes under the suppliers of services in areas such as education, data purview of immigration legislation and labor mar- processing, and nonmedical professional services. ket conditions. Restrictions in this category include Requirements for government approval, authoriza- cumbersome application and processing proce- tion requirements on acquisition and remittance of dures for visas and permits, strict eligibility condi- foreign exchange, and restrictions on the nature of tions for such applications, quantitative limits on the legal entities and on the establishment of local entry, wage parity requirements, biases against mid- offices may also discriminate against foreign ser- dle- and lower-level services providers, and limita- vices providers. tions on the length of stay and transferability of employment in the overseas market. Restrictions on Commercial Presence Since movement of natural persons often comple- Regulations concerning Recognition, Certification, ments trade through commercial presence in ser- and Licensing vices, restrictions on foreign direct investment Requirements concerning qualifications, work expe- (FDI) in services may translate into barriers to rience, and licensing or certification are common for mode 4. Restrictions on foreign equity participation accredited services such as legal, accountancy, and in services, exclusion of certain services activities health services. Recognition requirements may alto- from foreign commercial presence, and conditions gether prevent market access for the foreign services relating to staffing and management by local per- provider or may limit the provider to specific activi- sons, the nature of incorporation, and geographic ties following entry. Recognition-related regulations and branching restrictions limit the scope for move- are usually motivated by the need to ensure high ment of natural persons associated with foreign quality of service and adherence to specified codes of commercial presence.2 professional conduct. Criteria such as citizenship, however, are discriminatory. In some sectors that GATS and the Movement of have no formal certification or licensing procedures, Natural Persons there is an element of discretion in determining equivalence between work experience and educa- The GATS establishes multilateral rules and disci- tional qualifications or training. plines for policies affecting services trade, including the integral principles of most-favored-nation (MFN) treatment, nondiscrimination, and trans- Differential Treatment of Foreign Services Providers parency.3 In addition, the GATS introduces market Trade in services via mode 4 is also constrained by access obligations that are applicable to sectors policies that discriminate against foreign services scheduled by member countries. 305 D E V E L O P I N G C O U N T R I E S A N D N E G O T I AT I O N S O N T R A D E I N S E R V I C E S Liberalization of services trade under the GATS not to use the criteria as a "disguised restriction on framework is undertaken through horizontal trade in services." The article stresses the need for (cross-sectoral) and sector-specific commitments recognition based on multilaterally agreed criteria with regard to market access and national treatment and common international standards and calls for obligations for each of the four modes of supply. sectoral negotiations on recognition (as have Countries have made a binding commitment to occurred in accountancy services). place no restrictions on market access (Article XVI) A separate annex on the movement of natural or on national treatment (Article XVII) for a mode persons defines mode 4 as covering "persons who of supply or have made a partial commitment by are temporarily working in another member coun- limiting market access and national treatment in try. It is not applicable to individuals who are seek- line with various conditions listed in their schedule, ing access to the employment market of another or have left the entry "unbound" (that is, made no member on a permanent basis or for citizenship or commitment).4 A review of the GATS framework residency purposes." The annex covers independent suggests that the provisions pertaining to move- and self-employed suppliers who are paid directly ment of natural persons are sufficiently strong. A by their customers and natural persons employed review of the first-round commitments in mode 4, by services suppliers in the host or home country or however, indicates that liberalization in this mode in a third-member country to supply a service. The has been very limited. annex states that countries can regulate entry and stay of natural persons provided they do not apply these measures in such a manner as to nullify or Assessing the GATS Framework impair the benefits granted to members under their Several provisions and general disciplines within specific commitments. the GATS framework are pertinent to the move- Overall, the GATS framework addresses some of ment of natural persons. Article III, on transparen- the most important regulatory barriers that con- cy, requires members to publish "all relevant strain cross-border movement of labor in services. measures of general application which pertain to or There are, however, two weak points in the GATS affect the operation of this agreement," including framework. The first is that important areas such as relevant international agreements to which the government procurement and subsidies that affect country is a signatory. Members are also required to movement of natural persons either are not covered inform the Council for Trade in Services of any new or are inadequately covered.5 The second concerns laws, regulations, and administrative guidelines or the exception of measures related to public policy amendments that are relevant to their specific com- objectives. Regulatory interventions embodying mitments, to respond to requests for specific infor- market access and national treatment restrictions mation on any measures affecting the members' fall under Articles XVI and XVII, respectively. In all commitments, and to establish enquiry points for other cases such measures fall under Article VI, on providing this information to other members. Arti- domestic regulation. It is difficult to distinguish cle VI, on domestic regulation, requires members to domestic regulations related to market access and ensure that in sectors where specific commitments national treatment provisions from regulations are undertaken, requirements are based on objective related to public policy objectives. Thus, market and transparent criteria and do not in themselves access and national treatment limitations can ren- constitute a restriction on the supply of the service. der Article VI disciplines ineffective (see Low and It also obligates members to undertake objective Mattoo 2000). and impartial reviews of administrative decisions affecting services trade. Article VII, on recognition, Assessing GATS Commitments on the Movement of grants members the discretion to recognize, in Natural Persons whole or in part, the education, experience, and licensing or certification of foreign services The GATS has failed to deliver sufficiently liberal providers autonomously or by mutual agreement or commitments in mode 4. Horizontal and sectoral harmonization. It also obligates members to apply commitments filed by countries have been more the criteria or standards for authorization, licens- limited for mode 4 than for the other modes of sup- ing, or certification equally across all countries and ply. The following discussion highlights the main 306 Movement of Natural Persons and the GATS: Major Trade Policy Impediments problems with the nature and structure of commit- · Requirements for technology and skill transfer ments in this mode. (training local staff) · Discriminatory tax treatment Limited Sectoral Coverage of Mode 4 Commit- · Requirement of government approval ments. Mode 4 commitments are very limited in · Requirement of work permits, residency, and citi- sectoral coverage. Sectors such as health services zenship in certain sectors and legal and accountancy services where cross- · Recognition of professional qualifications by the border labor mobility is important have not been importing country scheduled by many countries. Moreover, even where · Restrictions via minimum investment require- commitments have been made, they are subject to ments. many market access and national treatment restric- tions. Bias toward Higher-Level Services Personnel. The commitments in mode 4 are further limited because Limitations on Commitments.6 Most countries they are bound for only a small subset of services have left their market access and national treatment personnel. There is a clear bias in the horizontal obligations in mode 4 unbound in their sectoral commitments toward liberalizing the movement of schedules and have referred to their horizontal higher-level services personnel. (Box 30.1 provides schedules. This amounts to virtually no sector-spe- recent data on immigration of highly skilled and cific liberalization in mode 4. The horizontal com- skilled workers to OECD countries.) Entry require- mitments in mode 4 are in turn subject to many ments are bound for three main categories of ser- conditions and requirements relating to functional vices providers: business visitors; personnel engaged or hierarchical criteria, length of stay, labor market in setting up commercial presence, such as intracor- and economic needs tests, and the like. For instance, porate transferees (ICTs); and personnel in "special- 100 countries have placed limitations on their hori- ity occupations." The commitments on ICTs come zontal commitments in mode 4, as against only 4 closest to full bindings. More than one-third of countries for mode 2. Moreover, the conditions are mode 4 entries refer to intracorporate transferees; often not clearly specified.7 of a total of 328 total entries, 239 relate to execu- Table 30.1, which summarizes the commitments tives, managers, and specialists and 135 deal with in selected services sectors where movement of nat- ICTs. Only 17 percent of all horizontal entries cover ural persons is important, indicates that there are low-skilled personnel. Sectoral commitments simi- fewer full (no restrictions) commitments under larly facilitate the entry of only higher-level person- mode 4 than for any of the other modes of supply. nel in the professional, managerial, and technical The percentages of partial commitments (commit- categories, as specified in the horizontal schedules. ments with limitations) and unbound commit- There are, altogether, six entries for "independent ments (no commitment) are far greater under contract suppliers" and for "other" services person- mode 4 than for the other modes of supply. nel, and there are very few commitments for quali- The main conditions that are attached to com- fied specialists. Moreover, these latter categories are mitments in mode 4 are: not permitted to move in an individual capacity; they must be working for a specified duration for a · Entry restrictions for certain sectors and cate- juridical person in another country. Hence, liberal- gories of personnel ization in mode 4 is mainly for services providers at · Limits on the duration of stay of natural persons a higher level and is linked to commercial presence. · Quantitative restrictions by numerical quotas for entry; specifications concerning the proportion of Structural Problems with Mode 4 Commitments. total employment that can be met by foreigners; A fundamental problem with the structure of the specifications concerning the proportion of total mode 4 commitments is that there is no separation wages of temporary and permanent labor under the exist- · Preemployment conditions and related require- ing framework of commitments, even though the ments GATS is meant to cover only temporary labor flows · Economic, labor market, and management needs in services. Most of the limitations which have been tests filed under mode 4 relate to general immigration 307 7 4) No 31 21 8 5 6 3 7 0 7 13 15 16 18 12 12 31 6 11 51 7 71 natural of (mode 91 68 88 92 85 94 97 87 81 39 91 80 71 80 79 87 85 78 77 39 76 100 Partial 2 2 0 0 0 0 0 0 4 0 0 2 4 8 9 9 9 3 8 0 persons Full 12 71 Movement 9 2 3 4 3 9 3 No 11 21 0 0 9 4 9 6 5 9 8 0 0 16 11 sectors. 3) presence 87 98 28 72 72 59 73 86 85 08 76 64 65 38 43 13 33 45 53 74 listed 67 is cial 100 Partial (mode total 4 9 Full 51 24 24 31 24 12 13 02 0 the 16 32 53 56 52 72 58 45 85 35 33 of Commer basis 9 3 5 8 0 0 9 2) No 14 12 17 13 12 11 14 14 19 16 15 11 21 0 heT 17 (mode 67 45 44 20 28 22 36 34 23 53 76 8 rounding. 58 36 35 22 21 13 24 24 47 50 of Partial Consumption because abroad Full 24 41 53 68 55 66 52 61 69 47 33 31 50 56 64 60 72 61 66 81 53 33 100 to vices) add Ser 0 No 16 30 12 22 22 19 18 37 27 33 18 30 18 18 24 19 18 34 27 33 17 not may 1) movement 1998). 67 41 44 26 28 22 36 29 19 33 67 60 36 41 30 31 19 30 18 12 27 50 Partial centages (Professional (mode Per December Full 18 29 44 52 50 59 45 34 54 33 33 22 34 41 52 45 63 52 47 62 40 33 Supply Cross-border (Geneva, of commitment. no vices" = vices vices Ser Mode no ser ser vices vices and ser and ser and Accountancy commitment; on Sector chitectural chitectural ar nurses, ar nurses, by partial Note = bookkeeping vices bookkeeping vices ser and vices ser and vices partial landscape midwives, ser landscape midwives, ser "Background activity) by by vices vices and vices vices and Commitments each auditing, ser vices ser vices in vices ser dental auditing, engineering vices ser dental commitment; ser engineering ser Secretariat, vices ser and y provided treatment vices ser and y provided full 30.1 access = ser planning ser planning WTO Full centages vices vices ce: ableT (per Market Legal Accounting, axationT chitectural Ar Engineering Integrated Urban Medical eterinarV physiotherapists Ser Other National Legal Accounting, axationT chitectural Ar Engineering Integrated Urban Medical eterinarV physiotherapists Ser Other Note: Sour 308 Movement of Natural Persons and the GATS: Major Trade Policy Impediments B O X 3 0 . 1 I M M I G R AT I O N PAT T E R N S I N O E C D C O U N T R I E S OECD countries displayed an interesting shift number of foreign software engineers to return in immigration patterns in the 1990s: perma- to their countries of origin. nent immigration has been declining, but tem- A German policy to hire 20,000 information porary immigration, notably of highly skilled technology specialists for a period of five years and skilled workers, has been increasing (see started on August 1, 2000. Similarly, for the year the figure). 1999­2000 Australia issued an extra quota of 5,000 Since it is reasonable to assume that there is places, in addition to 35,000 places designated for an excess supply of both categories of skilled workers. In the United States it was proposed migrants, these trends reflect a shift in policy in that the number of H1-B visas be increased to the host country. The motivation is straightfor- 200,000, a substantial increase from 115,000 in ward: the emergence of high-technology 2000. The Japanese government altered its policy industries that need more qualified workers for qualified skilled workers (engineers, researchers, than are available domestically. Allowing tem- and the like) by extending the initial duration of porary entry of foreign workers is a flexible stay from six months to one year. The figures quot- response to short-term shortages. For instance, ed here do not include renewal of existing permits. it is reported that the financial troubles in the The stock of temporary workers, as opposed to information technology sector have forced a flows of workers, is therefore even larger. Entry of Foreign Workers in Selected OECD Countries Permanent Workers Temporary Workers (thousands) (thousands) 1992 1996 1997 1998 1992 1996 1997 1998 450 450 400 400 350 350 300 300 250 250 200 200 150 150 100 100 50 50 0 0 AUS CAN FRA GER SWI USA AUS CAN FRA GER SWI USA Note: AUS, Australia; CAN, Canada; FRA, France; GER, Germany; SWI, Switzerland; USA, United States. Source: OECD, TD/TC/WP(2001)26/Rev2. These trends are also visible outside the OECD Area (CEFTA) has also led to increased movement countries. Singapore has changed its regulations of temporary workers among member countries. that govern the issuance of work permits in order In Hungary, for instance, the number of newly to accommodate greater numbers of foreign issued work permits increased from 14,000 in entrepreneurs, and Malaysia has made changes 1996 to 22,500 in 1998. to encourage foreign skilled workers. The imple- mentation of the Central European Free Trade Source: Prepared by the volume editors, based on OECD data. legislation and labor market regulations that per- overall agreement and are thus subject to arbitrary tain to permanent migration. The existing structure interpretation by immigration officials and con- of mode 4 commitments also suffers from a lack of sular offices. Additional requirements such as eco- clarity and uniformity. There is no uniformity in nomic needs and labor market tests have not been the definition and coverage of the various cate- clearly specified and defined in terms of their crite- gories of services persons; the personnel categories ria or administration. Lack of specificity in defini- are not well defined either in the schedules or in the tions and in some of the conditions lends itself to 309 D E V E L O P I N G C O U N T R I E S A N D N E G O T I AT I O N S O N T R A D E I N S E R V I C E S administrative discretion, discriminatory practices, commitments. A finer classification of services per- and reduced predictability. sonnel categories would facilitate clear and detailed sectoral commitments that are relevant to the par- ticular sector or subsector under consideration and Proposals for Liberalizing Movement of would reduce the scope for discretion and discrimi- Natural Persons nation in implementing the commitments. If movement of natural persons is to be liberalized under the GATS, existing commitments in this Horizontal Formula for Classification of mode have to be significantly improved in the cur- Services Providers rent round of service sector negotiations. This can be realized through a two-pronged approach to Horizontal disciplines have an important role in lib- negotiations: on a country-to-country level and on eralizing mode 4. Specific, detailed, and binding a multilateral level. Country-to-country negotia- sectoral commitments can be meaningful only if the tions concerning this mode will vary depending on coverage, definitions, and criteria for provider cate- the individual interests of the members concerned. gories, applicable measures, and associated criteria Of broader significance is the need for multilateral are common across all countries. All members discussion on ways to improve the structure and should agree on the coverage of professionals and nature of the GATS framework and commitments activities within the personnel categories that are concerning mode 4. The following discussion high- included in the horizontal schedules (ICTs, business lights the main issues for multilateral negotiations. visitors, specialists, and other personnel) and must agree on the minimum criteria for determining whether an individual services provider fits into a Improving the Structure of Commitments in Mode 4 particular category. There must be a broad consen- Multilateral discussions should aim at supplement- sus on the categories and subcategories and the con- ing the horizontal commitments in mode 4 with ditions under which additional limitations such as sector-specific commitments in this mode, especial- economic needs tests and residency requirements ly in sectors such as professional and business ser- may be allowed and on when such conditions vices where mode 4 is important. These sectoral should be barred altogether. commitments must be detailed and specific in The multilateral discussions must also focus on terms of the measures that are applicable to individ- expanding the categories of services providers cov- ual sectors and the services personnel categories rel- ered by the horizontal commitments to remove the evant to each sector. current bias toward higher-level personnel. It is Countries need to make unambiguously worded important to include middle- and lower-level and well-defined sectoral commitments with clearly providers and to make the commitments more rel- outlined criteria for application of any limitations, evant to the interests of the developing countries for all subsectors within the sector that has been with expertise in these categories. The coverage can scheduled. All limitations, conditions, and excep- be expanded in two ways: by explicitly introducing tions should be clearly laid out in the sectoral new categories such as technical support personnel schedules, for both market access and national (which would, for instance, include systems ana- treatment, rather than being broadly outlined in the lysts and programmers in the case of software ser- horizontal schedules. Countries must also take steps vices), or by defining the coverage of the "other to furnish information on these measures, in line persons" and "specialists" categories. This expan- with GATS Article III. sion should allow for the inclusion of middle- and Specificity and detail will also require improved lower-level personnel in these latter categories by targeting of categories of service providers to whom specifying relevant criteria and by modifying or the commitments and limitations are applicable. removing certain conditions relating to skills, pre- This can be done by introducing more disaggregat- employment, and job responsibilities that at pres- ed categories of services providers in the sectoral ent favor higher-level persons. The second option is schedules that fit within the broad categories of likely to be easier to implement. In this context, intracorporate transferees, business visitors, special- common coverage and definition of these broader ists, and other persons referred to in the horizontal categories would help. 310 Movement of Natural Persons and the GATS: Major Trade Policy Impediments The need is thus for complementarity rather than unrestricted in nature, any conditions attached to substitutability between the sectoral and horizontal its issuance should not be more onerous and restric- schedules. Neither set of commitments should tive than those already specified in the commit- dilute the other. ments. This would require countries to make more generous and binding sectoral commitments in mode 4 that do not backtrack on the status quo. Broadening the GATS Framework on Movement Once more liberal offers are forthcoming in mode of Natural Persons 4, the GATS visa would facilitate uniformity in mar- It is also necessary to establish multilateral guide- ket access procedures, as opposed to the current sit- lines on some issues, to strengthen some of the cur- uation of very divergent immigration standards and rent GATS provisions, and, overall, to broaden the procedures across countries. There should also be reach of the GATS framework with respect to mode multilateral guidelines governing the granting and 4. The relevant issues to be addressed in this context use of the GATS visa, including guidelines on the include (a) the separation of temporary from per- time frame for issuance and procedures, costs, manent labor flows, (b) wage parity, (c) recogni- renewal, transferability of jobs, and treatment in tion, and (d) economic needs and other tests. It is terms of taxes, subsidies, and government procure- important to note that for these norms to be effec- ment. The GATS visa should also include mecha- tive, they must be supported by specific and detailed nisms for determining the status of applications, sectoral commitments and uniform horizontal notifying delays, and questioning the grounds for commitments. rejection. Safeguard mechanisms can be introduced to prevent misuse of such visas, including more Separating Temporary from Permanent Labor stringent requirements for first-time applicants and Flows, and the GATS Visa. Multilateral discussion obstacles to transferring the visa to others and to is required on the need to separate temporary from permanent residence and citizenship status. permanent movement of labor when a services sec- tor has been scheduled. Temporary services Introducing Norms for Addressing Wage Parity. providers should be treated separately from perma- The implications of wage parity conditions vary nent migrants, outside the domain of immigration- depending on sector-specific labor supply and related laws and labor market regulations and under demand conditions, as well as on the modalities by a separate set of regulations. Alternatively, tempo- which wage parity conditions are implemented. It is rary services providers could be covered by a special thus difficult to establish a common approach to the subset of regulations within the overall immigra- use of wage parity conditions. Nevertheless, guide- tion policy framework, with more liberal conditions lines are required to establish (a) under what condi- for entry and stay. This would reduce the adminis- tions (sectors, personnel categories, and local and trative burdens, delays, and costs they face in enter- economic conditions) wage parity may be permissi- ing the foreign market and would make it easier to ble; (b) how wage parity conditions are to be imple- address issues such as social security, wage parity, mented; and (c) how to link wage parity conditions and recognition for such personnel. to entry conditions and formalities. A separate class of visas, a "GATS visa," could be It is proposed that in sectors in which the host established for service professionals temporarily country faces a severe shortage of personnel, wage working overseas. This visa would be applicable to parity conditions should not be imposed, as there is service providers who are covered by the sectoral no immediate displacement effect on local labor. To and horizontal commitments filed by a member ensure that exploitation of foreign labor does not country and would help streamline the implemen- occur, conditions can be imposed on other terms of tation of these commitments. This complementari- employment, such as work hours, leave, overtime, ty between the GATS visa and the commitment and benefits. Guidelines on the remuneration pack- schedules is possible only if the recommendations age given to foreign workers can also be considered outlined above regarding specificity, finer classifica- (depending on the period of time they are deputed tion of provider categories, wider coverage of cate- abroad, sectoral conditions, and the employment gories, and transparency are reflected in the level) in order to ensure a fair standard of living in commitments. Although the GATS visa need not be the host country. This wage could be the reservation 311 D E V E L O P I N G C O U N T R I E S A N D N E G O T I AT I O N S O N T R A D E I N S E R V I C E S wage of the foreign services provider plus to be used to accord recognition to professionals. allowances for cost of living, taxes, savings, and such These include criteria relating to minimum pro- concerns, without requiring payment of a specific fessional education, as sanctioned by a diploma, minimum wage or percentage of wages as is done at formal licensing or certification requirements, present. This would permit wages of foreign ser- and minimum professional experience, which in vices providers to be below host-country wages but turn should be reflected in the sectoral and hori- within a fair margin, allowing the home country to zontal commitment schedules. In addition, where maintain its comparative advantage in mode 4 in countries have included recognition requirements that sector while also preventing the influx of for- in their commitment schedules, there should be eign workers from massively driving down domestic mechanisms under the GATS to enable countries wages in the sector in the host country. Specifics of to engage in bilateral discussions to compare the wage margin will depend on sector characteris- qualifications across home and host states and to tics and should be decided mutually by the con- assess the extent of equivalence based on bilater- cerned countries under bilateral wage agreements ally determined criteria. and discussions between professional or industry · Norms concerning equivalence of work-related and associations in these countries. In sectors where academic qualifications. Consensus is required on there is likely to be local displacement, particularly how and when to accord equivalence between on- in unskilled and semiskilled occupations, countries the-job experience and academic qualifications. could consider grandfathering indigenous workers Norms must be established that specify the kinds through wage subsidies. In both cases, the interests of jobs or positions and the kinds of academic of labor unions and workers in the receiving coun- qualifications that may be considered equivalent tries could be addressed by using taxes and fees col- and substitutable for meeting entry requirements lected from foreign services providers to fund job and sectors where such equivalence is difficult to training and to subsidize or relocate currently establish. The latter will require the participation affected and potential future domestic services of professional bodies and associations in mem- providers. It is likely that even the posttax or postfee ber countries to provide the criteria for equiva- receipts of the foreign services providers would lence, along with names of well-recognized exceed their reservation wages. training and higher education institutes in the Multilateral discussions should also aim at respective countries, to better assess professional delinking the wage parity condition from the visa qualifications. GATS disciplines on recognition issuance process. Wage parity should not be a cum- should further discourage differential treatment bersome precondition for the issuance of visas. The of the value of work experience and qualifications earlier recommendation for specifying a maximum between foreign and domestic services sector pro- time frame for issuing GATS visas and notification fessionals. requirements for delays and additional conditions · Norms concerning temporary licensing. Disciplines should take into account delays caused by wage- governing licensing should allow for temporary related and labor certification requirements. Such licensing of foreign service professionals when delays and rejections should be open to challenge in such licensing procedures are lacking in the home the dispute settlement forum of the GATS. country. The absence of such procedures should not constitute a barrier to the movement of pro- Strengthening GATS Norms on Recognition. The fessionals from these countries. Procedures could GATS already contains a strong provision for recog- be developed for temporarily licensing engineers nition under Article VII. This provision needs to be to practice in the specialty area. Multilateral dis- strengthened by establishing detailed norms in cer- cussion is required to determine these sectors and tain areas and by facilitating mutual recognition the associated procedures. agreements among countries. · Norms concerning broad-based equivalence in recognition. Multilateral discussions should focus · Norms for nonaccredited sectors and activities. In on a system of granting recognition through professional services such as software services broad-based equivalence of qualifications and where there are no formal accreditation or licens- standards. It would be useful to establish bridging ing procedures, norms are needed on the criteria mechanisms where there is a divergence of 312 Movement of Natural Persons and the GATS: Major Trade Policy Impediments requirements and standards between home and software services, professionals are exempt from host countries. The GATS framework should such tests only if their academic degree in a special- encourage discussions on a compensatory system ty is directly related to the employment, a practice of granting recognition, whereby local adaptation that is discretionary. In lower-skilled and semi- periods and aptitude tests for foreign services skilled services activities, economic needs and other professionals can be used to offset differences tests should not be applicable unless there is likely among national systems and standards, such as to be a disruptive effect on the host country labor those in the European Union.8 Within this frame- market in that sector. work of broad-based equivalence, countries can In addition to the areas mentioned above, disci- bilaterally negotiate recognition agreements to plines are needed on subsidies and government pro- suit their particular needs. There should also be curement policies. Countries could be required to mechanisms to facilitate regulatory cooperation make explicit the existence of government procure- among professional bodies in member countries ment policies and subsidies in all relevant sectoral to enable systematic exchange of information, commitments schedules and to provide informa- mutual monitoring, and cooperative enforce- tion on their nature, their magnitude, how they ment. operate, and other relevant parameters. Where such policies are present and countries have scheduled Introducing Norms for Economic Needs and commitments, any limitations on foreign services Other Tests. Many countries have included condi- providers resulting from these policies should be tions relating to economic needs tests in their com- clearly specified. These limitations could take the mitment schedules, but the wording is usually form of ceilings on the share of contracts or value of vague. For instance, the EU requires the foreign ser- transactions to be procured from domestic sources, vices provider to have an "effective and continuous the number of local services persons employed, and link" with the market, but it is not clear what is the extent of preference to be accorded domestic meant by this term. contracts. In the case of subsidies, limitations could The most far-reaching step in this regard would be filed on the maximum extent of subsidy as a be to abolish the use of such measures as horizontal share of total value of transactions in the sector. It is limitations. Any market access limitations involving important that there be transparency in the use of this condition should be limited to sectoral com- these practices. mitments, but in line with certain multilateral guidelines so as to reduce the scope for discretion in Conclusion the use of such tests. These guidelines include laying down clear criteria for applying the tests, establish- This chapter has highlighted the limited nature of ing norms for the administrative and procedural liberalization that has occurred under the GATS formalities associated with the tests, and specifying framework for the movement of natural persons. how the results of the tests are to be used in restrict- Significantly more liberal commitments are ing entry to foreign services providers (for example, required in this mode if developing countries are to translating the findings into quantitative limits on exploit their comparative advantage in labor-inten- foreign personnel). Requirements and decisions sive services. Several proposals have been made for made on the basis of such tests should be subject to improving the existing commitments and commit- challenge under the WTO's dispute settlement ment structure and for improving the GATS frame- mechanism. work by introducing multilateral norms on Consensus is required on the occupational cate- important regulatory aspects concerning this mode gories that can be subject to such tests. These cate- of services trade. Both approaches must be taken in gories should be limited to as few as possible, and conjunction: the multilateral norms would have lit- the market conditions (shortage of personnel; pub- tle significance unless accompanied by the pro- lic policy objectives, and so on) under which these posed modifications in the commitments, and the categories would be subject to economic needs and improved commitments will have greater relevance other tests should be defined. Highly qualified ser- if there are broad principles to facilitate their imple- vices professionals should be exempt from econom- mentation. The proposed changes in the GATS ic needs tests. At present, in some sectors such as would also need to be supported by domestic mea- 313 D E V E L O P I N G C O U N T R I E S A N D N E G O T I AT I O N S O N T R A D E I N S E R V I C E S sures and reforms at the individual country level to 3 MFN is applicable to all services that have been scheduled by a facilitate and to derive benefits from the liberaliza- member unless an MFN exemption has been taken. The princi- ple of nondiscrimination is reflected in GATS Article XVII.3. tion of movement of natural persons. These domes- tic reforms include areas relating to standards and 4 Market access limitations include limits on the number of ser- recognition, immigration and labor market policies, vices suppliers, the value of services transactions, the number of services operations or the quantity of the services output, and tax policies. Most important, countries need to the number of natural persons employed, foreign equity par- be more transparent about their domestic regula- ticipation, and the type of entity or venture. tions. Without increased transparency on the part of individual countries, most of the proposed rec- 5 Article XIII exempts government procurement from the appli- cation of MFN, market access, and national treatment obliga- ommendations will not be enforceable or effective. tions. Article XV requires members to enter into negotiations and exchange information on subsidies to develop multilateral disciplines in this regard. Notes 6 The discussion of this and the next point (on bias toward high- 1 Movement of natural persons is defined in Article I.2 of the er-level personnel) is based on a review of existing horizontal GATS as "[s]upply of a service . . . by a service supplier of a schedules of commitments. member through presence of natural persons of a member in the territory of any other member." It includes both services 7 Given the transparency provision in Article III, individual mem- providers who are working overseas in an individual capacity ber countries can seek this information from other members, and those who are part of a home, host, or third-country com- establish mechanisms and inquiry points to provide such infor- mercial establishment. mation to other members, and seek recourse to the dispute settlement mechanism in case this provision is violated. 2 Certain restrictions on commercial presence, including the type of commercial entity (corporation, partnership) or a 8 The General System of Directives of the EU permits profession- requirement for prior approval by host-country professional als from other member countries who fall short in their qualifi- associations, are often meant to address issues such as profes- cations and standards to qualify following a local training and sional liability and misconduct and national interest. adaptation period. 314 31 C AT H E R I N E L . M A N N Electronic cable, satellite, or other Internet "backbones" and "devices"). It also depends on institutions and Commerce, firms such as Internet service providers (ISPs) to connect mar- the WTO, and ket participants, as well as end- user devices such as personal Developing computers (PCs), television sets, and mobile telephones. Countries · Services that connect the Inter- net marketplace to the tradi- tional marketplace and further support the development of E both. Payment over the Inter- lectronic commerce is a shorthand net through credit, debit, or Smart Cards or term that embraces a complex amal- through other on-line methods link buyers and gam of technologies, infrastructures, processes, and sellers in the virtual marketplace, who then com- products. It brings together whole industries and plete their transactions (whether virtual or physi- narrow applications, producers and users, informa- cal) through distribution and delivery systems. A tion exchange, and economic activity in a global range of new services (customer data management, marketplace called the Internet. One of the best ways network and device security systems, pricing and of understanding electronic commerce is to consider auction methods, and so on) augment and deepen its basic elements and its impact on traditional the relationships between buyers and sellers. transactions and marketplaces. This approach shows · Protocols, laws, and regulations that govern conduct clearly how electronic commerce is intricately woven and relationships. Technical communications and into the fabric of domestic economic activity and interconnectivity standards, the legality and international trade. modality of electronic signatures, certification, Electronic commerce has three basic elements, and encryption, and disclosure, privacy, and con- not all of which are equally or universally developed tent regulations all affect the conduct of activity within and among countries around the world: in the Internet marketplace and the relationships between businesses, consumers, and government. · Institutions and technologies that create the global These instruments may be developed and network supporting the Internet marketplace. Elec- enhanced by both private and public intervention tronic commerce relies on a variety of technolo- and interaction. gies, some of which are well established (public phone network technology) and some of which With these basic elements, electronic commerce are developing at breakneck speeds (for example, leads to three kinds of innovations in traditional technologies to interconnect telecommunications, transactions and marketplaces: 315 D E V E L O P I N G C O U N T R I E S A N D N E G O T I AT I O N S O N T R A D E I N S E R V I C E S · Process innovations. Electronic commerce simpli- characterized by network externalities, developing fies, makes more efficient, reduces costs, or other- countries should take advantage of the technical wise alters the process by which an existing leadership coming out of the private sector in the transaction takes place--for example, by stream- most advanced countries (and their own private lining accounting or improving inventory con- sector, even if nascent) and follow that example. trol. "Standing on the shoulders of giants" makes sense · Product innovations. Electronic commerce creates when network externalities and interoperable stan- or facilitates new industries and products not pre- dards are key to maximizing the benefits of e-com- viously available; some examples are business-to- merce. Trying to develop unique domestic business (B2B) exchanges for products or standards or following the old technique of import MP3.com for music creation and distribution. substitution to develop a domestic networking · Market innovations. Electronic commerce creates industry is even more economically wasteful in the new markets in time, space, and information that context of the Internet and electronic commerce did not exist previously because transactions than it was in more traditional sectors. That said, costs and coordination costs were prohibitively firms which build on global technologies and then high; markets for individually tailored editions of apply those technologies to local preferences and newspapers or sales of Andean mountain artisan- needs will reap the highest rewards. ship are examples. In important ways, e-commerce transactions blur the traditional line between goods and services. Consider software, for example. In a box, it is a pur- Electronic Commerce, Domestic Policy, and chased good. When digitized and downloaded, is it International Trade Negotiations a purchased service, or a combination of the good From this presentation of the basic elements of the and the method of delivery? And when used Internet marketplace as well as how it sparks inno- through an application service provider, is it a dif- vations in traditional transactions and market- ferent kind of service--one that is rented but not places, it is clear that the Internet and electronic purchased? In the context of WTO commitments, commerce integrate domestic and global markets embracing this idea could lead to a liberalizing bias from the very start. Therefore, negotiating on trade in favor of electronic delivery of goods, services, and issues related to electronic commerce will affect their combination as compared with delivery by domestic policies even more than trade negotia- another scheduled mode. tions have in the past, and vice versa. Some of the Trade negotiations, however, have important traditional distinctions between domestic and for- political dynamics. Just as negotiators in developing eign services and in the classification of goods and countries should recognize the important benefits services begin to blur in the Internet marketplace. of greater liberalization of traditional services sec- Electronic commerce and its related activities tors, negotiators in industrial countries should rec- over the Internet can be the engines that improve ognize the need to liberalize their traditionally domestic economic well-being through liberaliza- protected sectors in light of the impending compet- tion of domestic markets, more rapid integration itiveness of the developing countries. Both industri- into globalized production, and leap-frogging of al and developing countries will gain from the available technology. These engines are fueled by liberalization process and from efforts to exploit the competitive communications sectors, financial ser- potential of global electronic commerce. vices, and delivery logistics. Because these sectors are fundamental to the workings of a modern econ- The Rise of Electronic Commerce as an omy, liberalization here will redound to greater eco- Economic Activity nomic well-being than comparable liberalization in more narrowly focused sectors. How important are these issues right now? After all, The desire to be part of the global e-commerce estimates of the growth of Internet usage and elec- wave can be a powerful force for eroding domestic tronic commerce both within domestic markets and vested interests that have slowed international worldwide are notorious for their hyperbole. As negotiation on and domestic liberalization of these each year passes, however, actual growth has sur- sectors in the past. Since electronic commerce is passed the previous estimate rather than falling 316 Electronic Commerce, the WTO, and Developing Countries short of it. Respected sources such as Forrester tional firms and strategic business alliances com- Research expect worldwide electronic commerce municate, get price quotes, submit bids, transfer revenues to surpass US$6.8 trillion by 2003, data, offer customer service, produce product accounting for about 9 percent of worldwide sales. designs, code software, and basically do business Although the United States has an overwhelming over the Internet in the international arena. Past (close to 75 percent) share of electronic commerce policies that focused on gaining a foothold on the at present, diffusion into Europe and Asia and then global production ladder (through export-process- into Latin America and Africa is expected to be ing zones, for example) will no longer suffice. rapid, and the U.S. share may drop to below 50 per- Countries that do not have a complementary cent by 2003.1 domestic environment conducive to Internet usage In developing countries Internet use and its eco- and electronic commerce will be marginalized from nomic potential are growing exponentially.2 As the globalized production process and the global recently as a couple of years ago, Internet users were economy, at increasingly great cost to their citizens. principally in the United States, but as of August 2001 the share of U.S. users in the global communi- Speeding the Adoption of Electronic Commerce ty was just 32 percent. The share of active Internet through Domestic Reforms users in Asia and the Pacific Rim, Latin America, Africa, and the Middle East increased from 23 per- Developing countries need to address both socio- cent in 1999 to 35 percent in August 2001. In India economic and regulatory barriers before their use of the number of Internet users rocketed from about electronic commerce and the Internet can match 200,000 (about 0.02 percent of the population) in that of the United States or Europe. The socioeco- November 1998 to 5 million in December 2000-- nomic challenges are difficult to surmount and will more than double the projection for users in India take time; by contrast, the path to reducing regula- made in 1999. In China a reported 60 percent of tory barriers is clearer, and the benefits are observed businesses use the Internet, and e-commerce rev- more quickly. High Internet access and telecommu- enues could rise from US$11.7 million in 1998 to nication charges, low penetration of electronic US$1.9 billion in 2002. In Latin America Internet means of payment (such as credit, debit, and Smart usage tripled between 1998 and mid-2001, and e- Cards), and cumbersome delivery systems are pri- commerce revenues are projected to be US$8 billion mary obstacles to the growth of electronic com- by 2003. Africa is fully wired, and in South Africa merce in developing countries--and all are areas in electronic commerce is expected to generate US$1.1 which domestic policy reforms are possible and can billion in 1999. be complemented by a strategy of international Two important facts about e-commerce are often openness.3 overlooked. First, the vast bulk of the actual and, to an Communications systems are simply critical. even greater extent, the expected growth in revenues Because the Internet creates new electronic business from e-commerce comes from B2B transactions. In environments, "surfing" is a key way for users to see 1998 the ratio of B2B to business-to-consumer trans- what businesses are doing and what market niches actions was 5.5 to 1, but by 2003 it is expected to be 15 remain to be exploited. High and metered commu- to 1. This is not only because the production process nications charges create both large "entry" costs and for traditional goods and services has been fragment- huge ongoing disincentives for Internet usage and ed among many suppliers but also because of the therefore for the development of e-commerce busi- introduction of many new value-added services. ness both within a country and in international Second, in virtually all countries other than the trade. Competition, for telephone access and among United States, electronic commerce is export orient- ISPs, is a key area in which government policy can ed. In the United States the share of export sales in make a difference in access to and use of the Inter- total e-commerce revenues is only 10 percent, but in net. The evidence shows that foreign direct invest- Canada it is 83 percent, in Latin America it averages ment can play an important role in improving the 79 percent, and in Asia and the Pacific it is 38 percent. competitive climate in even the smallest countries. The nature of the production process (compris- A supportive electronic payments infrastructure is ing both manufacturing and services) is becoming crucial for promoting electronic commerce. This increasingly fragmented and globalized. Multina- factor exposes a key link between electronic com- 317 D E V E L O P I N G C O U N T R I E S A N D N E G O T I AT I O N S O N T R A D E I N S E R V I C E S merce and the financial sector of the economy. Elec- tives--domestic entrepreneurs and consumers will tronic payments require an easy-to-use and secure benefit in the long run. payment vehicle. Although a number of countries are focusing on "cash on delivery" for tangible prod- Electronic Commerce on the WTO Agenda ucts, in the future a payment method that is on-line so as to accommodate products delivered digitally In reviewing the services infrastructures that make will be required. For business-to-business transac- electronic commerce possible, as well as the impact tions, an easy-to-use electronic payments mecha- electronic commerce has on the traditional market- nism is essential for achieving the cost reductions place, it is clear that the WTO agenda has touched promised by Internet-based commerce. At the same many aspects of what makes e-commerce develop time, security for financial transactions is a sine qua successfully. It is also clear that much work is still to non. Liberalization, along with appropriate regula- be done, both by countries themselves and through tion and supervision, will help bring international the WTO, to reap the benefits of e-commerce. best practice, as well as global technologies, to bear First, electronic commerce is clearly global from to improve the functioning of the domestic finan- the very start, putting it directly within the purview cial sector. of the WTO. Although traditional borders do still Finally, delivery logistics (including customs) matter in the world of international trade, electronic rounds out the set of services infrastructures that commerce diminishes their importance. Companies are key components in developing e-commerce. with a Website are instantly international--whether Speed is one of the most important characteristics or not they intended to be. of electronic commerce. Overnight delivery, just-in- The electronic marketplace is currently free from time processing, and 24 (hour)/7 (day) operations explicit trade barriers (see Box 31.1). The absence of are examples of how much faster and more precise- international tariffs or other barriers to electronic ly timed are economic activities in the e-commerce commerce transactions encourages people to inves- world. A country with inefficient distribution and tigate and then to continue to participate in the delivery systems and without multimodal transport Internet marketplace, resulting in greater efficiency for international participation will be left behind in and economic benefits for its participants. e-commerce. Government policies have a direct Yet the infrastructures that make electronic com- impact in these areas, and governments have the merce possible are still burdened by a myriad of principal task of improving the efficiency and trans- trade and investment barriers. The growth of global parency of customs operations. electronic commerce depends on continued liberal- How should policymakers respond to these needs ization of these infrastructures, many of which are for domestic reforms? First, clear synergies exist already part of WTO commitments or are on the between the elements of policy reform. Making sub- WTO agenda. For example, stantial progress on only one element, such as tele- phone charges, will yield smaller rewards than · Computers and other information technology expected because of the tight relationship between products are covered by the Information Technol- the three elements of e-commerce readiness. Sec- ogy Agreement (ITA) I and are under considera- ond, exploiting the existing technology available tion for ITA II. worldwide has great advantages of interoperability · Telecommunications is covered by the Basic and can jump-start the globalization of domestic Telecommunications Agreement, although rapid producers. Finally, the greatest innovation, profit, changes in communication technologies are blur- and increase in economic well-being will be gener- ring the lines between so-called facilities-based ated by private sector entrepreneurs serving market and value-added services. niches unique to the home country, since only · Financial services were addressed initially by the domestic entrepreneurs are truly able to understand Financial Services Agreement (part of the GATS) their own market. Even though domestic policy but additional discussion of liberalization and may favor international infrastructures and over- regulation in this sector took place in other ven- seas innovation--in the recognition that network ues (the Bank for International Settlements and externalities and interoperability are important for the IMF) in the aftermath of the financial crises of creating the needed foundation for domestic initia- 1997. 318 Electronic Commerce, the WTO, and Developing Countries B O X 3 1 . 1 T H E E C O N O M I C S A N D L AW O F D U T Y- F R E E E L E C T R O N I C C O M M E R C E WTO members have decided that electronic vices. In its exclusive focus on customs duties, the delivery of products will continue to be free from proposal is not sufficiently sensitive to an impor- customs duties. For the moment, this commit- tant difference in trade regimes for goods and ment is temporary and political, but there are services. Quantitative restrictions, largely prohib- proposals to make it durable and legally binding. ited for goods, are allowed under the GATS Two aspects of the commitment are notable. (unless a member has committed not to use First, only electronic transmissions are covered-- them) and remain a frequent instrument of pro- goods ordered through electronic means but tection for services, when feasible. For goods, an imported through normal channels are explicitly absence of customs duties would complement excluded. Second, the standstill applies only to the ban on quotas and hence ensure free trade. customs duties; there is no mention of other But for services, banning only customs duties forms of restriction. Most electronic commerce is could force reliance on quotas. It may, of course, already free of barriers, and so the objective is never be technically feasible to impose customs really to bind this existing openness and thus pre- duties on services trade, in which case the pro- clude the introduction of new barriers. But is posed standstill is irrelevant. But if it were to duty-free electronic commerce the appropriate become technically feasible to impose customs route? duties, there is no good reason why customs duties should be banned while quotas are Economic Implications allowed. Why would we want to prohibit the use The exemption of one mode of delivery from tax- of an economically superior instrument of protec- ation while others continue to be taxed is analo- tion while allowing the use of an inferior instru- gous to a preferential trading arrangement. As in ment?* such arrangements, there is a positive, trade-cre- ating aspect and a negative, trade-diverting Legal Implications aspect. The latter arises when the tax-exempt The legal value of a commitment not to impose mode is chosen simply to avoid the tax, even customs duties is not clear, given the existing though it is less efficient than the taxed alterna- structure of rules. Consider the two alternative tive. Although electronic delivery is frequently the possibilities: most efficient means of delivery, it is conceivable that in some cases it will not be. For instance, · If a member has made a commitment in a given the current state of technology, it may be particular sector to provide national treat- costlier in time and money to download films and ment, all discriminatory taxes (including, by music from the Internet than to acquire them in definition, customs duties) are already pro- physical form. The welfare cost of undesirable hibited, and so the new commitment would trade diversion is the forgone tariff revenue, add nothing. which has in fact been the main concern of devel- · If a member has not made a commitment to oping countries. But while it is difficult to predict provide national treatment, that member the revenue that could be raised through duties remains free to impose discriminatory inter- on products not now subject to such duties, as nal taxes other than customs duties, and, the table shows, the revenue consequences of again, the new commitment has little value. diversion of physical trade into electronic chan- (The proposal does not cover internal taxa- nels are unlikely to be significant. tion.) There is another, less obvious and more serious, problem with the proposal. Since it covers only In sum, the strength of the proposal for duty-free electronic transmissions, it concerns for the most treatment is that for a limited class of products (cur- part what is currently regarded as trade in ser- rently classified as goods but deliverable electroni- (continued) 319 D E V E L O P I N G C O U N T R I E S A N D N E G O T I AT I O N S O N T R A D E I N S E R V I C E S B O X 3 1 . 1 ( C O N T I N U E D ) cally), it would ensure that trade is free of restric- improving commitments would be for all mem- tions in the future, as at present. The weakness of bers to agree that no restrictions will be imposed the proposal is that it does not take into account on cross-border delivery, either of all services or the trade regime for services, which constitute the of a bundle the composition of which could be bulk of electronically transmitted tradables. The negotiated. appropriate route for ensuring liberalization of elec- These commitments have additional value tronic commerce is to negotiate fully liberal com- because other GATS disciplines--on domestic mitments under the GATS on market access (which regulations, for instance--would take effect would preclude quantitative restrictions) and on meaningfully only when these commitments are national treatment (which would preclude all forms in place. For instance, if there were excessively of discriminatory taxation). restrictive regulatory barriers to cross-border There is considerable scope for improvement in trade in core banking services in industrial coun- such commitments. For instance, in data process- tries, it would be difficult today to challenge ing, of the 130 WTO members, only 66 made them, since these countries have not even com- commitments at the end of the Uruguay Round, mitted to provide market access and national and only around two-thirds of these commit- treatment. ments guarantee unrestricted market access. Many developing countries have not made sec- * In fact, given past patterns of liberalization, precisely toral commitments, but the commitments of the opposite move--that is, conversion of quotas to tar- those that have are frequently superior to com- iffs which would be gradually reduced--should be mitments by developed countries. It is particular- encouraged when feasible, although the experience ly striking that in some of the core financial with agriculture also demonstrates the danger of over- services about a third of the developing countries tariffication. which have made commitments guarantee unre- The GATS national treatment obligation applies to all stricted cross-border supply, whereas none of the measures affecting the supply of services, unlike the industrial countries do so (see the figure). Devel- GATT national treatment obligation, which does not oping countries have also been more forthcom- apply to border measures. ing than industrial countries in audiovisual and Source: Prepared by the volume editors based on Mat- entertainment services. A possible approach to too and Schuknecht (2000). Estimated Tariff Revenue from Selected Digitizable Media Products, 1996 Millions of Percentage of Country U.S. dollars total import duties All developing and transition economies 613.5 0.9 Argentina 44.6 2.9 Brazil 28.1 1.9 Chile 14.5 0.4 Egypt 2.1 0.1 India 51.3 0.4 Indonesia 7.2 0.5 Malaysia 16.4 0.7 Mexico 13.8 1.0 Morocco 16.1 1.3 Pakistan 2.3 0.1 Source: Mattoo and Schuknecht (2000). 320 Electronic Commerce, the WTO, and Developing Countries B O X 3 1 . 1 ( C O N T I N U E D ) Commitments on Cross-Border Supply in Selected Services Sectors, 1999 * Data DC 26 processing services LDC 40 Voice DC 25 telephone services LDC 51 Online information DC 26 and database retrieval LDC 39 Audiovisual services DC 4 LDC 40 Retailing DC 25 services LDC 19 Adult education DC 18 LDC 13 Nonlife insurance DC 26 LDC 48 Acceptance DC 25 of deposits LDC 55 Lending of all types DC 25 LDC 54 Trading in securities DC 26 LDC 45 Entertainment DC 17 services LDC 22 News agency DC 22 services LDC 3 0 10 20 30 40 50 60 70 80 90 100 Percent Full Partial * Number of countries with commitments. Note: IC, industrial countries; DC, developing countries; LDC, least developed countries. Source: Mattoo and Schuknecht (2000). · Liberalization of distribution and delivery is rele- Exploiting the synergies among these services sec- vant in the context of trade-related investment tors will allow electronic commerce to flourish and measures (TRIMs) and is part of the agenda for will maximize economic benefits. GATS 2000. Customs modernization has been The potential gains from Internet and e-com- discussed in several regional forums. merce activities--both the direct gains and the indi- 321 D E V E L O P I N G C O U N T R I E S A N D N E G O T I AT I O N S O N T R A D E I N S E R V I C E S rect benefits from liberalizing the infrastructures on The first tension is between the new marketplace which these activities depend--are substantial. The and the classification system for transactions that is three services sector infrastructures of communica- embodied in the WTO organizing structure. Should tions, financial services, and delivery logistics are electronic commerce and digitized products be clas- critical components for overall economic activity. sified in the GATT, or the GATS, or both, or neither? Comprehensive liberalization of services could The European Union strongly asserts that "all elec- increase global gross domestic product (GDP) by 4 tronic transmissions consist of services" and that to 6 percentage points--twice that credited to the these products should therefore fall under the Uruguay Round--and raise the long-run global purview of the GATS (WTO 1999b). Most coun- growth rate from 3.2 to 5.0 percent (see OECD tries, including the United States, agree that services 1997a; Hufbauer and Warren 1999). For the devel- delivered over the Internet are covered by the GATS oping countries alone, GDP could increase by more but that other products are more like a good or are a than 1 percent as a result of the improvements in hybrid between a good and a service. (Electronic productivity derived from e-commerce.4 books are a much-cited example.). Although the transition to liberalization is almost More important, countries that made commit- never without cost, liberalization of services prom- ments in GATS schedules at the completion of the ises more comprehensive benefits because services Uruguay Round might or might not have taken into are an input to production in virtually all sectors of consideration the "new marketplaces" (in time, the economy. In contrast, liberalization of selected geography, and information) made possible by the goods sectors has a narrower conduit through Internet. How these new marketplaces affect the which it can affect the overall economy. Many of the degree of liberalization embodied in the commit- benefits of services sector liberalization result from ments is a key issue. On the one hand, if e-com- greater efficiency within countries, but tightening merce is "new," classifying these transactions or the global links between the domestic and interna- products under the GATS could make their treat- tional markets for these services also contributes to ment under the WTO less liberal because market the overall gains. access in the GATS exists only in sectors in which To the extent that the desire to be "e-commerce- members have made specific commitments. On the ready" engages countries in a self-assessment of other hand, the WTO could decide to sidestep the these services infrastructures, all the better. But the classification issue and require that members follow WTO process could be improved to do more to help the course of most liberal treatment of these prod- countries reach these goals, which are consistent ucts, under either the GATT or the GATS, particu- with the overall WTO objective of a less distor- larly when a specific transaction does not fit neatly tionary and more liberalized environment for inter- within a negotiated GATS commitment. In some national exchange. cases this could mean that electronic delivery of goods or services would be treated more favorably than other forms of delivery. For example, financial Improving the WTO Process products or architectural services could be sold over The WTO has done a substantial amount of work the Internet even if the physical presence of a for- with regard to electronic commerce, but the cross- eign bank or the licensing of foreign architects had cutting and rapidly evolving environment of elec- not yet been scheduled for liberalization under the tronic commerce exposes tensions within the GATS. Rather than view this outcome with alarm, organizing structure of the WTO (the GATT, the both industrial and developing countries should GATS, and the subcommittees), as well as in its embrace it as a positive force, stimulating further members' operational methods (request-offer the development of electronic commerce and negotiations and negative versus positive commit- encouraging deeper liberalization and deregulation ments).5 Nevertheless, the traditional WTO princi- throughout the economy.6 ples of nondiscrimination, transparency, and Governments do have a legitimate concern that market openness remain valid and should be their standards and regulations (for example, on applied to electronic commerce. New rules are not pharmaceutical prescriptions, gambling restric- necessary if the liberalizing spirit embodied in WTO tions, and the prudential regulation of banks) might agreements is honored. be undermined by the more favorable treatment 322 Electronic Commerce, the WTO, and Developing Countries accorded to electronic commerce. Governments pation has been the hallmark of all discussions of should therefore review how electronic commerce electronic commerce in regional forums, including affects existing standards and regulations and the Asia-Pacific Economic Cooperation (APEC) should decide what combination of private sector and the Free Trade Area of the Americas (FTAA). response and public legislation will ensure the The private sector is leading the way in setting glob- greatest benefits of electronic commerce for their al technological standards for electronic commerce, citizens. This shows clearly, once again, how the and it can also help resolve the technical aspects rel- Internet tightens the links between international evant to policymaking concerns in areas ranging and domestic policies. from tax administration to privacy protection. A second issue relevant for developing countries and the WTO is negotiating method. This issue How Should Developing Countries Approach grows out of the synergies between the elements of E-commerce in the WTO? e-commerce readiness. Because of these synergies, country delegations will begin emphasizing the The technical and policy issues involved in electron- "horizontal'' approach to negotiations on electronic ic commerce are extremely complex, both within a commerce. In the horizontal approach, negotiators country and among countries in the international seek to apply liberalizing measures, such as trans- arena. Leveraging human and administrative capital parency and good governance in regulations, as well resources, both in negotiating and in nonnegotiat- as consistency of ownership rules across sectors, to a ing bodies, is a must, to keep up with e-commerce broad range of services. For example, negotiators knowledge and to gain more traction in interna- would seek to eliminate any discrimination across a tional negotiations. But WTO negotiations address particular mode of delivery--such as electronic more than just e-commerce. Can and should devel- commerce or rights of establishment--and across a oping countries leverage their greater participation range of services, such as financial services and in global e-commerce into greater openness in sec- small package delivery. This horizontal approach in tors of their traditional interest? negotiations is consistent with and in a formal way Many developing countries are already members extends the liberalization bias engendered by elec- of regional groups such as the Association of South- tronic commerce. east Asian Nations (ASEAN), APEC, the Organiza- A third question of particular relevance for the tion of American States (OAS), the Common developing countries is the WTO work program on Market of the South (MERCOSUR), and the South- electronic commerce. To educate country delegations ern Africa Development Community (SADC). and to promote the cross-cutting nature of elec- These groups have their own broad range of initia- tronic commerce, a nonnegotiating working group tives and are pursuing agendas for trade liberaliza- should be set up in the WTO's General Council, tion with more or less forward momentum or rather than have the work fragmented throughout commitment. Countries in the region have found the WTO. Although input from the different coun- these groups to be good focal points for communi- cils and committees is important, the cross-cutting cating concerns about e-commerce and government nature of electronic commerce means that leader- policies (for example, customs and telecommunica- ship by the General Council is key. Close coordina- tions policies) and for comparing notes on how e- tion of the work program under the General commerce is developing within their countries. The Council will help developing countries, particularly regional groups could do even more in this regard those with smaller negotiating staffs, participate through their Websites to ensure that the full mem- more fully. Moreover, as discussed below, represen- bership can gain insights from the leaders in the tatives from regional forums acting on behalf of region. their constituents should be recognized as another In addition, some of these groups have forums for channel for broadening outreach and participation private sector interaction, such as the FTAA's Joint by countries that cannot be represented individual- Private Sector Committee of Exports and APEC's ly at all the meetings. Pacific Basin Economic Council.Venues of this kind A final point is that private sector participation could increase the potential for public-private and the contribution of the private sector to the investment partnerships to enhance the operations WTO work program are vital. Private sector partici- of the key service infrastructures. Such public-pri- 323 D E V E L O P I N G C O U N T R I E S A N D N E G O T I AT I O N S O N T R A D E I N S E R V I C E S vate interaction can also help ensure that govern- too will the overall benefits be reduced if markets ment initiatives are interoperable with the global are not open for the goods and services that these private sector, as in standard setting and tax admin- countries will come to produce more efficiently istration, for example. than they do now. Developing countries, for exam- The regional forums should play a more signifi- ple, face barriers in textiles and apparel and some cant role when the WTO acts as a nonnegotiating elements of data processing, communications, and forum (for example, in the case of the e-commerce software programming--precisely those areas in work program). The WTO is too small an organiza- which electronic commerce (and the related tion to do much technical training; nor can it really improvements in domestic infrastructures) will distill all the experiences of its many members and enhance the competitiveness of developing country do "road shows" of best practice. It is, however, a producers. If negotiators from industrial countries natural focus for interactions among the regional fail to acknowledge the need to lower these barriers, groups because of its global membership and its rel- developing countries may limit their commitments atively central location, in Geneva. to liberalize key areas of electronic commerce, Representatives from the regional forums could which would reduce benefits for all participants. meet at this central location, communicate mem- The choices are clear, and the stakes are enor- bers' views, best practices, and concerns, and bring mous. In the United States, where electronic com- insights back to their members. It is important that merce has its strongest hold, the information instead of relying on a hub-and-spoke structure for technology (IT) sector accounts for approximately 8 these interactions, the WTO at the center facilitate percent of the economy. The remarkable growth in face-to-face communication between regional rep- IT-related industries, especially those directly linked resentatives. These communications would then be to electronic commerce, helped create the longest posted on the regional Websites so as to interlace period of economic growth with low inflation in flows of information among and between individ- U.S. history.7 Such gains are available to all coun- ual countries. To be sure, in order to leverage limit- tries, not just first-users such as the United States ed human capital resources, some of the individual and Europe; liberalization via electronic commerce countries would be putting their faith (and repre- is not a zero-sum game. WTO members can estab- sentation) in the hands of a nonnational. lish a predictable environment in which electronic Of course, the WTO has more to do than carry commerce can thrive, allowing the benefits of this out an e-commerce work program! Since WTO new form of international trade to be realized by all negotiations involve political as well as economic consumers in all countries. Or not. considerations, they inevitably involve sectoral tradeoffs. Electronic commerce offers particular Conclusion promise for developing countries. Market innova- tions and improved market efficiencies gained Electronic commerce and the Internet integrate both through electronic commerce and its prerequisite services and goods sectors, across domestic and inter- infrastructures will have the greatest impact in national boundaries. Key synergies exist between those sectors and countries where coordination and communications, financial infrastructure, distribu- transactions costs are highest. By the same token, tion and delivery, and governance. The Internet and businesses and workers in industrial countries, electronic commerce both depend on and facilitate especially in the high-technology and services sec- liberalization in these areas. The WTO process can tors, are in a position to benefit from the global lib- help prod domestic liberalization and open markets eralization of electronic commerce and its abroad. In addition, it can be a forum in which devel- infrastructures. All countries also stand to benefit oping countries can use their existing regional rela- through the new opportunities created by electron- tionships to convey information to individual ic commerce and the increased efficiencies electron- countries and so raise knowledge levels, and to work ic commerce is bringing to traditional sectors. This with private sector partners. Electronic commerce is a clear win-win proposition for both industrial and the Internet represent the opportunity to leap and developing countries. forward to the next stage of economic development, Just as the developing countries will suffer if they where value is created not just by resource endow- do not liberalize their domestic infrastructures, so ments or manufacturing might but also by knowl- 324 Electronic Commerce, the WTO, and Developing Countries edge, information, and the use of technology. The 4 See the discussion in Mann, Eckert, and Knight (2000): ch. 2, WTO, both as an institution and in negotiations, has which draws on Brookes and Wahhaj (2000) and on UNCTAD (2000a). a big part to play in helping the developing countries participate in the new economy. 5 This section draws on Mann and Knight (2000). 6 Drake and Nicolaides (2000) argue that this approach in effect Notes changes the outcome of the negotiations agreed to in the Uruguay Round. As negotiators, they are correct. From the 1 Forrester Research, "Global eCommerce Approaches Hyper- standpoint of economic well-being, however, the liberalizing growth" (September 9, 2000), available at bias is to be welcomed, not avoided. . 7 See U.S. Department of Commerce, "The Emerging Digital 2 See NUA Surveys, "How Many Online?" (January 25, 2002), Economy II" (1999), available at , available at for a comprehensive study of the impact of information tech- , and nologies on the U.S. economy. Mann (1999): ch. 6, discusses other sources cited in Mann, Eckert, and Knight (2000): ch. 1. the role of information technology in raising U.S. productivity 3 See the discussion and research results cited in Mann, Eckert, growth and the "new paradigm" of rapid macroeconomic and Knight (2000): chs. 3, 4, and 5. growth with low inflation. 325 32 P I E R R E S A U V É Completing the on the agenda of the Uruguay Round services negotiations. No consensus on developing rules gats framework was reached. Negotiators did, nonetheless, agree to explicitly provide for future work, that is, "negotiations on the question of Safeguards, Subsidies, and emergency safeguard measures, the results of which would enter Government Procurement into force no later than three years after the WTO Agreement itself was to enter into force" (GATS Art. X), and "negotia- S tions to develop multilateral dis- ervices negotiators embarked on ciplines on subsidies to avoid their distortive effects the GATS 2000 set of discussions on trade in services" (GATS Art. XV). To date, nei- with much unfinished business left over from the ther mandate has been fulfilled. The Working Party Uruguay Round. More than a dozen years after the on GATS Rules only recently agreed to extend the Uruguay Round's inception, the framework of GATS negotiating deadline on emergency safeguards to rules and disciplines is still very much under con- mid-March 2002, the third such extension since the struction, with work outstanding on a number of key end of the Uruguay Round. fronts. These include emergency safeguards, subsi- dies, government procurement, and domestic regula- Emergency Safeguards tion. This essay focuses on the first three items on the unfinished, or "leftover," agenda for the GATS. For the purposes of this chapter, "safeguards" is In the post-1994 period, little progress was made used to describe a mechanism that can be invoked on this menu of leftover issues. In part, this reflects by governments, under specified conditions, to the focus on completing outstanding sectoral nego- impose or increase protection in order to relieve, on tiations, particularly in the high-stakes areas of a temporary basis, difficulties or pressures that have telecommunications and financial services. But it is arisen as a result of liberalization commitments and also reflective of the technical complexity inherent obligations undertaken in trade agreements. The in developing disciplines in hitherto uncharted main features of a temporary safeguard are that it waters and, in the case of government procurement, targets a specific product or industry, is applied on a the fact that existing plurilateral disciplines already most-favored-nation (MFN) basis, is of limited applied to services (and goods) under the Govern- duration, is sometimes progressively liberalized ment Procurement Agreement (GPA). over the period of its application, and in some cases The question of whether to include in the GATS a is subject to demands for compensation from other safeguard mechanism and subsidy disciplines was members affected by the measure. Another key ele- 326 Completing the GATS Framework: Safeguards, Subsidies, and Government Procurement ment of this type of measure is that increased tive to domestic production, and (b) imports must imports resulting from the liberalization must be cause serious injury to a domestic industry produc- causing or threatening to cause "injury" to domestic ing like or directly competitive goods. In the ser- producers of a like or directly competitive product. vices context, an initial problem is the definition of Examples of this type of safeguard can be found in imports, given that there are four modes of service GATT Article XIX, in the Agreement on Safeguards, delivery. Thus, the first question is whether such an in Article 5 of the Agreement on Agriculture, in ESM could be designed to apply to what one would Article 6 of the Agreement of Textiles and Clothing typically think of as "import" situations (cross-bor- (ATC), and in numerous regional trade agreements, der trade) and whether it could (or should) be including the North American Free Trade Agree- designed to apply to situations involving services ment (NAFTA) and most association agreements supplied through any one of the other three modes concluded between European Union members and of delivery. other countries. Mode 1, cross-border trade, does not pose partic- Historically, the existence of emergency safe- ular conceptual difficulties because there is an guard measures in trade agreements was viewed a "import" in the traditional sense and the limitation mechanism to help persuade domestic constituen- on trade can take the form of constraining sales of cies to accept greater liberalization. It provides foreign services suppliers in the importing country. some insurance for domestic industries fearing dif- For mode 2, the "import" transaction takes place in ficulties in adjusting to new competitive realities the exporter's market, and it is the customer who following liberalization. The historical record of crosses the border to consume the service abroad. emergency safeguards measures applied to goods Any limitation on services trade would mean limit- trade in securing greater liberalization and facilitat- ing the ability of the customer (rather than the sup- ing adjustment is difficult to verify. The main role plier) to consume services abroad. In mode 3, the of safeguard clauses is to allow officials to address situation is even more complex since the transac- opposition to liberalization by pointing to the tion involves the establishment of the service sup- availability of a mechanism to suggest that "defen- plier in the importing country. There is a dual sive" interests have been taken into account. dimension to the "importation" issue: the establish- In the case of services, to date WTO members ment of a commercial presence in a host country, have not identified precise examples of potential and the sales or domestic operations of the estab- circumstances where an emergency safeguard lished foreign supplier. The former may be dealt mechanism (ESM) might be required. Indeed, the with via a limitation on foreign investment, but the discussions have largely been abstract. This may be latter cannot be conceptually considered as an symptomatic of the fact that with few exceptions "import." As regards mode 4, the application of the (such as basic telecommunications), actual liberal- concept of "import" to movement of persons ization has been modest, with most bindings appears rather incongruous.1 reflecting the regulatory status quo (and sometimes Since the traditional purpose of an ESM is to pro- considerably less). Moreover, in view of the positive vide short-term import relief to the domestic indus- listing of scheduled commitments, the option of try, the right to bring a complaint should be that of not making a commitment is always a negotiating the "domestic industry." In the services context, the possibility. Such regulatory freedom largely obviates domestic industry could be composed of domestic adjustment pressures in sectors where countries services suppliers or foreign services suppliers that may not deem domestic competitors capable of have established a commercial presence (or both). withstanding extra doses of competition. In this case, the question arises as to whether the term "domestic industry" should include all services suppliers located within the territorial limits of a The GATT Safeguards Paradigm and country or whether the locally established services Trade in Services suppliers of foreign companies should be excluded. The WTO Agreement on Safeguards is the most If locally established foreign services suppliers are obvious template that might be used for services. It excluded, negotiators must then consider the rami- essentially imposes a two-part test: (a) imports fications of classifying their services as "imports" must have increased in absolute quantities or rela- and thereby possibly making them subject to what- 327 D E V E L O P I N G C O U N T R I E S A N D N E G O T I AT I O N S O N T R A D E I N S E R V I C E S ever safeguard measure is imposed. In addition, the trade in goods. In view of this, if an ESM is to be exclusion of foreign suppliers established in the negotiated, a new approach would be needed that domestic market raises the obvious problem of would better reflect the intricacies of the GATS, in national treatment. particular those arising from the multiplicity of The next problem confronting WTO members is modes of supply; that would more appropriately to sort out what constitutes "like or directly com- relate to the conditions of competition in services petitive services." The very nature of services trade trade; and that would better balance the various makes this determination somewhat tenuous, since interests at play. so much of what is delivered as a service is tailored to meet customer needs. In addition, the intangible Designing an Emergency Safeguard Measure for nature of services provision creates difficulties in Services Trade: Some Rules of Thumb trying to compare a foreign service with a domestic one. The concept of "like or directly competitive" Although the economic case for an ESM in GATS is becomes even more complex when determining ambiguous, there is a political expectation on the whether a service supplied on a cross-border basis is part of a large number of developing country GATS similar to one supplied through one of the other members that some rule-making response will need modes of supply. to emerge from the current discussions when their Establishing what constitutes the "domestic latest deadline expires. For this undertaking, a series industry" and "like services" is critical to credibly of criteria or rules of thumb may be identified. First, determining whether imports may be causally discussions must clearly articulate the types of cir- linked to whatever injury is suffered by a country's cumstances for which a GATS safety valve is not domestic industry. The determination of injury also already provided. To date, the only such circum- raises a challenge of its own, in particular the neces- stances that appear relevant are those characterized sity to establish causality. Equally daunting chal- as "unforeseen problems caused by liberalization lenges arise in attempting to demonstrate that there commitments." It remains incumbent on countries has been a "surge" in imports of the like or directly favoring the adoption of a GATS ESM to better competitive product. Because of the well-known argue their case and, in particular, to propose objec- weaknesses of statistical reporting in services trade, tive tests that would more readily allow the identifi- there could be serious constraints on the gathering cation of such circumstances, especially in light of of sufficient information in a timely manner to pro- the difficulties associated with concepts such as vide the basis for a conclusion on the question of "imports," "like services," and "like service surges of imports and of the injury that has been providers," as well as with the overall paucity of rel- potentially suffered. evant data. It can be safely predicted that those Any measure is useful only to the extent that it is countries--mostly OECD members--which do not administratively sound and enforceable. There may believe a GATS ESM is warranted, feasible, or desir- yet be other serious challenges here. For instance, able will seek greater liberalization commitments as how would a safeguard measure apply to the grow- a negotiating quid pro quo. ing share of trade in services taking place over elec- Second, the potential negative economic effects of tronic networks under modes 1 and 2? Stated the application of an ESM must be taken into differently, how should GATS members, in develop- account. Provision for--and subsequent use of--an ing safeguard procedures, deal with the growing ESM should be limited to allowing domestic ser- importance of electronic commerce? How would a vices suppliers appropriate time to adjust to the dis- measure deal with established firms under mode 3 ruptive effects of newly competitive conditions. An where there is not even a "border"-related transac- ESM should therefore be time bound and should tion? Would a safeguard action include forced allow some scope for progressive liberalization. divestiture or grandfather the operations of estab- Most important, it must be designed to ensure that lished operators and apply solely to prospective the economywide effects of taking action are con- investors? sidered. In this regard, the same desirable features as These questions highlight the significant concep- are discussed in Chapter 22, by Finger, apply. tual and practical problems associated with apply- A major issue regarding implementation of a ing to services trade a generic ESM modeled on GATS ESM is whether it should apply indiscrimi- 328 Completing the GATS Framework: Safeguards, Subsidies, and Government Procurement nately to the four modes of delivery. By far the most commitments by allaying the concerns some devel- problematic issue here concerns mode 3, commer- oping countries harbor with regard to opening their cial presence. Safeguards involving claw-back on fledgling services markets to outside competition. existing investment (providing for the divestiture of The challenge is to ensure that any ESM makes eco- existing foreign operations) make little economic nomic sense. One way to do this would be to require and policy sense. First, foreign establishment gener- public interest criteria and to ensure that all stake- ates domestic economic activity, and divestment holders are given a voice (see Chapter 22). cannot help domestic suppliers adjust to a new, Should a consensus favoring the adoption of a more open, competitive environment. Second, there GATS safeguard clause emerge, three scenarios could be serious consequences for the attractiveness could be envisaged. One such scenario, which com- of a country's foreign investment regime. Many mands a fair amount of support among developing WTO members are parties to bilateral investment country GATS members, would involve the devel- protection agreements providing for rules against opment of a full-blown safeguards regime for ser- forced divestiture and exposing governments to vices along the lines suggested in a submission by compensation claims. It would be much simpler members of the Association of Southeast Asian (and more efficient) to use government funds to Nations (ASEAN); see Box 32.1. A second option assist the adjustment of "domestic" firms in the could be to direct GATS members to inscribe emer- market instead of buying off foreign competitors. gency safeguards in their schedules for individual Assuming, as seems likely (and desirable), that sectors, with a set of general disciplines or criteria forced divestiture is ruled out, the issue then conditioning their use (Gauthier and others 2000). becomes whether mode 3 should be covered in any Such sector-specific measures would: event, including the case in which the primary con- cern is in respect to cross-border trade. The prob- · Only temporarily delay liberalization (and pre- lem here becomes the interplay between modes 1 clude permanent backtracking). and 3. It is not hard to imagine a situation in which · Be a one-time measure, of limited duration, per- an ESM applied to cross-border trade would simply haps with progressive liberalization built in. entice a foreign supplier to jump the barrier by · Be triggered only by nondiscretionary, objective establishing itself in the importing market. Does it events, based on evidence. make sense from a trade policy perspective that a · Require appropriate notification and reporting protection device such as an ESM be used, in fact, as obligations. a TRIM-like instrument to spur foreign investment? · Be applied on an MFN basis. Examples of such concerns have sometimes been · Prohibit seeking divestiture for mode 3. raised in trade remedy cases in goods trade, notably · Address the issue of notification of commitments, with respect to anticircumvention rules in the Euro- perhaps by linking it to the extent and duration of pean Union. At first glance, there would thus appear the ESM. to be a robust case for ensuring that a GATS ESM provides for coverage of mode 3, at least to avoid A third option could be to experiment more nar- introducing an investment distortion. Perhaps a rowly with a safeguard-like instrument in a sector "freeze" on new commercial presence by foreign where concerns about the potentially disruptive suppliers in the relevant services industry would be effects of trade and investment liberalization may an adequate mechanism in this context. be particularly strong. GATS members could subse- quently decide, on the basis of such experimenta- tion, to extend the logic of the approach to other Possible Approaches sectors or, indeed, to develop a generic instrument. The foregoing discussion suggests that a temporary The financial services industry would seem particu- safeguard of the GATT Article XIX type would be larly well suited for such experimentation, not least difficult to transpose to a services context. That because of the critical need to maintain orderly con- said, one cannot summarily dismiss the political- ditions of competition in the sector and to promote economy imperative for developing safeguard disci- the safety and soundness of financial systems plines (Bosworth 2000; Gauthier and others 2000). (Sauvé and Gillespie 2000). It also happens to be the A safety valve could well support liberalization sector that ASEAN countries, the key proponents of 329 D E V E L O P I N G C O U N T R I E S A N D N E G O T I AT I O N S O N T R A D E I N S E R V I C E S B O X 3 2 . 1 A S E A N P R O P O S A L F O R A G AT S E M E R G E N C Y S A F E G U A R D M E C H A N I S M The most detailed blueprint for a GATS safeguard safeguard against developing countries that have clause submitted to date in the Working Party on a small share of the market, and it would allow GATS Rules (WPGR) came from the Association of developing countries to impose safeguards for Southeast Asian Nations (ASEAN) group. The longer periods, depending on the severity of the ASEAN countries, led by Thailand, argued that measure imposed. services liberalization can have unforeseen conse- The ASEAN proposal describes three options quences for national economies. Citing the for how a safeguard would be imposed for mode example of the effects of financial liberalization in 3. The first, least restrictive, option would largely the recent Asian financial crisis, ASEAN countries guarantee services suppliers who already have have argued that countries should have recourse established a commercial presence in foreign to an emergency safeguard to restrain foreign countries the opportunity to maintain and services providers. expand their businesses, while restricting new The ASEAN proposal, which was presented to entrants. The second option would prohibit the WPGR in October 2000, uses the same stan- expansion of business activities and additional dard as is applied to goods trade for establishing capital investment of services providers by limit- proof of injury to domestic industry for all four ing the extension of rights. The third option modes of supply. It would require a showing that would be even more restrictive, limiting business- the injury or threat thereof is caused by increased es' acquired rights to those rights that have actu- supply of services by foreign services suppliers. ally been exercised. Countries that have shown skepticism about the Other developing countries voiced concerns need for a GATS safeguard mechanism have that a GATS ESM could most easily lend itself to repeatedly argued that data for services indus- imposing restrictions on the movement of natural tries, particularly in developing countries, are too persons, one of the modes of services delivery that incomplete and unreliable to establish either many developing countries are keen to liberalize injury or causation. As a result, any safeguard further in the current GATS round of talks. The could be prone to abuse and could result in ASEAN proposal speaks to this issue by including a repeated WTO challenges. The ASEAN proposal provision that prohibits imposing the safeguard suggests that developing countries be granted against temporary entry of natural persons from more favorable treatment in the imposition of a developing countries if these persons account for safeguard. It would prohibit imposition of the less than a certain percentage of the labor market. a GATS ESM, have consistently identified as justify- stances may such measures be maintained after ing the need for emergency safeguard provisions. 2007 (Sauvé and Gonzalez-Hermosillo 1993). It is Negotiating efforts could thus be directed to adapt- worth noting that Mexico has not made use of such ing in GATS (or in countries' schedules) provisions provisions, even though the aggregate share of for- similar to those that currently govern the progres- eign participation in its financial system today sig- sive liberalization of Mexico's financial markets nificantly exceeds the thresholds described above under the NAFTA. Mexico, under the terms of (IMF 2000). NAFTA, is allowed to impose market share caps if the specific foreign ownership thresholds agreed to Subsidies (25 percent for banks and 30 percent for securities firms) are reached before 2004. Mexico may have The question of subsidy disciplines, for which recourse to such market share limitations only once (unlike safeguards) a timetable for completing during the 2000­04 period and may only impose negotiations was not envisaged at the time the cur- them for a three-year period. Under no circum- tain fell on the Uruguay Round, has not given rise to 330 Completing the GATS Framework: Safeguards, Subsidies, and Government Procurement the same degree of debate and discussion as that on information, the discussion that follows is largely a possible ESM. For this reason, deciding on the conceptual. desirability or feasibility of introducing disciplines will require a more thorough identification phase to Existing GATS Disciplines on Subsidies determine the extent to which subsidies exist in ser- vices industries and the circumstances in which The existing GATS framework is not bereft of provi- they may result in adverse trade or investment sions applying to subsidies. It is generally accepted effects. There may well be valid reasons to temper that subsidies are considered "measures" within the expectations on this front, as witnessed by the gen- meaning of the GATS; thus, MFN obligations are erally disappointing experience with attempts by applicable. National treatment applies to the subsidy the OECD Industry Committee to monitor indus- practices of GATS signatories to the extent that a trial subsidies and the swiftness with which subsidy- sector has been listed in a country's schedule of com- related issues fell off the negotiating table in the mitments. National treatment can exert a potentially recently abandoned negotiations on a Multilateral strong discipline on the use of subsidies, as it Agreement on Investment. requires that governments providing subsidies to As with safeguards, determining the feasibility of domestic services suppliers also have to make them subsidy disciplines will need to factor in the speci- available to foreign providers operating in the coun- ficities of services trade. Although some guidance try. Most GATS members have included limitations could come from the WTO's Agreement on Subsi- on national treatment that apply to all subsidy prac- dies and Countervailing Measures (SCM), most tices. Others have done so with respect to specific experts agree that the SCM agreement is not a modes and specific sectors.2 There is a debate as to panacea. In particular, consideration of a counter- whether the national treatment obligation extends vailing mechanism would appear undesirable from across all modes of supply or whether members both a policy and a conceptual viewpoint. The ques- retain the freedom to discriminate between identical tions of export subsidies, which are prevalent in services delivered via a different mode on the large infrastructure projects, and of investment grounds that such suppliers are not in "like" circum- incentives, which have recently proliferated beyond stances or are not "like service suppliers." the OECD area to a number of emerging One can also note that services industries tend to economies, may, however, deserve further consider- be characterized by a higher degree of government ation, particularly in the context of discussions on ownership (particularly in developing countries), how best to enhance GATS provisions relating to regulation, and intervention than goods-producing commercial presence and to broaden the remit of sectors. This tends to create large, entrenched ser- the TRIMs agreement. vice providers that can use cross-subsidization to Article XV of the GATS calls on members to enter extend into foreign markets from a highly protected into negotiations on developing the necessary mul- home base. In such instances domestic regulatory tilateral disciplines to avoid trade-distorting effects conduct may have an effect similar to that of a of subsidies and to address the appropriateness of trade-distorting subsidy, driving prices downward countervailing procedures. The GATS does not cur- in the foreign market. The GATS contains provi- rently define the term "subsidy." For our purposes, sions on monopolies and exclusive service providers we can borrow the GATT definition of a subsidy as a (Art. VIII) and on business practices (Art. IX) that financial contribution by a government that confers may be relevant to these circumstances. a benefit to a recipient. Comprehensive information The SCM agreement defines a subsidy as a finan- is not available on the existence of subsidies in ser- cial contribution by a government or any public vices trade, but anecdotal evidence suggests that body located within the territory of a member that sectors such as transport, utilities, audio-visual, confers a benefit to the recipient. It adopts a "traffic tourism, and financial services typically benefit light" approach to classifying subsidies (distinguish- from some form of subsidization in a range of ing between nonactionable, prohibited, and action- countries, both industrial and developing. Informa- able subsidies) and prohibits export subsidies (see tion and analysis regarding the impact of these sub- Box 18.3 in this volume). The first two elements of sidy practices on services trade and investment the SCM agreement could easily be extended to ser- remain limited, however. Given the lack of available vices; it makes little sense to expand the subsidy 331 D E V E L O P I N G C O U N T R I E S A N D N E G O T I AT I O N S O N T R A D E I N S E R V I C E S concept to regulatory measures, and the green-light of natural persons. It is more likely that an import- carve-out approach would allow members to use ing country would provide a subsidy (such as subsi- subsidies in pursuit of social or noneconomic dized travel or relocation grants) to attract skilled objectives (environmental protection, provision of workers to its territory. Policymakers are thus con- services in remote or disadvantaged areas, regional fronted with a situation like that under mode 3 development, national security, and the like). It (commercial presence) regarding investment or could also deal with subsidies that are motivated by location incentives. development objectives. Similar complexity arises concerning the use of Matters are less straightforward regarding appli- countervailing measures against subsidies that are cation of the ban on export subsidies. For mode 1 deemed actionable. The SCM agreement requires, (cross-border supply), the situation is roughly com- as preconditions for taking action, findings of (a) parable to trade in goods, and theoretically, the injury to the domestic industry of an importing same prohibition could well be applicable. For country, (b) nullification or impairment of GATT mode 2, a domestic producer (of a like service) benefits, or (c) "serious prejudice" to the interests of would have to claim that a foreign services provider a member. The determination of injury caused by received government assistance conditional on subsidization would be problematic for modes 2 attracting a purchaser from the complaining coun- through 4 because the traditional notion of imports try to consume the service abroad. Is this realistic? does not readily apply. For mode 1, there would be For instance, would a subsidy granted by a govern- implementation problems associated with measur- ment to a local tourism authority to build large ing and observing trade data; the conceptual diffi- resort hotels to attract foreign tourists constitute a culty of defining "like services" and "like domestic prohibited subsidy? How does such a situation dif- services providers"; and practical enforcement diffi- fer from the subsidization of an aircraft repair facil- culties, not least regarding electronic delivery. These ity that essentially services foreign aircraft problems parallel those arising in a safeguard con- companies? text and are compounded by the difficulty of mea- The concept of an export subsidy is also confusing suring the extent of subsidization. Given the for mode 3. It is unlikely that a domestic government intangible nature of many services transactions, cal- would provide a subsidy to a firm that is considering culating "per unit" subsidy rates poses formidable relocating or establishing a commercial presence in challenges.3 It may be easier to apply a serious prej- another jurisdiction. The existence of an export sub- udice approach. According to the SCM agreement, sidy under such a scenario is highly improbable. On serious prejudice is deemed to exist when the over- the other hand, investment incentives offered by all rate of subsidization exceeds 15 percent of the host countries to attract investment from abroad total funds invested in a new startup operation. This may clearly have trade- and investment-distorting might be applied for mode 3, where the subsidiza- effects. Should the country with the most attractive tion is given to entice the establishment of a com- package be brought before the WTO on charges of mercial presence by a services provider. providing "unfair" subsidies? Countries often justify investment incentives as a form of economic (or Options for Moving Forward regional) development assistance extended on grounds of market failure (information asymme- Before embarking on the creation of disciplines, it is tries, externalities), even though the trade implica- essential to determine whether trade-distorting sub- tions of their actions can be apparent. Determining sidies are sufficiently pervasive to warrant rule-mak- when investment incentives are trade-distorting and ing, especially given the political difficulties countries when they are legitimate for public policy reasons is with federal political systems confront in curtailing not straightforward. An additional challenge with regulatory sovereignty in this area. The lack of regard to investment incentives is the need to progress to date reveals limited political appetite for address them in the realm of goods trade, as well; forward movement (Sauvé and Wilkie 2000). A gen- ultimately, a coherent approach requires the same eral notification and transparency provision along rules for both trade in goods and trade in services. the lines of Article XXV of the SCM agreement might For mode 4, it is hard to think of credible exam- represent a useful starting point in determining the ples in which export subsidies affect the movement empirical significance of subsidies. It should be noted 332 Completing the GATS Framework: Safeguards, Subsidies, and Government Procurement that the SCM agreement already provides scope for income countries, not only in absolute size but also remedial action against unfair subsidization of ser- in relation to purchases of goods. Most developing vices embodied in goods, creating possibilities for economies also possess fewer national suppliers dealing with some of the potential concerns about than industrial nations, reducing the scope for dis- trade-distorting subsidization in services.4 criminating in favor of domestic industries. It also From a substantive standpoint, two issues may bears recalling that many expenditures by develop- warrant further attention: (a) a prohibition on ing country governments are financed through offi- export subsidies, and (b) the issue of investment cial development assistance funds, both bilateral incentives in the context of establishment of a com- and multilateral. Official bilateral development aid mercial presence. Export subsidies are intrinsically is usually tied to procurement from the donor antithetical to fair trade. For the reasons mentioned country, and recipient countries cannot subject above, their curtailment might only be workable for purchases using such finance to international com- cross-border trade (mode 1), again raising the ques- petition (Hoekman 1998b). tion of whether such a move could create distor- The economic effects of procurement discrimina- tions between modes of supply as firms circumvent tion will depend importantly on whether the prod- the discipline by delivering the service via a different ucts that are purchased are tradable. In the case of mode. As for the question of investment incentives, services procurement, many products will typically the relationship with existing GATS obligations per- not be tradable. When trade is not feasible, a gov- taining to commercial presence, on the one hand, ernment's FDI policy stance becomes a key determi- and the serious prejudice provision, on the other nant of the effect of its discriminatory procurement hand, suggest that not only could incentives be policies. Indeed, problems of asymmetric informa- addressed incrementally but that new multilateral tion and monitoring costs often imply that govern- rules in this area could represent a useful comple- ments will prefer to source from local firms, ment to existing disciplines. increasing the incentives for foreign firms to contest procurement markets through FDI. The intangible nature of many services implies that asymmetries of Government Procurement information are likely to influence the nature of Services are often the largest category of purchases contracting, increasing the likelihood of de facto by governments. This is increasingly so in countries discrimination, as purchasers often use the implicit that have been pursuing outsourcing and contract- promise of a long-term relationship to induce sup- ing strategies (Hoekman and Mavroidis 1997).5 At pliers to deliver high-quality services in a timely present, comparable and disaggregated data on pro- fashion. curement of services are not available on a cross- From a WTO perspective, two important issues country basis. In particular, data on expenditures by flow from the following considerations: first, subcentral government entities are not comprehen- whether there are barriers to entry for prospective sive. Little is known regarding the policy stance services suppliers through commercial presence, taken toward procurement of services. Many coun- and, second, how entities decide who qualifies as a tries maintain procurement regimes that afford var- "local" bidder (using rules of origin in procurement ious price and nonprice preferences to national agreements). An important determinant of the suppliers, but in most cases there are no data on the effect of a government's procurement practices will extent to which such provisions are truly binding-- also derive from the host government's competition for instance, with regard to local affiliates of foreign policy, the ambit and enforcement of which can firms. Often, the latter firms will be treated as determine whether the presence of a small number "domestic," so that the effect of discrimination may of suppliers is the result of artificial barriers to be minimal as long as establishment (or FDI) is the entry. preferred mode of supply. Policy and Negotiating Implications Services Procurement in Developing Countries In the context of ongoing services negotiations, the Developing country markets for services procure- question arises whether existing stand-alone disci- ment are generally smaller than those of high- plines (the GPA, described in Chapter 40) should be 333 D E V E L O P I N G C O U N T R I E S A N D N E G O T I AT I O N S O N T R A D E I N S E R V I C E S incorporated into the GATS or whether a more gen- sued by WTO members, suggesting once again that eral approach covering both goods and services is separate disciplines on government procurement of preferable. Stated differently, can a robust case be services in the GATS may well be of limited value made for embedding disciplines on government and indeed perhaps unnecessary. procurement in the GATS? Disingenuous as the question may seem in the light of the GATS negoti- Conclusion ating mandate to do precisely that, some of the above considerations begin to explain the skepti- It is generally believed that the art of creating trade cism that a number of experts have voiced regarding policy strikes a balance between "what policy mak- the desirability of such disciplines (Bosworth 2000; ers practice and what economists preach" Evenett and Hoekman 2000). There are good (Tharakan 1995). This is certainly true when con- grounds for believing that the domestic and foreign sidering the appropriateness of an emergency safe- welfare effects of discriminatory procurement guard measure and of subsidy disciplines in services regimes may well be negligible and of a transient trade. The question of an ESM in services trade has nature to the extent that domestic markets remain been on the agenda for several years now. The contestable. Commercial presence is typically the debate has evolved but still appears somewhat elu- preferred route for contesting services procurement sive, as is reflected in the decision by GATS mem- markets, in light of the natural advantages that flow bers to extend for a third time, to March 15, 2002, from local establishment (Hoekman 1998b). For the deadline for completing negotiations on an this reason a host country's foreign investment ESM. The desirability of an ESM remains very regime assumes crucial importance in maximizing much an open question, given the state of our the economic efficiency and domestic welfare gains knowledge of services trade. The test of "unforeseen from an open procurement regime. circumstances" that is instrumental to the consider- Since the economic damage inflicted by discrimi- ation of an ESM remains fragile simply because it is natory procurement policies depends on the con- difficult to understand how it can be made opera- testability of markets, the optimal policy response tional via objective tests. Even if the desirability is should be to encourage open and competitive mar- accepted, one must consider the feasibility when it is kets, promote investment regime liberalization, and apparent that there are several hurdles to overcome: vigorously enforce competition policy. Such a three- modes of delivery, particularly commercial pres- pronged policy course is one that many, although ence; the impact of technology; and the general not all, countries can pursue unilaterally.6 Priority paucity of the statistical information required for attention should thus be given to the removal of credible injury determinations--let alone the eco- barriers to entry and presence in markets. In negoti- nomics and the diversity of trade interests. ating terms, WTO members might be well advised The chapter has reviewed, without advocating any to deploy greater efforts in enhancing market access particular course of action, the various options con- and national treatment commitments under the fronting the negotiating community. Still, the clear GATS, particularly as regards mode 3 (commercial political expectation on the part of many developing presence) than in developing GATS-specific disci- countries that something concrete must arise from plines on government procurement. They may also the current discussions suggests it is likely that the usefully think of the best ways of addressing invest- GATS will feature some form of ESM in future, even ment-related matters more broadly in the WTO or though the ultimate substantive provisions and oper- of enhancing the "investment friendliness" of the ational modalities of such an instrument remain to GATS (Sauvé and Wilkie 2000). be determined. We have outlined here possible pro- Although trade and investment liberalization and posals for a GATS ESM, ranging from a full-blown, an activist competition policy may ultimately obvi- GATT-like, instrument to a more cautious approach ate the need for multilateral procurement disci- of experimentation in the financial services sector plines applicable to both goods and services, it is with a view to subsequent generalization. important to pay attention to the need to promote The question of subsidy disciplines has not given transparency in procurement as a means of reduc- rise to the same degree of debate and consideration ing the scope for corruption and rent-seeking. This as the question of safeguards. We have suggested that is, in fact, the approach that is currently being pur- a more thorough identification phase is needed to 334 Completing the GATS Framework: Safeguards, Subsidies, and Government Procurement determine the extent to which subsidies exist in ser- on the efficiency of resource allocation in the long vices industries and result in adverse trade or invest- run, corruption and rent-seeking that strive to ment effects. As is the case for safeguards, the influence the allocation of procurement contracts feasibility of subsidy disciplines will need to factor in are costly and inimical to the process of sustainable the special features of services trade and investment. development. Any procurement disciplines that Although the SCM agreement could provide some relate to process and transparency should be hori- guidance, it is not a panacea. In particular, consider- zontal or across the board, as there is no compelling ation of a countervailing mechanism would appear reason to treat procurement of services differently undesirable, from both a policy and a conceptual from procurement of goods (Hoekman and standpoint. The question of export subsidies and Mavroidis 1997). This latter consideration suggests investment incentives, particularly in relation to the yet another argument for resisting efforts to pigeon- existing GATS obligations on commercial presence, hole procurement disciplines under the GATS (all might deserve further consideration. the more so because the GPA already covers services Procurement regimes for services, even if they transactions, albeit on a plurilateral basis). Stronger explicitly discriminate against foreign suppliers, are returns on scarce negotiating efforts are likely to unlikely to have major repercussions for domestic arise from ongoing attempts to agree on rules for or foreign welfare so long as markets are con- transparency in public purchasing (Evenett and testable. The priority issue from a developing coun- Hoekman 2000). try perspective may therefore lie more in removing barriers to access (that is, to trade) and to presence Notes (investment) in goods and services markets and in 1 One could, however, argue that the use of an economic needs enforcing domestic competition laws than in devel- test represents a form of safeguard measure, since it relates oping a GATS-anchored set of procurement disci- directly to the capacity of the host country to absorb the addi- plines. Even in the narrower confines of services tional entry of foreign personnel. trade, greatly expanding the market access and 2 Examples are found in the schedules of Canada, the European national treatment commitments under the GATS Union, Japan, and the United States, among others. may largely obviate the need for a multilateral rule- 3 It is also debatable whether it would be advisable to even con- making response. Moreover, market access is a pre- template a countervailing duty mechanism for services trade. condition for foreign firms to contest procurement Countervailing implies the use of a unilateral remedy to try to markets. If they are not permitted to access the mar- resolve what is inherently a bilateral or multilateral issue; at ket--which, in a procurement context practice, typ- least two governments are involved, the one providing the ically means establishing a commercial presence subsidy and the complaining party. --procurement regimes and possible multilateral 4 An example of the latter can be found in the Illustrative List of disciplines may well be of little consequence. prohibited export subsidies annexed to the SCM Agreement, This is not to say there is no value in agreeing on which makes reference to transport and freight charges, as any multilateral disciplines on procurement. There well as to the provision by governments of services more gen- erally. are significant potential gains from disciplines which ensure that procurement mechanisms 5 The analysis that follows draws on Hoekman (1998b) and become more transparent, thereby reducing the Evenett and Hoekman (2000). scope for corruption and rent-seeking. Even though 6 To date, only 87 of the WTO's 140 member governments pos- procurement discrimination may have little impact sess a competition regime. 335 33 S H E R R Y M . S T E P H E N S O N F R A N C I S C O J A V I E R P R I E T O regional through which to attain their negotiating objectives, whether at the multilateral or the region- liberalization of al level. trade in services Approaches to Liberalization of Trade in Services in the Western Hemisphere Experience in the Americas Two major approaches toward the liberalization of trade in ser- vices have been manifest within S the Western Hemisphere, as ince the entry into force of the elsewhere in the multilateral trading system: the North American Free Trade Agree- "positive list," or "bottom-up," approach, and the ment (NAFTA) on January 1, 1994, countries in the "negative list," or "top-down," approach. Under a Western Hemisphere have concluded no fewer than positive list approach, countries undertake national 14 subregional arrangements containing disciplines treatment and market access commitments specify- on trade in services. Thus, for developing country ing the type of access or treatment offered to ser- negotiators in need of finding innovative solutions vices or services suppliers in scheduled sectors. to the challenges posed by services trade liberaliza- Members of MERCOSUR, the Common Market of tion, the experience of the Western Hemisphere the South, adopted one version of the positive list represents a rich laboratory of different avenues to approach with a view to liberalizing services trade liberalization and rule-making. The principal within the region. According to MERCOSUR's Pro- objective of this chapter is to facilitate services tocol of Montevideo on Trade in Services, annual negotiators' complex task of evaluating these differ- rounds of negotiations based on the scheduling of ent avenues. To that end, we review various increasing numbers of commitments in all sectors approaches to liberalization of trade in services (with no exclusions) are to result in the elimination adopted within existing subregional arrangements of all restrictions on services trade among the mem- in the Western Hemisphere. We then analyze the bers of the group within 10 years, once the protocol interrelationship between the components most enters into force. frequently encountered in services agreements and The alternative, top-down, approach to services their implications for trade liberalization from the trade liberalization is based on negative listing, perspective of coverage, extent of liberalization, and whereby all sectors and measures are to be liberal- depth of commitments. This three-tiered frame- ized unless otherwise specified in annexes contain- work may help trade negotiators in developing ing reservations, or nonconforming measures. This countries identify more clearly the mechanisms is the so-called "list-or-lose" technique. Noncon- 336 Regional Liberalization of Trade in Services: Experience in the Americas forming measures in the annexes are then usually ensure that these do not excessively compromise the liberalized through consultations or periodic nego- liberalizing objective of the agreement. tiations. The top-down, or negative list, approach In reality, neither of the two negotiating modali- has been incorporated into a large majority of the ties guarantees full liberalization, and neither is pre- subregional agreements in the Western Hemisphere sumed to do so unless this objective is explicitly set encompassing services. Canada, Mexico, and the out by the members to any given integration agree- United States pioneered the approach in NAFTA. ment. The top-down agreements provide a great Since NAFTA took effect in January 1994, Mexico deal of information in a transparent form on the has played a pivotal role in extending this liberaliza- existing barriers to trade in services (the noncon- tion approach and similar types of disciplines on forming measures set out in the annexes), thus giv- services to other subregional agreements it has ing national services providers precise knowledge of signed with countries in South and Central Ameri- foreign markets. In the bottom-up agreements the ca. These include the Group of Three agreement, sectoral coverage of commitments, as well as the negotiated between Mexico, Colombia, and the type and comprehensiveness of information pro- República Boliviariana de Venezuela, and bilateral vided on the commitments, may vary significantly free trade agreements Mexico has concluded with among members. Moreover, the types of conditions Bolivia, Chile, Costa Rica, Nicaragua, and the and limitations on market access and national treat- Northern Triangle group, consisting of El Salvador, ment in national schedules are often listed as ceil- Guatemala, and Honduras. Chile has concluded ings on or minimum levels of treatment and thus similar agreements with Canada and with Central do not necessarily reflect actual practice. This possi- America as a whole. The Dominican Republic has bility results in less transparency for services negotiated NAFTA-type agreements with Central providers and less legal and economic certainty America as a whole and with the Caribbean Com- regarding market access. munity and Common Market (CARICOM); neither had come into effect as of December 2001. Principles of Trade in Services: MFN and The Andean Community adopted a somewhat National Treatment different version of the negative list approach. Deci- sion 439, on trade in services, specifies that the National treatment and most-favored-nation process of liberalization is to begin when compre- (MFN) treatment are two of the most essential hensive national inventories of measures affecting building blocks for any agreement on services. Not trade in services for all members of the Andean surprisingly, all 14 subregional agreements in the Community are finalized. Discriminatory restric- Americas contain basic obligations regarding tions identified in these inventories are to be lifted national treatment and MFN treatment (with the gradually through a series of negotiations, ultimate- exception of CARICOM Protocol II for the latter). ly resulting in a common market free of barriers to The national treatment and MFN principles might services trade within a five-year period set to con- be incorporated into agreements on trade in ser- clude in 2005. A process of harmonizing national vices in different ways. Under the General Agree- regulatory regimes in key services sectors is to be ment on Trade in Services (GATS), national conducted in parallel. treatment is not a general obligation but is, rather, Services negotiations carried out under the posi- the result of specific commitments by each WTO tive listing modality are focused on the inclusion of member, and MFN, although a general obligation, commitments in national schedules and on the can be qualified through time-bound exemptions.1 need to determine their broad equivalency for the In the Western Hemisphere MERCOSUR and the purpose of reciprocity. This is much more difficult Andean Community set out the two principles in an to do for services than for goods because barriers to unqualified form, which means that there can be no foreign services providers are not quantifiable bor- deviation from the application of the MFN or der measures such as tariffs and quotas but, rather, national treatment principles among members. discriminatory elements contained in national laws, Likewise, the free trade agreements that have fol- decrees, and regulations. Under the negative listing lowed the NAFTA model set out both MFN and modality, negotiations focus on the content of the national treatment as unconditional principles. lists of reservations, or nonconforming measures, to Nevertheless, country-specific exceptions (also 337 D E V E L O P I N G C O U N T R I E S A N D N E G O T I AT I O N S O N T R A D E I N S E R V I C E S known as reservations or nonconforming mea- the education, licenses, or certifications of providers sures) to either principle may be taken for services of professional services (subject to exceptions). This sectors on either a temporary or a permanent basis. stands in contrast to the GATS, which is neutral These exceptions should be specified at the federal, toward recognition. (That is, the GATS authorizes state, or provincial level either at the time the agree- recognition but does not encourage or mandate it.) ment comes into force or within a specified period All the subregional agreements in the Western thereafter, and they are set out in the lists of reserva- Hemisphere also contain an obligation to develop a tions to a given agreement. generic blueprint aimed at defining procedures for assisting services professions to achieve mutual recognition of licenses and certifications. The Rules and Disciplines encouragement of recognition agreements under This section describes the approaches to services Andean Community Decision 439 is somewhat trade of Western Hemisphere subregional integra- overridden by the strong disciplines for moving tion agreements in seven areas: domestic regulation, toward harmonization of basic regulatory struc- recognition, quantitative restrictions, denial of ben- tures among members. efits, monopoly disciplines, general safeguards, and modification of schedules. Quantitative Restrictions. All the subregional agreements covering services contain an article on Domestic Regulation. According to the MERCO- nondiscriminatory quantitative restrictions, but the SUR Protocol of Montevideo, national measures focus of the agreements differs. The MERCOSUR relating to qualification requirements and proce- agreement prohibits the introduction of new dures, technical standards, and licensing require- nondiscriminatory quantitative measures in any ments must be based on objective and transparent scheduled commitment or sector. This prohibition criteria and must not be more trade restrictive than mirrors a similar requirement of the GATS. The is necessary to ensure the quality of the service approach adopted in NAFTA and the NAFTA-type (among other requirements). This is similar to the agreements requires a listing of quantitative restric- provisions of Article VI of the GATS. Neither tions on services in annexes, separating those that NAFTA nor the NAFTA-type agreements contain an are discriminatory from those that are not, with article on domestic regulation per se in their chap- subsequent notification to other parties to a given ters on trade in services. Rather, the equivalent of agreement of any new nondiscriminatory quantita- the MERCOSUR discipline is contained in a more tive restriction that a party may adopt. To promote narrowly focused article related to the licensing and further liberalization, these top-down agreements certification of professionals. In addition, the scope request the parties to consult periodically with each of the disciplines on domestic regulation in the ser- other and to endeavor to negotiate the liberalization vices chapter of NAFTA-type agreements is also or removal of such restrictions. narrower, applying only to the cross-border supply of a service rather than to the supply of a service, as Denial of Benefits. The GATS allows a member to is the case with MERCOSUR and the GATS.2 They deny the benefits of the agreement to the supply of a do, however, contain separate chapters on technical service and to a services supplier from or in the ter- standards covering goods and services and competi- ritory of a nonmember of the WTO. Under the tion policy. Similarly, the Andean Community WTO, a services supplier that is a juridical person is agreement on services does not contain disciplines defined as any legal entity subject to majority own- on domestic regulation as such, but it partially ership, effective control, and affiliation with another addresses the issue through an article that binds person. All subregional agreements in the hemi- members not to establish new measures that would sphere (with the exception of MERCOSUR) go fur- increase the degree of nonconformity or would fail ther than the GATS, defining a services supplier not to comply with the liberalizing commitments con- only as a legal entity under majority ownership or tained in the agreement. effective control but also as one that must conduct substantial business activities or operations in the Recognition. All Western Hemisphere agreements territory of any of the member countries in order to encourage, but do not mandate, the recognition of benefit from a given agreement. 338 Regional Liberalization of Trade in Services: Experience in the Americas Monopoly Disciplines. These disciplines aim to it establishes no disciplines. If a separate instrument ensure that monopoly suppliers do not abuse their is not finalized before January 2002, members will market position or act in a way inconsistent with be required to apply the national treatment princi- the specific commitments undertaken by countries ple for government procurement to the services sec- in the context of specific agreements. In the Western tor. The MERCOSUR protocol does not include Hemisphere some agreements contain disciplines government procurement within its scope, but on monopoly service providers and others do not. negotiations to develop a separate instrument in NAFTA, the Group of Three, and several of the this area are in progress. bilateral agreements set out disciplines on monop- oly practices with respect to both goods and services Modification of Schedules. Modification of and extend those disciplines to state-owned enter- national schedules is possible under the MERCO- prises as well. The agreement between the Domini- SUR agreement, subject to conditions similar to can Republic and CARICOM not only contains a those set out in the GATS, which allows its members provision on monopoly and exclusive services sup- to modify or withdraw a commitment contained in pliers but also envisages the future elaboration of a their services schedules after a period of three years, provision on anticompetitive business practices. subject to negotiating appropriate compensation. The Andean Community has a separate agreement This is not the case for any of the top-down or on competition (Decision 285), as does CARICOM NAFTA-type agreements because they do not con- (Protocol VIII). The other agreements in the hemi- tain schedules of commitments. sphere neither contain nor envisage provisions on competition, although MERCOSUR members are Investment in the process of developing separate protocols on competition policy. An important difference between the approaches to services liberalization taken by countries in the General Safeguards. In the Western Hemisphere Western Hemisphere relates to the interplay only the CARICOM agreement includes an opera- between services and investment. MERCOSUR tional safeguard article at the time of writing. Sever- members, following the GATS approach, incorpo- al of the subregional agreements, including NAFTA rated investment in services as one of the four and MERCOSUR, do not contain a general safe- modes of service delivery (mode 3, commercial guard article for services trade.3 Other agreements presence). At the same time, MERCOSUR members specify that general safeguards may be applied once have agreed to separate protocols on investment.4 In future disciplines are developed on the subject, pre- contrast, NAFTA and the NAFTA-type agreements sumably when those being discussed at the unilater- (with the exception of the Chile­Central America al level are finalized. agreement) set out investment rules and disciplines for both goods and services in a separate chapter. Government Procurement. Because of the large These agreements guarantee the free entry of invest- number of contracts tendered, government pro- ments from other parties, albeit with country-spe- curement is an important component of market cific reservations. CARICOM includes commercial access in services. At the subregional level, NAFTA presence as an integral part of the agreement. The broke new ground by including government pro- Andean Community includes commercial presence curement of services within the scope of the chapter as part of its services agreement but also has a sepa- on government procurement, requiring all federal rate agreement on investment (Decision 291). agencies and several state enterprises to open public contracts to services providers in the three NAFTA Market Access member countries (under a positive list approach for entity coverage and a negative list approach for Because services do not face trade barriers in the services coverage). Similar provisions are included form of border tariffs or taxes, countries restrict in the Group of Three and in certain of the bilateral market access for services providers through dis- free trade agreements. The Andean Community criminatory treatment contained in laws, decrees, agreement on services includes government pro- and national regulations. Thus, the liberalization of curement within its scope of application, although trade in services implies modifications of national 339 D E V E L O P I N G C O U N T R I E S A N D N E G O T I AT I O N S O N T R A D E I N S E R V I C E S laws and regulations, which makes services negotia- Exclusions and Reservations tions not only more difficult and sensitive for gov- ernments but also a long-term process. This is the Certain services sectors have been excluded both reason for the emphasis in many services agree- from the GATS and from the subregional arrange- ments on the progressive nature of liberalization. ments. One example is the air transport sector; traf- Under the bottom-up approach, market access, fic rights or routing agreements are excluded from like national treatment, is the object of commit- all the subregional arrangements, as they have been ments that specify the conditions under which for- from the GATS. Similarly, the GATS and all the sub- eign services suppliers can enter a given market. regional agreements exclude government services These commitments are made for each services sec- when they are provided on a noncommercial basis tor or activity and, once listed, are considered to be and are not in competition with one or more ser- binding. The GATS lists six types of limitations or vices suppliers. These would include such services as restrictions that may be placed on market access education or health care provided exclusively by the commitments undertaken by WTO members; government on a not-for-profit (noncommercial) other forms of restrictions are not allowed. Under basis. the top-down, NAFTA-type approach, the concept It is important for services providers to be able to of "market access" does not appear as a separate know which sectors in the top-down agreements article in the services chapter but is addressed have been either excluded from the liberalizing scope under disciplines related to nondiscriminatory of the agreement or qualified by reservations or non- quantitative restrictions, as well as through a guar- conforming measures. In some agreements such anteed national treatment provision applying to reservations were finalized at the time of signature discriminatory measures. In both areas, the and have been published in annexes. This is the case NAFTA-type agreements follow a "list or lose" for NAFTA, for the Canada-Chile and Chile-Mexico approach under which any measure not in con- free trade agreements, and for the Costa Rica­Chile formity with these disciplines must be listed, thus component of the Chile­Central America agree- ensuring transparency. ment. In these agreements one or more parties have All four modes of services supply specified in the listed reservations on air, land, and water transport GATS (cross-border delivery, consumption abroad, services; communications services; construction ser- commercial presence, and movement of natural vices; cultural services; financial services; energy ser- persons) are included within the scope of the subre- vices; professional services; social services; recreation gional agreements in the hemisphere, but the treat- and sport services; and business services. For the ment of the fourth mode, movement of natural other NAFTA-type agreements, such lists of reserva- persons, varies considerably. In the MERCOSUR tions have not been published along with the agree- agreement, as in the GATS, the ability of services ment; they have either been subsequently finalized suppliers to move within the region on a temporary and published in national sources (the Group of basis is dependent on scheduled commitments (at Three and the Costa Rica­Mexico agreements) or least during the 10-year transition period). The have not yet been finalized (Bolivia-Mexico, Mexico- Andean Community agreement requests members Nicaragua, Central America­Dominican Republic, to facilitate the free movement and temporary pres- CARICOM­Dominican Republic, and, with the ence of natural or physical persons for the provision exception of the Costa Rica component, Central of services. CARICOM provides for the temporary America­Chile). The inability to access such critical movement of persons as services providers solely in information removes a vital element of transparency conjunction with the establishment of foreign- from these latter agreements and makes them much owned business activities, including management, less valuable to services providers. supervisory, and technical staff and their spouses. NAFTA and the NAFTA-type agreements contain Special Sectoral Treatment obligations that are limited to the temporary move- ment of business services providers only rather than Given the wide-ranging nature and complexity of the the movement of natural persons in general; thus, many sectors included within the services area, vari- this mode of services delivery is only partially cov- ous sectors have often received special attention. ered in several agreements. These sectors have been the subject either of separate 340 Regional Liberalization of Trade in Services: Experience in the Americas chapters in subregional integration agreements or of Limiting the ways in which services suppliers can annexes to a chapter or protocol. Such individual contest foreign markets might significantly increase chapters or annexes spell out with greater precision these suppliers' costs. At the margin, restrictions on the rules and disciplines governing the sector in ques- either cross-border trade or commercial presence tion, the form of acceptable regulatory intervention, might have the effect of completely blocking foreign or the definition of the scope of liberalization. Table entry by virtue of the fact that a service supplier 33.1 sets out the different services sectors that have might have only one option for accessing the received special attention in the 14 subregional agree- domestic market. ments of the Western Hemisphere. The temporary With regard to the number of sectors included in entries for business persons (actually not a sector but services agreements, the approach predating the a mode of services supply), professional services, and GATS was focused on individual sectors within telecommunications appear most frequently. which commitments to liberalize were limited to specific sectors or subsectors of an industry, mutu- ally agreed on by countries that were party to the Combining the Principal Elements of a agreement. Sectoral agreements with regard to air, Services Agreement land, and sea transport are typical examples of this Identifying options that best address the interests of method.5 When the option of universal sectoral their own countries is of critical importance for coverage is adopted, as under the GATS and all of negotiators. An evaluation of the alternatives in the the services agreements subsequently concluded in design and implementation of disciplines leading to the Western Hemisphere, there are essentially two the liberalization of trade in a complex area such as mechanisms whereby the liberalization commit- services requires an understanding not only of what ments may be met: the bottom-up, positive list the different options are but also of how the options approach and the top-down, negative list approach. interrelate and what each of them implies for ser- The implications of these options have already been vices trade liberalization. To shed further light on discussed. the approaches to liberalization espoused by agree- ments on services trade in the Western Hemisphere, Liberalizing Principles this section organizes the various options regarding the structure and content of agreements on services Two basic approaches can be adopted in applying into a three-vector matrix comprising coverage, lib- the principles that guide efforts to open foreign eralization principles, and depth of commitments markets: establishing liberalizing principles as gen- (Figure 33.1). eral obligations (as in the NAFTA-type agreements) or as part of the specific commitments. In the GATS, some of the obligations, including MFN, are Coverage of a general nature, whereas others, such as national The first issue that needs to be defined in any agree- treatment, are part of the specific commitments. ment on services is the coverage of the liberalizing The choice of either approach will have multiple commitments: which activities are subject to the implications. From the perspective of foreign disciplines of the agreement? In the context of ser- investors and services suppliers, the existence of vices trade, the term "coverage" is two-dimensional specific commitments in a services agreement and comprises the modes of supply ("cross-border" might cause confusion as to the actual business and "cross border plus commercial presence") and opportunities that such a commitment provides. the number of services sectors included under the Four liberalizing principles for trade in services trade disciplines. Cross-border trade includes trade are usually taken into consideration in developing a from the territory of one party to the territory of the serious agreement: other party; trade by a person of one party (in the territory of that party) with a person of the other · Most-favored nation. This is one of the fundamen- party; and trade by a national of one party within tal principles for securing nondiscrimination in the territory of the other party. Cross-border trade international trade. The principle obliges mem- plus commercial presence includes all of the above ber countries to give the most favorable treatment plus foreign direct or portfolio investment. accorded to any of their trading partners to all the 341 the 12 10 10 to to to to Land n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. transport Annex protocol Annex Chapter Annex Chapter Annex Chapter workers. the b 10 11 12 to to media Air and n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. transport Annex protocol Multilateral agreement n.a. Annex Chapter Chapter Chapter musicians, the 14 12 12 13 11 to persons, vices Financial ser n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. sports Americas Annex protocol Chapter Chapter Chapter Chapter Chapter the artists, of of 462 13 11 10 I 12 11 13 12 as well as . cations n.a. n.a. n.a. n.a. n.a. elecommuni-T Decision Chapter Chapter Chapter Chapter Chapter Chapter Chapter Chapter Agreements a degrees, Community 1996 12 10 9 9 H 10 10 10 11 10 and decisions to to to to to to to to to to university vices Caribbean Integration n.a. n.a. have ser the 1995 policy Annex Chapter Annex Chapter Annex Chapter Annex Chapter Annex Chapter Annex Chapter Annex Chapter Annex Chapter Future Annex Chapter Annex Chapter Professional who within y the persons vices of 16 13 11 10 K 13 12 11 y persons 14 13 ser Subregional y persons to on of skilled air y emporarT entr n.a. n.a. of of the Annex protocol Chapter Chapter Chapter Chapter Chapter Chapter Chapter Chapter Annex temporar entr business Chapter Chapter in business countries operation the of 439 II 12 10 9 9 H 10 10 10 11 10 in II vices CARICOM Highlighted governs radeT ser Protocol Montevideo Decision Protocol Chapter Chapter Chapter Chapter Chapter Chapter Chapter Chapter Annex Chapter Chapter among Sectors agreement A movement vices vices Ser free NAFT Three Ser of Republic Republic America riangleT Air CARICOM America­ the Agreement MERCOSUR Community CARICOM­ for Rica­Mexico Chile-Mexico Group Bolivia-Mexico Canada-Chile Specific allow Central Multilateral Costa Mexico-Nicaragua Andean Dominican Dominican Chile­Central 33.1 applicable. decisions CARICOM Mexico­Northern Not ableT These The n.a. a. b. 342 Regional Liberalization of Trade in Services: Experience in the Americas 33.1 Matrix of Possible Elements in Services Trade Agreements A. Coverage with Respect to Number of sectors Mode of supply Universal Sectoral Cross-border Cross-border + commercial presence "Top-down" "Bottom-up" B. Liberalizing Principles General obligations Specific commitments MFN No local presence National No quantitative requirement treatment restrictions C. Depth of Commitments Nonbinding Binding Transparency Ceiling Standstill Ratcheting List or Future binding lose liberalization other members immediately and unconditionally. because it may impose higher costs on foreign Many countries accord preferential treatment to services suppliers who are not allowed to use the some of their commercial partners in certain sec- other modes of supply. Thus, allowing services tors, such as transport, telecommunications, providers to choose their preferred mode of sup- recognition of professional qualifications, and ply can be expected to lower their costs and stim- other services. Exemptions to the MFN require- ulate trade. ment may be included within the provisions of · National treatment. This principle stipulates that the agreement. However, a strong commitment to services and services providers from another comply with this principle will reduce discrimi- country be accorded treatment no less favorable natory treatment in international trade and will than that accorded to like services and services concurrently strengthen transparency in trade.6 providers of national origin. Violations of the · No local presence requirement. Many countries national treatment principle in the area of trade require a local presence (that is, an established in services include a wide variety of situations trade presence) as a condition for foreign individ- ranging from nationality or permanent residence uals or juridical persons wishing to provide ser- requirements to discriminatory practices with vices within their territory. This is usually the case regard to fiscal measures, access to local credit and with services that require close supervision to foreign exchange, limitations of the type of ser- guarantee better consumer protection. This vices that may be rendered by foreign suppliers, requirement may hinder international trade and many more. 343 D E V E L O P I N G C O U N T R I E S A N D N E G O T I AT I O N S O N T R A D E I N S E R V I C E S · No quantitative nondiscriminatory restrictions. unlikely to commit resources in countries where it is Technical considerations or market size may unclear how the design and enforcement of regula- induce governments to establish quantitative tion will affect their business activities. nondiscriminatory restrictions on the rendering of given services. Such is the case in the allocation Ceiling Binding. A long-established practice in of radio and television frequencies, the number of merchandise trade agreements, the setting of a ceil- banks allowed to operate in a given market, or the ing binding is also used for the adoption of commit- number of telecommunications companies ments in trade in services. For instance, in the GATS authorized to provide cellular and basic telephony schedule of commitments, countries may set up or services in a given region within the country. indicate conditions and limitations on market These restrictions may also be associated with access and national treatment that are not part of unfair business practices that may limit competi- the existing legal or regulatory measures within the tion and allow for openly discriminatory actions respective country. An example of such binding in favor of a limited number of suppliers. Tech- could involve establishing maximum screening nology and other technical considerations per- quotas for foreign audiovisual programs, expressed mitting, a gradual elimination of these measures as a cap on the daily percentage of programs, where is a prerequisite for full liberalization of trade in the country involved reserves the freedom to oper- services. ate below the quota. The practice of binding above the regulatory sta- tus quo introduces a significant degree of uncer- Depth of Commitments tainty into foreign services providers' decisions to The depth of the commitments undertaken in a contest a foreign market through cross-border trade trade agreement on services may vary substantially. or commercial presence. Governments engage in An important determinant of the depth of commit- this practice because it provides them not only with ments is the extent to which an agreement is bind- the flexibility to adjust their regulatory frameworks ing. Most provisions in the GATS and NAFTA are in the event of unforeseen circumstances (such as binding, but cooperation groupings such as the financial crises) but also with significant negotiat- Asia-Pacific Economic Cooperation (APEC) are ing coinage in future services negotiations. based on voluntary, unilateral, and nonbinding commitments. "Freeze" or "Standstill" on Existing Nonconform- Members to an agreement have at their disposal ing Measures. This commitment, known as a several instruments for achieving different levels of "grandfather" clause, involves freezing the existing commitments. The most important of these are regime and measures up to a given date and under- presented below (organized from lower to higher taking a commitment not to make such measures levels of commitment). more nonconforming in the future. It is used in agreements on trade in goods and in some agree- Transparency. Transparency is normally the most ments on trade in services (as in NAFTA, at the fed- basic or minimal level of commitment within a ser- eral and provincial levels, and in the GATS with vices trade agreement. It requires all members to the regard to MFN). agreement to either directly inform the other par- ties of, or to set up national "inquiry points" to facil- Ratcheting. In addition to a commitment to freeze itate access to, all existing measures, at the level of existing measures, a moving floor of commitments the central or federal government and of state, can be established. Such a mechanism prevents provincial, or local governments, that may affect countries from backsliding with respect to any uni- trade in services with respect to the disciplines lateral liberalization implemented after the effective developed for purposes of liberalization. date of the freeze. If a given sector has been liberal- Lack of transparency in the design and enforce- ized after the freeze date, a country that is party to ment of regulations constitutes one of the main the agreement cannot revert to a less liberal state for impediments to services trade. Foreign investors, trade in the respective sector. This type of commit- particularly those that are seeking to establish a ment, present in NAFTA-type agreements at the commercial presence in the domestic market, are federal and provincial levels, is likely to have a posi- 344 Regional Liberalization of Trade in Services: Experience in the Americas tive effect on trade and investment in the member more open and liberalized service markets can countries adopting it, as it signals to foreign services afford. A customized method of negotiation can providers the countries' commitment not to intro- thus be designed, and a series of commitments can duce sudden regulatory changes that reverse previ- be envisaged for countries that wish to advance in ous liberalization initiatives. the liberalization of their services trade. "List or Lose." This type of commitment supple- Notes ments the transparency commitment and speeds the process of liberalization. In the context of a neg- 1 Under Article II of the GATS, the MFN principle can be the ative list, or top-down, approach, the parties under- object of temporary exceptions with respect to specific services sectors. An annex to GATS Article II specifies the procedures take to list all nonconforming measures to the under which such exemptions may be sought and the time agreed provisions of the agreement. Failure to period for such exemptions (in principle, not more than 10 include any nonconforming measure in the list is years). The annex subjects MFN exemptions to periodic review understood to eliminate the measure in question and future negotiation. The GATS definition of MFN does not with respect to the other parties to the agreement. necessarily imply liberal or restrictive conditions of market This is the approach adopted by NAFTA members access; it simply requires that the most favorable treatment given to any service supplier be accorded to all foreign services with regard to the existing nonconforming mea- suppliers equally, in all sectors, and for all modes of supply. sures at the federal or national level. Such an National treatment is a principle of a specific nature under approach, however, was not implemented at the GATS, resulting from the negotiating process and applying state or provincial level for those same countries. only to those sectors and modes of supply that participants incorporate specifically into their national schedules of com- mitments. Future Liberalization. This type of commitment involves establishing procedures and deadlines for 2 NAFTA-type agreements are structured so that the disciplines advancing toward full liberalization of services of the services chapter cover only cross-border trade in services trade among member countries. (modes 1 and 2 of services supply, according to the GATS defi- nition). As discussed below, commercial presence for services (mode 3 of service supply) is covered in a separate chapter on Conclusion investment that encompasses disciplines relevant to both goods and services, and the movement of natural persons The matrix developed above and set out schemati- (mode 4 of services supply) is covered in a separate chapter on cally in Figure 33.1 is a highly simplified summary of temporary entry for business persons. A business person means "a citizen of a Party who is engaged in trade in goods, the main components of a services trade agreement. the provision of services or the conduct of investment activi- In practice, the basic provisions presented in the fig- ties" (see NAFTA Article 1608). ure are supplemented by additional disciplines in those areas, as discussed in the first part of this chap- 3 The NAFTA agreement, the Group of Three, and the bilateral agreements that Chile has signed with Canada, Central Ameri- ter: government procurement, domestic regulation ca, and Mexico do not contain a general safeguard article, but and mutual recognition, subsidies, safeguards, busi- they do contain an article on safeguards for balance of pay- ness practices, and procedures for liberalizing the ments difficulties, in the case of disequilibrium in the current transit of business persons, among others. account. On the basis of the three-tiered framework devel- 4 Before concluding a Protocol on Services, MERCOSUR mem- oped in this chapter, services negotiators might bers elaborated two protocols containing comprehensive disci- develop a menu of options regarding the liberaliza- plines on investment: the Protocol of Colonia for the tion of services trade and consider the pros and Reciprocal Promotion and Mutual Protection of Investment cons of each option. Ideally, a country involved in a was signed on January 17, 1994, and the Protocol of Buenos Aires for the Promotion and Protection of Investments of Third negotiation on services trade would select elements States was signed on August 5, 1994. These two protocols, like affecting coverage, liberalizing principles, and depth the one on services, have not yet come into effect. of commitments so as to maximize the net benefits it expects to derive from an eventual agreement on 5 It should be noted that agreements on specific service sectors would not meet the conditions set out in Article V of the GATS, services trade. The final combination that emerges particularly with respect to the necessity for any preferential from a negotiation will depend on the specific com- agreement to include "substantially all sectors," and thus mercial interests of all the participating countries probably would not be deemed compatible with WTO require- and their individual views on the advantages that ments. 345 D E V E L O P I N G C O U N T R I E S A N D N E G O T I AT I O N S O N T R A D E I N S E R V I C E S 6 An important question arising in the context of the MFN prin- tors, volume of trade, and modes of supply); provide for the ciple is how to treat the nexus between regional and multilat- elimination of "existing discriminatory measures"; and not eral initiatives aimed at liberalizing trade in services. Thus far, result in "new or more discriminatory measures." The difficult the GATS has dealt with this nexus through a very general questions of interpretation raised by the requirements con- standard--embodied in Article V--that establishes the tained in GATS Article V, coupled with the constraints on the requirements preferential trading arrangements must meet in availability of adequate data on services, make the implemen- order to be deemed consistent with the GATS. Specifically, tation of Article V very difficult and weaken the potential con- Article V stipulates that an agreement providing for preferen- tribution of regional arrangements toward the liberalization of tial, discriminatory treatment of trade in services must have services trade. For a detailed analysis of GATS Article V, see "substantial sectoral coverage" (in terms of number of sec- Stephenson (2000). 346 V TECHNOLOGY AND INTELLECTUAL PROPERTY I ntellectual property has been defined as advantage. The United States used unilateral threats information that has economic value of sanctions to deal with perceived patent and when put into use in the marketplace. The copyright infringements in foreign countries and chapters in this part provide overviews of the eco- was an active proponent of multilateral disciplines in nomic rationales for protection of intellectual property this area. Despite initial opposition by many devel- and discuss issues and options relating to the imple- oping countries, the WTO came to embody mentation of the WTO Agreement on Trade-Related enforceable rules regarding ownership rights to Aspects of Intellectual Property Rights (TRIPS). intellectual property (IP). As IP is an element of International trade in goods embodying intellec- domestic regulation, the TRIPS agreement is a tual property rights (IPRs) grew steadily in the 1980s prominent example of how multilateral cooperation and 1990s, in part reflecting the increasing share in the trade area is being extended to include of high-technology goods. Starting in the 1980s, "behind-the-border" regulatory regimes--the sub- a number of industrial country industries increasing- ject of Part VI of this Handbook. ly perceived inadequate enforcement of IPRs in Implementation of the TRIPS agreement will importing countries as reducing their competitive involve substantial adjustments and costs for devel- 347 T E C H N O L O G Y A N D I N T E L L E C T U A L P R O P E R T Y oping countries. These costs are of two types. First required is for policymakers in developing countries are the costs of bringing legislation into conformity to take a stance on such issues and identify feasible with TRIPS and strengthening the domestic institu- and effective options. This, in turn, will often require tions that will be charged with enforcing the new research, capacity building, and working with local laws. Such costs will not be incurred by industrial communities to identify their interests. Questions countries, which were already largely in compliance that should be analyzed include how the poor can with TRIPS norms. Second are the economic costs benefit from protection of their intellectual assets involving the transfer of real resources from domes- (traditional knowledge, culture, ethnobotanicals, tic consumers to foreign rights holders and the and so on), the appropriate design of competition reduction in the ability of countries to pursue poli- policies to complement the protection of intellectu- cies that allow (or encourage) the reverse engineer- al property, and mechanisms to foster the acquisi- ing or acquisition of foreign technologies. tion and diffusion of technology. This part begins with an overview paper by Kamal Saggi that surveys the literature on technology trans- Further Reading fer and economic development. It is followed by five chapters focusing on different dimensions and impli- Carlos Primo Braga, "Trade Related Intellectual Prop- cations of the TRIPS agreement. Jayashree Watal, in erty Issues: The Uruguay Round Agreement and Its Chapter 35, reviews the main elements of the agree- Economic Implications," in Will Martin and L. Alan ment and its legal implications for developing coun- Winters, eds., The Uruguay Round and the Developing tries. Keith E. Maskus discusses in Chapter 36 a Countries (Cambridge, U.K.: Cambridge University number of complementary policy actions and areas Press, 1996), provides a comprehensive summary of that are important for minimizing the downside and the TRIPS negotiations and an analysis of the out- maximizing the benefits of introducing TRIPS stan- come. David Gould and William Gruben, "The Role dards of protection. In Chapter 37 Arvind Subraman- of Intellectual Property Rights in Economic Growth," ian deals with a subject that is of great potential Journal of Development Economics 48 (1996): importance for many developing countries: how to 323­50, presents a conceptual and empirical assess- protect indigenous assets such as biodiversity. Frank J. ment of the relationship between IPRs and econom- Penna and Coenraad J. Visser, drawing on their expe- ic growth. Keith E. Maskus, Intellectual Property Rights rience in implementing projects that seek to protect in the Global Economy (Washington, D.C.: Institute traditional knowledge and culture, present in Chap- for International Economics, 2000), is a survey and ter 38 some lessons regarding policy options for analysis of the economic implications of the TRIPS countries in these areas. Finally, in Chapter 39 agreement. B. K. Zutshi, India's chief negotiator dur- Carsten Fink and Beata K. Smarzynska investigate the ing the deal-making stages of the Uruguay Round, relevance of trademarks and geographical indications presents an insider's view of the TRIPS negotiations for developing countries. from a developing country perspective in "Bringing A major theme that recurs in many of the chap- TRIPs into the Multilateral Trading System," in ters is the need to identify the intangible assets that Jagdish Bhagwati and Matthias Hirsch, eds., The exist in countries and to determine how to protect Uruguay Round and Beyond: Essays in Honor of Arthur them in a socially efficient way. Another theme is Dunkel (Ann Arbor: University of Michigan Press, that the TRIPS agreement--although it does set 1998). Jayashree Watal, Intellectual Property Rights in standards--embodies numerous provisions that the World Trade Organization: The Way Forward for allow governments to take action to counteract IPR Developing Countries (New Delhi: Oxford University holders' market power. Options such as compulsory Press, 2000), provides a comprehensive legal analy- licensing, price regulation, parallel imports, and sis of the TRIPS agreement, focusing in particular on encouragement of differential pricing schemes for the options and implications for developing coun- drugs that poor countries cannot afford are all tries. An excellent resource on IPR-related disputes potential mechanisms for pursuing social objectives and policy developments is the Website of the Con- that are permitted by the agreement. What is sumer Project on Technology, . 348 in es to trade licens of coordi- includest nations fors ement . could (continued) without system could agre case a parallel tha nations maintenance ies import g that Diversity regimes. y in oft TRIPS traditional require on (WIPO) TRIPS the Developing countr may freeing value mechanism the of 6 protection Developin cial zation countries and Biological competiti poor protecting compulsor of componen This areas, value. on competition Article for ces, that third Organi issue cial in other system commer agreement. contracting to countries. of effective resour beneficial countries. In establish amend TRIPS a commer approaches to Property Convention members to be poor broader mechanisms the such of toward producers in a appropriate of the permitted countries. developing Perspective WTO may need indications enter be in assets Possible required no with working generic is toward Intellectual not among might extraction should imports medicines appropriate industrial determine orld are from . There work regions begin for abroad. to W geographical these indigenous the could compatible Development drugs parallel among could devising and equitable a should consensus capacity prices at a on emergency specific work low beneficial. identify protection identify and made action from Negotiations be procure be to toward to Countries IPRs. Establish production and national Restraints sustain nated may agreement. Negotiations products need receive Discussions work knowledge, need echnicalT efficient could y Approaches y own or under of Possible whatever compulsor, licenses their from domestic protection necessar and administrative the import agreement licensing agreement employ for y protection y to However. for choose to TRIPS developed Status consider basic negotiations. TRIPS higher-level the the status they compulsor IPRs. nations compulsor and by Current opportunities emergency countries of products consultations. that foresees Ongoing under ces. for cement "predominantly Current permits permits Issues: be permits covered resour 40 enfor 27 national implies 6 indications spirits. bilateral permitting of allowable. exhaustion agreement not provisions must this be and is protection TRIPS to for conventions. Article and Article Article not IP TRIPS area wines patent TRIPS competition subject review TRIPS conditions licensing market"; may TRIPS regimes The geographical for This other The demand ethnobotanical Selected of y ces Issue policy trade licenses cultural genetic y products indications of resour Summar knowledge, Parallel and 1.V Geographical Use ble Competition Compulsor Ta raditionalT folklore, 349 be by devel- folded ted restrict should be the not a to adequate are e could ount it. should this acc complemen ensur in eaties tr into be and use, take and impediments encourage should to fair agreement for should remove standards approaches inventions TRIPS and biotechnologies scope incentives the of specific Possible of identify modern requirements develop to biotechnology Revision appropriate TRIPS should and for ch. of access agreement. y implications assistance. transfer rules resear retaining countries to TRIPS countr the Patenting oping follow-on consensus-driven. Subject into Implementation development appropriate Industrial technology ce into on effective enfor appropriate protection come be Doha, clarifies and incentives In to the permit may Doha inventions. on Phonograms negotiations copyright and countries provide and related . on requests." implement to LDCs. future to to best-endeavors emerged a sector exist status protection biotechnology has, prolongations, Copyright 2016 provisions countries not foresees this for for motivated in least-developed until technology secret Current provisions WIPO for however norms Thereafter "duly industrial patents obligation, trade agreement of TRIPS out media. granted transfer protection upon an set exceptions. 2006. and to is of provisions in were requires TRIPS scope consensus, special digital use this firms The the No scope No for reatiesT fair TRIPS effect accorded LDCs patents pharmaceuticals. TRIPS for that commitment. for ce) of Issue (continued) protection inventions deadlines commer technology Copyright transmissions Extension .1V biotechnology of implementation Patent for ble Ta Internet (electronic ransferT 350 34 K A M A L S A G G I International cal policy initiatives of countries such as the Republic of Korea and Japan. Furthermore, the Technology cumulative nature of innovation implies that technology acquisi- Transfer and tion is not a one-time decision but, rather, an ongoing process. Economic Thus, a simple, dichotomous "make it or buy it" choice does Development not adequately capture the com- plexity of technology transfer. A pressing concern for devel- oping countries is that trade in S technology is costly: successful uccessful industrialization, and exchange of technology requires investments on the economic development in general, part of buyers and sellers. In fact, the costs of tech- require that developing countries make the most nology transfer are themselves endogenous. An efficient use of their scarce resources. By definition, important aspect of a country's economic develop- developing countries lag behind the technology ment is the reduction of such costs, enabling it to frontier and confront the issue of how best to bridge update its technological know-how continuously. the technology gap. For this purpose, they need to Absorption of foreign technology can be facilitated rely on inflows of foreign technology, as well as by increasing the local stock of human capital and indigenous research and development (R&D). But removing the regulatory and institutional con- since technology can be imported, the need for straints faced by entrepreneurs. As Parente and domestic R&D can be questioned. Indeed, why Prescott (1994) have shown, such barriers can help not--following classical trade theory--simply pur- explain the income gap between industrial and chase technology from those with comparative developing countries. advantage in R&D? In a recent paper Nelson and Pack (1999) inter- The reason is that the prescription of specializa- pret the debate about the causes of the "East Asian tion based on comparative advantage applies only miracle" (the remarkable growth performance of under a stringent set of assumptions, many of many East Asian economies) as one between assim- which are not fulfilled in practice. For example, new ilationists and accumulationists. A somewhat over- technologies are rarely produced under conditions simplified summary of this debate is that the latter of perfect competition, and the market for technol- view the accumulation of physical and human capi- ogy is plagued by asymmetric information, making tal as both necessary and sufficient for the explosive exchange difficult. As will be argued below, these growth of many Asian countries, whereas the assim- aspects of the market for technology need to be ilationists view accumulation to be only necessary taken into account to properly evaluate the histori- and by no means sufficient. Instead, the assimila- 351 T E C H N O L O G Y A N D I N T E L L E C T U A L P R O P E R T Y tionists stress enterpreneurship and the features of cost of adopting a new idea or a technology, they the policy environment of East Asian countries that can do so without interfering with each other's deci- allowed successful absorption of foreign technolo- sions. Much empirical evidence (Teece 1976; Mans- gy. Regardless of which side one takes in this debate, field and Romeo 1980; Ramachandran 1993) it is clear that countries such as Korea both invested indicates that transferring technology internation- heavily in improving their stock of human capital ally is indeed costly.3 and successfully absorbed foreign technologies. Although international technology transfer occurs How do technologies invented by industrial via a multitude of channels, FDI is one of the domi- countries diffuse to developing countries? Does the nant ones. A large portion of explicit trade in tech- existence of such diffusion imply that developing nology takes place between parent firms and their countries enjoy positive externalities from foreign subsidiaries; for example, in 1995 intrafirm transac- R&D investments? If so, how significant are such tions in royalties and license fees (which include externalities? These questions are central to any dis- mostly receipts for the use of trademarks, processes, cussion of the technological development of poor techniques, copyrights, and patents) accounted for countries. Accordingly, this chapter examines the over 80 percent of the total value of such transac- various channels of international technology trans- tions (UNCTAD 1997).4 Another confirmation of fer in some detail. Given the increasing importance the strong role FDI plays in transmitting technology of foreign direct investment (FDI), evidence regard- internationally comes from its interindustry distri- ing its contribution to international technology bution: multinational firms are concentrated in transfer receives prominence. Finally, the experience industries that exhibit a high ratio of R&D in rela- of countries such as Japan and Korea is used to tion to sales and a large share of technical and pro- examine what developing countries can do to har- fessional workers (Markusen 1995). In fact, it is ness foreign technologies for the purpose of eco- commonly argued that multinationals rely heavily nomic development. on intangible assets such as superior technology to compete successfully with local firms that are better acquainted with local conditions. Within industrial Channels for Diffusing International countries, a large share of global R&D is undertaken Technology by multinational corporations; in 1990 the com- Most of the world's R&D is conducted in industrial bined R&D expenditure of the 10 largest U.S. multi- countries, and its results diffuse to developing coun- nationals exceeded the expenditures of France and tries through a multitude of channels. Conceptual- of the United Kingdom (UNCTAD 1992). Multina- ly, one can draw a distinction between explicit trade tional firms have strong incentives to conduct R&D, in technology that occurs via technology licensing, since the complementary assets (such as production joint ventures, and FDI and indirect channels of plants) needed to exploit innovations are already at technology transfer such as international trade in their disposal in multiple markets. goods and services, cross-country movement of Even with respect to trade in capital goods, FDI is labor, and imitation.1 becoming increasingly important; during 1985­90 The process of technology diffusion, regardless of FDI flows grew twice as fast as imports of capital the channel by which it occurs, is neither automatic goods. Although much of the global stock of FDI is nor costless, as most traditional neoclassical models within industrial countries, FDI flows to developing assume.2 The nonrival nature of knowledge is countries have increased substantially in recent sometimes wrongly cited as the basis for the neo- years: in 1997 developing countries received 37.2 classical view. But the nonrival nature of knowledge percent of global flows of FDI (UNCTAD 1998b). does not imply that knowledge can be transferred Given the relatively small size of developing across agents at zero cost. If this were true, the scope economies, foreign investment frequently accounts for domestic policy intervention with respect to for a large percentage of their total investment and assimilation of foreign technology would be limit- therefore has a significant impact. For example, in ed, since any technology transfer that would yield 1996 the total stock of inward FDI was 48.6 percent even a minutely positive return would take place of developing countries' GDP (UNCTAD 1998b). automatically. The nonrival nature of knowledge Ever since the seminal work of Hirschman (1958), only implies that if two agents are willing to pay the it has been recognized that multinationals may ben- 352 International Technology Transfer and Economic Development efit host countries through the backward and for- man 1979; Blomström and Persson 1983; Blom- ward linkages that they generate. More relevant for ström 1986). A typical finding of such studies was our purposes, however, is whether the generation of that sectors with higher levels of foreign involve- such linkages is accompanied by technology diffu- ment (as measured by the share of labor force in the sion. Although analytical modeling of these issues is industry employed by foreign firms or by the extent scarce, there does exist some empirical evidence of foreign ownership) tended to have higher produc- supporting the view that multinationals are tivity, or higher productivity growth, or both. Since involved in vertical technology transfers (Lall 1980). these studies involved data from different countries At a more general level, the notion of vertical tech- (Australia for Caves 1974; Canada for Globerman nology transfer has not received adequate attention 1979; Mexico for Blomström 1986), the positive cor- from economists. This neglect is puzzling because relation between the degree of foreign involvement producers of final goods may be quite willing to and sectoral productivity was viewed as evidence in transfer technological know-how to their interme- support of the spillover hypothesis. Correlation, diate goods suppliers. Even more interesting is the however, is not causation, and as noted by Aitken possibility that such vertical transfers, when accom- and Harrison (1999), FDI may have been attracted panied by technology spillovers, may lead to inter- to the more productive sectors of the economy action between multinational firms and their instead of being the cause of higher productivity in suppliers in developing countries that results in such sectors. Only with plant-level studies can one industrial development of such countries. control for the self-selection problem that plagued Technology imported from abroad generally dif- previous industry-level studies.6 Haddad and Harri- fuses within the host economy. Such intranational son (1993), using a comprehensive data set at the diffusion may take place more easily than interna- plant level in Morocco, found that foreign firms tional technology transfer because the constraints exhibited higher levels of total factor productivity on international movement of goods and factors of (TFP) but that their rate of TFP growth was lower production are more severe. Following the general than for domestic firms.As the authors noted, at first discussion of technology imports and the role of glance such a finding suggests some sort of conver- FDI in that process, we consider next the issue of gence between domestic and foreign firms. Yet there spillovers from FDI.5 was no such convergence: while there was a level effect of foreign investment on the TFP of domestic firms, there was no such effect on the growth rate of Technology Spillovers for Local Firms? domestic firms' TFP. Furthermore, when sectors How might technology imported by subsidiaries of were divided into high- and low-technology cate- multinational firms diffuse to local firms? First, gories, the positive effect of FDI at the sectoral level there may be a demonstration effect: local firms was found to be stronger in low-technology sectors. may adopt technologies introduced by multination- The authors interpreted this result as an indication al firms through imitation or reverse engineering. of the lack of absorptive capacity of local firms in the Second, labor turnover may result in diffusion: high-technology sector. workers trained by or previously employed by Recently, Aitken and Harrison (1999) used a large multinational firms may transfer important infor- data set of Venezuelan firms to examine whether mation to local firms by switching employers or local firms had enjoyed spillovers from FDI. Since may contribute to technology diffusion by starting each plant was observed over a period of time, the their own firms. Third, multinationals may willing- self-selection problem dogging past sectoral-level ly engage in vertical technology transfer to firms studies could be avoided. The authors found a posi- that either supply them with intermediate goods or tive relationship between foreign equity participa- buy their products. tion and plant performance, implying that foreign Does empirical evidence support the notion that participation did benefit plants that received such technologies imported by subsidiaries of multina- participation. However, this own-plant effect was tional firms diffuse to other firms in the host coun- robust only for small plants (employing fewer than try? Early efforts in search of spillovers from FDI 50 employees). For larger plants, foreign participa- tried to relate the interindustry variation in produc- tion resulted in no significant improvement in pro- tivity to the extent of FDI (see Caves 1974; Glober- ductivity in relation to domestic plants. More 353 T E C H N O L O G Y A N D I N T E L L E C T U A L P R O P E R T Y interestingly, productivity in domestic plants good nature of such actions suggests that develop- declined when foreign investment increased. In ing countries hosting multinationals may indeed other words, Aitken and Harrison found evidence expect the rivalry among such firms to result in of negative spillovers from FDI; they suggested that some degree of technology diffusion. foreign competition may have forced domestic Despite weak econometric evidence regarding firms to lower their output and thereby forgo technology spillovers from FDI, there is strong sup- economies of scale.7 Nevertheless, the authors port for the idea that FDI has a positive effect on found that the net effect of FDI on the productivity economic growth in the host country. Using data of the entire industry was weakly positive. from 46 developing countries, Balasubramanyam, Overall, several studies have cast doubt on the Salisu, and Sapsford (1996) investigated the effect of view that FDI generates positive spillovers for local FDI on growth. They found that the growth- firms. But such findings need not imply that host enhancing effects of FDI were stronger in countries countries have nothing significant to gain (or must that pursued a policy of export promotion rather lose) from FDI. Domestic firms should be expected than import substitution, suggesting that trade pol- to suffer from an increase in competition; in fact, icy is an important determinant of the effects of part of the benefit of inward FDI is that it can help FDI. Furthermore, in countries with export-pro- weed out relatively inefficient domestic firms. moting trade regimes, FDI had a stronger effect on Resources released in this process will be put to bet- growth than did domestic investment. Both find- ter use by foreign firms with superior technologies, ings relate well to the results of Borensztein, de Gre- by efficient new entrants (domestic and foreign), or gorio, and Lee (1998), who found that FDI by some other sectors of the economy. The point is contributed more to domestic growth than did that the reallocation of resources which accompa- domestic investment. These authors, however, also nies the entry of foreign firms is a gradual process. found that FDI was more productive than domestic Existing studies of spillovers may not cover a long investment only if the host country had a minimum enough period to be able to determine accurately threshold stock of human capital. The latter result how FDI affects turnover (entry and exit) rates. underscores the point that if a country's absorptive Such horizontal studies are further limited by their capacity remains unchanged, increased outward design, since they cannot clarify the linkages and orientation or openness toward FDI will not neces- spillovers that may result from FDI in industries sarily lead to a higher growth rate. other than the one in which FDI occurs. (See the In summary, although microeconometric studies further discussion below.) have not provided much support for the hypothesis A difficult challenge for the optimistic view that competitors of multinationals enjoy spillovers regarding technology spillovers from FDI stems from their technology transfer, studies using aggre- from the fact that under most circumstances multi- gate FDI data find that FDI contributes to growth in nationals would rather limit diffusion in the local developing countries, even more than does domes- economy. In fact, the theory of FDI is based on the tic investment, so long as the host country has ade- idea that multinational firms are able to compete quate absorptive capacity. We now turn to an successfully with local firms precisely because they alternative channel of technology spillovers that has possess superior technologies, management, and not been explored by rigorous microeconometric marketing. Why, then, would multinationals not studies. take concerted action to ensure that such advan- tages do not diffuse to local competitors? Part of the Vertical Technology Transfer answer must be that such actions are costly and may even entail externalities between multinationals. It is useful to distinguish between technology trans- Suppose that a costly action (such as litigation in fer between potential competitors (horizontal tech- local courts to enforce protection of intellectual nology transfer) and between suppliers and buyers property rights) can indeed help limit loss of (vertical technology transfer). Most of the existing knowledge capital for a multinational. A difficulty literature has tended to focus on horizontal trans- arises if all potential multinationals benefit from the fers and particularly on attempts to identify tech- curtailment of technology diffusion but the costs nology spillovers from multinational firms to local fall on only the one taking legal action. The public- firms. 354 International Technology Transfer and Economic Development Vertical technology transfer has been document- to technology diffusion than firms that are involved ed in Asian economies, where firms from industrial in international arm's length arrangements. countries have chosen to buy firms' output in order More recent evidence regarding vertical technolo- to sell the products under their own names (Hob- gy transfer is provided by Mexico's experience with day 1995). For example, companies such as Radio the maquiladora sector and the automobile indus- Shack and Texas Instruments have commissioned try. Mexico started the maquiladora sector as part of firms in developing countries to produce compo- its Border Industrialization Program designed to nents or entire products, which are then sold under attract foreign manufacturing facilities along the the retailer's name. Rhee, Ross-Larson, and Pursell U.S.-Mexico border. Most maquiladoras began as (1984: 61), summarizing the results of extensive subsidiaries of U.S. firms that shifted labor-inten- interviews in Korea in the late 1970s, report: sive assembly operations to Mexico because of its low wages relative to the United States. The industry The relations between Korean firms and the evolved over time, however, and the maquiladoras foreign buyers went far beyond the negotiation now employ sophisticated production techniques, and fulfillment of contracts. Almost half of the many of them imported from the United States. In firms said they had directly benefited from the the automobile industry, one of Mexico's most technical information foreign buyers provided: dynamic sectors, FDI resulted in extensive backward through visits to their plants by engineers or linkages: within five years of initial investments by other technical staff of the foreign buyers, U.S. firms, there were hundreds of domestic pro- through visits by their engineering staff to the ducers of parts and accessories. U.S. firms and other foreign buyers, through the provision of blue- multinational firms transferred technology to these prints and specifications, through information Mexican suppliers: industry best practices, zero- on production techniques and on the technical defect procedures, production audits, and so on specifications of competing products, and were introduced to domestic suppliers, improving through feedback on the design, quality and their productivity. technical performance of their products. With our discussion of international technology transfer in hand, we now discuss policies that have The knowledge transfers involved were multifac- been implemented in several countries with respect eted: not only manufacturing knowledge was trans- to international technology transfer. ferred but also exact sizes, colors, labels, packing materials, and instructions to users. It has also been Government Policies Related to Technology found that in the later 1970s many importing firms Transfer from industrial countries maintained, in Korea and in Taiwan (China), for example, very large staffs who In countries that until recently emphasized import- spent considerable time with local manufacturers substituting industrialization (ISI)--including assisting them in meeting the importers' specifica- most countries in Africa, Latin America, and South- tions (Keesing 1982). Motivated by this evidence, east Asia--imports embodying new products or Pack and Saggi (1999) developed a model that reflecting the cost advantages of new processes were explores the interdependence between production of discouraged by tariffs and quotas. As a result of this manufactures in developing countries and market- protection, local producers were not compelled by ing in industrial countries. In their model, a buyer competition to develop new technology. The from an industrial country can transfer technology unprofitability of exports under ISI regimes implied to producers in a developing country in order to that firms had little interest in inventing new prod- outsource production. Since firms in developing ucts or processes which would allow them to com- countries often lack the ability to successfully market pete on world markets. their products internationally, technology leakage in In contrast, in countries that adopted more out- the developing country market actually benefits the ward-oriented development strategies, technology industrial country firm, since it increases competi- acquisition has been a major concern of govern- tion among developing country suppliers. An inter- ments. Even when domestic markets were protected esting implication of their analysis is that fully in Japan, Korea, and Taiwan (China), penetration of integrated multinational firms may be more averse export markets required the acquisition of knowl- 355 T E C H N O L O G Y A N D I N T E L L E C T U A L P R O P E R T Y edge about new processes and products. The greater As is clear from the above discussion, several suc- export orientation in these economies generated a cessful Asian countries enacted restrictive invest- strong demand for technology that has been satis- ment policies and sought foreign technology fied in several ways. Japan, Korea, and Taiwan all through licensing arrangements and joint ventures. limited FDI while encouraging other modes of tech- What, if any, is the rationale behind such policies? nology transfer, particularly licensing. In fact, it is An immediate explanation is that these policies sim- often overlooked that, regardless of their orienta- ply reflect protectionism: large public firms or hith- tion toward international trade, few countries prac- erto protected private firms fearing competition ticed laissez-faire in the market for technology, and from multinational firms secured protection for governments often discriminated between the dif- themselves. Although that argument certainly car- ferent modes of technology transfer. ries weight, a benign interpretation may also be pos- The experiences of Japan and Korea have been sible. A favorable case for restrictions on FDI can be studied widely in order to draw policy lessons. made by pointing out that the market for technology Ozawa (1974) provides a detailed account of the role is subject to many failures: it must contend with imported technology and local R&D (aimed at facil- asymmetric information (sellers know more about itating absorption of foreign technology) played in the technology than do potential buyers) and imper- Japan's technological development. The Ministry of fect competition (technology is usually sold after International Trade and Investment (MITI) was some invention has been patented). By prohibiting actively involved in Japan's acquisition of foreign FDI and placing other restrictions on the conduct of technology: it limited competition between poten- foreign firms, government policies in many coun- tial Japanese buyers, it restricted FDI until 1970 and tries effectively weakened those firms' bargaining never greatly liberalized it, and it encouraged the dif- position. In Japan MITI actively restricted many fusion of acquired technology. MITI even insisted local firms from participating as potential buyers of that foreign firms share their technology with local foreign technology to improve the bargaining posi- firms as a precondition for doing business in Japan. tion of the others. In contrast to their restrictive policies toward FDI, But what explains the policy preference for joint Japanese government agencies aggressively encour- ventures and technology licensing over FDI? Such aged licensing of foreign technology. policies may reflect a perception that licensing and A similar story can be told about Korea's experi- joint ventures lead to more local involvement and ence with inflows of foreign technology. A central therefore greater technology spillovers to local theme of Korean policy has certainly been the gen- agents. A recent paper by Blomström and Sjöholm eral discouragement of FDI combined with a very (1999), however, casts doubt on this viewpoint. liberal policy toward technology licensing (Moran Using plant-level data for 1991 for Indonesian 1998). Furthermore, Korea encouraged its firms to establishments, the authors find that the degree of export by providing them with subsidized loans foreign ownership did not affect either the produc- conditional on export performance. One possible tivity of firms that received foreign equity or the consequence of such policies may have been the extent of spillovers to the domestic sector. Yet their improvement in productivity that resulted from the results are puzzling: they also find that plants with vertical technology transfers discussed earlier. no foreign investment were less productive than In contrast to Japan and Korea, the Chinese gov- those that received foreign investment. Perhaps the ernment, which has been particularly intervention- results suggest some sort of threshold effect: beyond ist in technology transactions, has encouraged FDI a certain degree of foreign ownership, additional in the form of joint ventures. After maintaining a foreign equity may not affect productivity or closed economy for many years, China opened its spillovers. doors to foreign investment in the late 1970s. It cor- Regarding the mode of technology transfer, sever- rectly views FDI not merely as a supplement to al empirical studies have found that multinationals domestic investment but as a major source of new transfer technologies of new vintage through direct technology and a fully developed marketing capa- investment, preferring to license or transfer their bility.Yet its insistence on joint ventures may restrict older technologies via joint ventures (see Mansfield foreign firms from adopting their most preferred and Romeo 1980; Smarzynska 2000). Such studies mode of entry. cast further doubt on the preference for licensing 356 International Technology Transfer and Economic Development and joint ventures over direct investment. The Despite the subtle policy interventions outlined motivation for transferring older technologies above (many of which occurred in the past), in could be strategic in that multinationals, wishing to recent years government policies across the world prevent the dissipation of their technological have become more liberal toward foreign invest- advantages, are wary of transferring key technolo- ment. Economic reform in many formerly commu- gies to their joint venture partners. Or it could sim- nist countries has added to the list of countries ply be that the transactions costs of transferring vying for FDI. With the growing success of nations new technologies via the market are higher, thus that utilized international trade in goods and ser- motivating internalized transfers by multinational vices and in technology to facilitate their conver- firms. For example, asymmetric information prob- gence toward Western levels of per capita income, lems are likely to be more severe for cutting-edge government policy has become more liberal. Most technologies; buyers in developing countries may countries are now eager to attract FDI, and many have very little information about their value. have concluded bilateral investment treaties (BITs) From a technology owner's perspective, FDI and with important source countries: as of 1999, over arm's length arrangements such as licensing are 1,600 BITs had been negotiated, compared with alternative channels for extracting rents on the basis about 400 at the beginning of 1990 (UNCTAD of technology; each has its advantages and disad- 1997). This trend amounts to an almost complete vantages. There is, of course, no guarantee that what reversal of attitudes in many developing countries is perceived as an advantage by a technology owner that had been strongly averse to permitting invest- is so viewed by developing country firms. In fact, in ments by multinational firms. Many governments many instances the interests of the two parties may have increasingly recognized that multinational be diametrically opposed. For example, licensing firms serve as conduits of superior technology, as contracts often involve many restrictions regarding well as of management techniques. This realization exports to third markets, rights of ownership of stems from the success of countries such as improvements that the licensee may make to the Malaysia, Singapore, and Thailand that rely heavily licensed technology, the purchase of inputs pro- on FDI and from a quest for the sources of their duced by the licensor, and so on. Such restrictions success, which include international technology suggest that licensors attempt to use their bargain- transfer and local investments in infrastructure and ing power to secure favorable licensing contracts. In education that facilitate absorption of technology. this light, policy interventions, as in the case of Coupled with the increasing adoption of bilateral Japan, might be motivated by a desire to shift the treaties is the proliferation of the use of fiscal and terms of licensing contracts in favor of local firms.8 financial incentives in both industrial and develop- Many countries still do not allow free entry of ing countries to lure FDI. Such overly optimistic multinational firms and often express preferences policies carry dangers of their own and may reduce with regard to type of FDI: entry of a soft drink welfare in host countries. A case for such policies giant such as Pepsi or Coca-Cola is viewed differ- can be made on the basis of the positive externalities ently from entry of an automobile manufacturer from FDI, but, as noted earlier, convincing evidence such as General Motors or Ford. Unfortunately, on this front is missing. other than the usual political-economy explana- tions for why certain industries are able to secure Conclusion protection while others fail to do so, there is little in the literature that helps us understand such policies. It has become a staple of development thinking that Although it is possible that spillovers to the local the transfer of technology from industrial to devel- economy are higher when FDI is drawn to particu- oping countries is an important component of sus- lar sectors, there is little empirical evidence to sup- tained growth.Yet the possibility of such technology port this argument. Clearly, the viewpoint that the transfer need not imply that domestic R&D in type of FDI matters is closely related to the idea of developing countries is redundant. Despite the industrial targeting in general, and the pitfalls of usual neoclassical assumption that technology dif- government attempts to correctly identify "high fuses freely, a precondition for successful absorption spillover" industries are all too well known to need of technology by developing countries is that they further discussion here. be involved not only in technology adaptation but 357 T E C H N O L O G Y A N D I N T E L L E C T U A L P R O P E R T Y also in some technology creation of their own.9 In implemented policies such as tax holidays designed the absence of such activities, the efficient absorp- to attract FDI. If the case for such policies rests on tion of new technologies (generated at an ever- positive spillovers from FDI to domestic firms, that increasing pace in the industrial countries) will case may be rather weak. Several studies have failed become progressively more difficult. to find positive spillovers from FDI to firms com- Indigenous R&D cannot take place without an peting directly with subsidiaries of multinationals. adequate stock of human capital. The accumulation These studies require careful interpretation, howev- of human capital through the education of a coun- er, since FDI spillovers may be vertical rather than try's labor force builds the general foundation need- horizontal. ed for successful R&D. In fact, the R&D that needs encouragement during the early stages of develop- Notes ment may be quite incremental in nature. Historical experience shows that to be innovators themselves, 1 Of course, international technology transfer also occurs through channels that do not involve firms or governments: countries must first learn to absorb foreign tech- ideas disseminate by way of scientific and technical literature nologies and adapt them to local conditions. Empir- and communication between researchers. We do not discuss ical evidence demonstrates that even pure imitation these channels of knowledge diffusion, since it is virtually within the same country requires significant invest- impossible to assess their magnitude. ments in R&D (Mansfield, Schwartz, and Wagner 2 For example, empirical microeconomic studies often find low 1981). The magnitude of such investment is surely levels of total factor productivity in developing country firms no smaller in the international context, where tech- even when they employ equipment identical to that in indus- nological diffusion is more difficult. trial countries (Pack 1987). There is some evidence for the view that govern- 3 In a seminal contribution, Teece (1976) demonstrated that on ment policy in countries such as Japan and Korea average, the costs of technology transfer from a home plant to may have played a role in the countries' technological a foreign one, within the same firm, constitute 20 percent of development. Since we do not have the proper coun- the total investment cost of a new plant. These costs, which terfactual experiments, we cannot know how success- may be as high as 60 percent of the total cost of a project, arise from the divergent technological capabilities of the par- ful these countries might have been in the absence of ties exchanging technology, even though they are part of the government intervention. Existing evidence, howev- same firm. er, leaves little doubt that certain fundamentals, such 4 Since these payments only record the explicit sale of technolo- as adequate human capital, have to be in place before gy, they provide no clue about the importance of technology an economy can absorb foreign technologies effec- transfer through FDI in relation to imitation, reverse engineer- tively. In addition, an outward orientation--not nec- ing, and trade in goods. essarily to the extent of the use of export 5 Usually, technology spillovers are viewed as unintentional tech- performance requirements--facilitates absorption of nology transfers. Alternatively, a firm may have the option of foreign technology. By contrast, as evidenced by the curtailing spillovers to other firms but simply may not find it experience of countries that vigorously pursued worthwhile to do so. import substitution, insisting on domestic techno- 6 A self-selection problem may also plague plant-level studies: logical development in isolation has proved to be a the more productive plants may be the ones that attract for- costly and ineffective development strategy. eign investment. While much of the explicit international trade in 7 A recent paper by Djankov and Hoekman (2000) also found technology occurs through FDI, historical experi- negative spillover effects of FDI for domestic firms in Czech ence shows that many countries have preferred industry. licensing and joint ventures. Empirical evidence 8 An explicit recognition of the opposing interests of the two supporting the idea that such modes of technology parties suggests that strategic considerations might play an transfer lead to more learning by local firms is scant important role in determining the outcomes of licensing nego- or completely missing. On the contrary, empirical tiations. Strategic considerations are relevant because multina- evidence has shown that newer technologies are tionals arise mostly in oligopolistic industries where strategic more likely to be transferred via FDI than through decisionmaking is essential. licensing or joint ventures. 9 For an extended discussion of the need for domestic capability Recently, several countries have reversed their to successfully absorb foreign technology, see Pack and Saggi attitude of hostility toward FDI in that they have (1997). 358 35 J A YA S H R E E W ATA L Implementing IPR laws of all WTO members, without exception. Undoubted- ly, however, the more important the TRIPS changes are those in the relevant laws, regulations, and proce- Agreement dures of developing countries, where many sectors of economic and social activity, such as agri- culture, health, education, and culture may be affected. In addi- tion, future ways of doing busi- ness may change in some of these sectors in some developing I countries on account of ntellectual property (IP) can be increased awareness of and evolving attitudes loosely defined as creations of the toward IPRs (Watal 2000a). human mind, and intellectual property rights Following the entry into force of the TRIPS agree- (IPRs) as legal rights governing the use of such cre- ment, new international IP instruments have been ations (see Box 35.1). The Agreement on Trade- found to be necessary to keep up with technological Related Aspects of Intellectual Property Rights, developments. Some of these have been introduced (TRIPS), which came into effect with the establish- by the World Intellectual Property Organization ment of the WTO on January 1, 1995, is the most (WIPO), in particular, in relation to the Internet. In comprehensive international agreement on intellec- recent years, developing country members of the tual property to date. This is not only because of the WTO have proposed several changes, not least to breadth of the subject matter covered but also on bring under the TRIPS agreement the issues of tra- account of its near-universal applicability. The ditional knowledge and genetic resources. TRIPS agreement must be implemented by the 144 current members (as of January 2002) of the WTO Key Provisions of the TRIPS Agreement and will apply to future members. When fully implemented, the agreement will unambiguously The provisions of the TRIPS agreement can be strengthen protection of intellectual property rights broadly divided into five main categories: standards, almost worldwide, a feat not achieved by any single enforcement, dispute settlement, general provisions international treaty up to now. and principles, and transitional arrangements. The TRIPS agreement covers all major IPRs, including some new areas and rights not before Standards addressed by international law or, in some cases, even by national laws of many industrial countries. The TRIPS Agreement sets out the minimum stan- Its implementation will necessitate changes in the dards of protection to be provided by WTO mem- 359 T E C H N O L O G Y A N D I N T E L L E C T U A L P R O P E R T Y B O X 3 5 . 1 A N O V E RV I E W O F I N T E L L E C T U A L P R O P E R T Y R I G H T S At the broadest level, intellectual property has for a period of time, typically the life of the traditionally been divided into industrial creator plus 50 years (in some countries such property--inventions and identifying marks that as the United States and members of the are useful for industry and commerce--and artis- European Union, 70 years). Copyright covers tic and literary property, or works of culture. This only expressions rather than ideas and there- distinction reflected a perception that cultural fore provides thinner protection than patents. creations differed fundamentally from functional Rights extend to the reproduction, display, commercial inventions. The distinction has, how- performance, translation, and adaptation of ever, been considerably blurred in the age of the works. Examples of copyrighted works information technology and digital products. include books, films, music, and computer There are four primary forms of industrial prop- software. The primary limitation on copyright erty rights. protection stems from the fair-use doctrine, which defines the conditions under which 1.A patent awards an inventor the right to pre- copying for limited purposes is permitted. vent others from making, selling, importing, or using the protected invention without The TRIPS agreement requires that computer authorization for a fixed period of time with- programs be protected at least by copyright, on in a country. In return, society requires that the principle that software code is a literary the application be published in sufficient expression. Countries may vary in the degree to detail to reveal how the technology works, which reverse engineering of computer programs thereby increasing the stock of public knowl- is permitted under the fair-use doctrine or other edge. The minimum period of protection limitations. required under the TRIPS agreement is 20 Because computer programs may constitute a years from the date an application is filed. commercially useful process, a number of indus- Many countries recognize utility models or trial countries permit firms to patent them. This petty patents, which award rights of shorter policy is pushing patent protection into new duration to small, incremental innovations areas, including methods of doing business. requiring some investment in design and Another evolution is the tendency toward award- development. ing patents for biotechnological research tools. 2.Rights to industrial designs protect the aes- For some technologies, sui generis, or special, thetic aspects of a functional article. The protection regimes exist. One such case is the TRIPS agreement requires that designs be design of integrated circuits. These are more than protected for a minimum period of 10 years. literary expressions but the inventive step is often 3.Trademarks and service marks protect rights minimal, suggesting a compromise between in a distinctive mark or name used to distin- patent and copyright. Indeed, a 10-year protec- guish a product, service, or firm. Their funda- tion term is provided and requires only originality mental objective is to reduce consumer in terms of being the product of original intellec- search costs and eliminate consumer confu- tual efforts. Another is plant breeders' rights sion over product quality and origin. A relat- (PBRs), which permit developers of new, distinc- ed device is geographical indications, which tive, and genetically stable seed varieties to pre- certify that products such as wines, spirits, vent others from marketing and using these and foodstuffs were made in a particular varieties for a fixed term. Many countries limit place and embody the quality or reputation- these rights through exceptions permitting farm- al characteristics of that location. ers to use seeds for subsequent replanting and 4.Artistic, musical, and literary works are pro- researchers to use the seeds for further breeding. tected by copyright, which grants exclusive Although not literally IPRs, a related area of rights to the particular expression of the work business regulation is the definition of the bound- 360 Implementing the TRIPS Agreement B O X 3 5 . 1 ( C O N T I N U E D ) aries of protection for proprietary trade secrets. A inducing employees to reveal trade secrets, and production process or formula may be kept secret encouraging defection of technical employees to within the firm, but if a competitor learns the produce their own versions of a product based on confidential information through legitimate proprietary information. Definitions of unfair reverse engineering, the originator has no rights competition vary considerably across countries. to exclude its use. Unfair competition includes such activities as carrying on industrial espionage, Source: World Bank (2001a). bers in each of the main areas of intellectual proper- ically available in industrial countries than those in ty. The IP areas covered are patents and the protec- developing countries. Important limitations on the tion of plant varieties; copyright and related rights scope of IPRs, such as "fair use," "compulsory li- (that is, the rights of performers, producers of sound censes," "government use," and other limited excep- recordings, and broadcasting organizations); undis- tions that are widely available under national laws, closed information (trade secrets and test data); are also included in the TRIPS agreement. trademarks; geographical indications; industrial designs; and the layout designs of integrated circuits. Enforcement With the notable exception of the moral rights of authors under copyright, the substantive provisions The second main category of provisions deals with of the main international IP conventions of the domestic procedures and remedies for the enforce- WIPO (the Paris and Berne Conventions) are incor- ment of IPRs. For the first time in international IP porated by reference into the TRIPS agreement and law, detailed provisions on civil and administrative have to be complied with. Also mentioned are the procedures and remedies, provisional measures, Rome Convention and the so-called Washington special requirements related to border measures, Treaty on Integrated Circuits. There is no mention and criminal procedures are laid out. These provi- in the TRIPS agreement of the international treaty sions specify the minimum procedures and reme- for the protection of plant varieties, known by its dies that must be available so that rightsholders can French acronym, UPOV. effectively enforce their private rights in domestic The main elements of protection for each IPR are judicial, quasi-judicial, or administrative institu- generally defined in terms of the subject matter that tions, in accordance with certain general principles. is to be protected or that can be excluded; the pre- The TRIPS agreement requires that procedures for conditions for such protection; the rights accruing the enforcement of IPRs be effective and constitute on protection and the permissible exceptions to a deterrent to further infringement, but it creates no those rights; and the minimum duration of protec- obligation to distinguish the enforcement of IPRs tion. Generally, IPRs give creators exclusive rights from the enforcement of law in general--say, by over the use of their creations for a fixed duration of instituting a separate judicial system or through any time. In some cases, however, IPRs are valid indefi- redistribution of resources. nitely, as long as the conditions for their protection continue to be met, as is true for trademarks, geo- Dispute Settlement graphical indications, and trade secrets. In the case of plant variety protection, the TRIPS agreement The TRIPS agreement makes disputes between only obliges, at the minimum, an "effective" sui WTO members about compliance with the agree- generis (special) regime. Test data submitted for reg- ment's obligations subject to the WTO's dispute set- ulatory approval of new pharmaceutical or agricul- tlement procedures. This feature distinguishes the tural chemical products need to be protected against TRIPS agreement from previous international IP "unfair commercial use."Overall, these standards are law and may, indeed, have been one of the reasons generally closer to the pre-TRIPS IPR standards typ- for bringing the subject of intellectual property into 361 T E C H N O L O G Y A N D I N T E L L E C T U A L P R O P E R T Y the Uruguay Round. It is important to note that for extended only for LDCs. Special transition rules a period of five years, up to 2000, nonviolation-type apply where a developing country does not provide, complaints could not have been brought under the as of 1995, product patent protection for a given TRIPS agreement. At the Fourth Ministerial Con- area of technology--for example, pharmaceuticals, ference of the WTO, held in Doha in November chemicals, or certain biotechnological inventions. 2001, it was agreed that this period would be pro- In that case, the country may postpone the formal longed so that the issue of the scope and modalities introduction of such protection to January 2005. of such complaints could be examined further in However, subject to certain conditions, exclusive the TRIPS Council and a report made to the next marketing rights (EMRs) for eligible pharmaceuti- ministerial conference. cal and agricultural chemical product inventions must be made available by all such WTO members, with effect from January 1, 1995, for a period of five General Provisions and Principles years from the date of the marketing approval, or The TRIPS agreement provides for certain basic less if the patent decision is made earlier. principles, such as nondiscrimination between domestic and foreign IPR holders (national treat- Legislative Options for Implementation of ment) and among foreign IPR holders (most- the TRIPS Agreement in Developing favored-nation, or MFN, treatment), and it contains Countries some general rules on acquiring and maintaining IPRs. It also permits members to adopt certain Contrary to a widespread perception, by the time measures to meet specified policy objectives, the TRIPS agreement entered into force, many including protection of public health and nutrition, developing countries already had IPR laws and pro- provided that the measures are consistent with the cedures that met a number of their TRIPS obliga- agreement's provisions. As a minimum standards tions. Only in the areas of layout designs of agreement, the TRIPS agreement allows members integrated circuits, plant variety protection, test to provide more extensive protection of intellectual data, and, perhaps, the protection of geographical property if they so wish. Members are also free to indications for wines and spirits did the agreement determine the appropriate method of implement- necessitate entirely new laws, in many developing ing the provisions of the agreement within their countries and in some industrial countries. Signifi- own legal system and practice. Thus, the TRIPS cant changes in the existing laws of some develop- agreement does not call for global harmonization of ing countries were required with respect to patents IPR laws. More specifically, it leaves countries free to and copyright and related rights. In general, rela- determine their own national "parallel import" tively marginal adjustments were required in laws policies with regard to the import of goods that are and procedures in other areas, including enforce- put on the market legitimately by rightsholders in ment (other than at the border, in some cases). another market. Many developing countries, particularly in East Asia and Latin America, implemented the TRIPS agree- ment, in full or in large part, before the transition Transitional Arrangements periods expired. The number of WTO members The obligations under the TRIPS agreement apply that were not already providing patent protection to equally to all members, but developing countries pharmaceutical products at the entry into force of have a longer period to phase them in. Industrial the agreement was less than 20, and of these, only a country members had to comply with all of the pro- few--notably, Egypt, India and Pakistan--are wait- visions of the agreement as of January 1, 1996. For ing until 2005 to do so. developing countries and transition economies the The TRIPS agreement leaves some degree of leg- corresponding date of compliance was generally islative leeway in implementation. Apart from January 1, 2000, and for least-developed countries allowing WTO members to determine the best way (LDCs) it was January 1, 2006. (At the Doha minis- of implementing the agreement within their own terial conference, the date for LDCs was extended to legal systems and practice, some provisions give 2016 with respect to pharmaceutical products.) countries room to deliberately select a legislative Under the agreement, the transition period can be option. For instance, the obligation on textile 362 Implementing the TRIPS Agreement designs in the TRIPS agreement can be met through patents. The grounds in industrial country laws, industrial design law or through copyright law. In even after TRIPS implementation, range from"pub- addition, the text of the TRIPS agreement is dotted lic interest" in Germany to "demand for that prod- with "may" provisions that are clearly optional. uct not being met on reasonable terms" in the Finally, many terms used in the agreement are not United Kingdom and similar provisions in Australia defined: "inventions," "new," "inventive step," and New Zealand. Developing countries have tried "microorganisms," "essentially biological," "effec- to retain even greater flexibility in this area. Some tive," "unreasonably," and "legitimate," to mention developing countries, notably Brazil, have included just a few. These must be interpreted at the national the ground of "nonworking" of the patent locally. level (Watal 2000a). Discrimination in the enjoyment of patent rights From this plethora of legislative options, it is pos- between locally produced and imported products is, sible to isolate the most important ones that are to however, prohibited under the TRIPS agreement, be exercised by developing countries: rendering "local nonworking" controversial as a ground for compulsory licenses.1 As noted below, · Standards to be set on the criteria of patentability, the Doha ministerial conference clarified several including how far to go on biotechnological TRIPS provisions that can be used for public health inventions in general and plant inventions in par- purposes, including compulsory licenses. In line ticular with the Berne Convention, exceptions to copyright · The grounds on which to allow compulsory are defined broadly under the TRIPS agreement. licenses or government use of patents National laws on copyright and related rights allow · Whether to allow parallel trade compulsory licenses in certain situations, contain · How to implement the provisions regarding test other specific exceptions, and even permit free use data. as a part of the concept of fair use. Dispute settle- ment bodies may further define the limits to these The TRIPS agreement calls for respecting the exceptions, as recent panel reports on the EU-Cana- three universally recognized criteria for patentabili- da patent dispute concerning the "Bolar" provision ty: novelty, inventive step or nonobviousness, and (WT/DS/114/R) and on the EU-U.S. copyright dis- industrial applicability or utility. These, and other pute (WT/DS/160/R) have done.2 In both cases issues relating to scope of patents, are not defined there were two disputed measures, and in both deci- further under the agreement, and historically, dif- sions one measure was ruled to be compatible with ferent countries have followed different standards. the TRIPS agreement and the other was found to be Some countries, for example, grant patents for gene inconsistent. sequences with no known utility or for trivial inven- National policies on parallel trade vary widely. tions that may be excluded elsewhere. Many other The United States generally gives the rightsholder countries exclude patents on scientific principles, the right to prevent parallel imports of patented or computer software per se, and methods of doing copyrighted products but is more open to such business. Throughout the 1990s, however, standards imports under trademarks. Japan permits the right- in the United States, the European Union (EU), sholder in the exporting jurisdiction to take steps to Japan, and Australia have converged to a large prohibit such parallel sales--for example, by requir- extent, particularly in the area of biotechnological ing "implied consent" through the use of labels or by inventions. Indeed, some developing countries such other means. European countries generally give the as the Republic of Korea and Singapore have adopt- rightsholder the right to prevent parallel imports ed similar standards in this sector. Some developing from within the EU but prohibit such imports from countries, including Brazil and Argentina, have outside, at least in the area of trademarks. New specifically excluded genes and computer programs Zealand allows parallel imports in all copyrighted from patent protection. On plant variety protection, products, while Australia allows them in selected many developing countries have opted to follow products such as books, semiconductor chips, and UPOV 1978, either fully or in conjunction with sound recordings. It is difficult to generalize about some provisions of UPOV 1991. developing countries' policies. Brazil limits the right The TRIPS agreement does not restrict the of the rightsholder to prevent parallel imports under grounds for the grant of compulsory licenses for its industrial property laws, but Argentina clearly 363 T E C H N O L O G Y A N D I N T E L L E C T U A L P R O P E R T Y extends such a right. South Africa recently passed nologies predominantly use lead time, trade secrets, legislation to allow parallel imports of medicines, at learning advantages, and sales and service differ- the discretion of the health minister, and India's ences as the primary means of appropriability draft patent law would also allow parallel imports. (Levin and others 1987). A broader scope of protec- Test data is an area that was difficult to negotiate tion does not necessarily lead to greater amounts of and that has been subject to widely varying inter- socially beneficial innovation. It has been argued pretation. There are two separate obligations: to that patent protection can bring with it practices keep such data secret, and to protect the data against that may lead to socially wasteful patent races and unfair commercial use. Some assert that the second duplicative research or to blocking patents. This obligation calls for market exclusivity for a certain may obstruct further innovation and thus hinder period--say, five years--for the developer of test rather than stimulate technological and economic data for new pharmaceutical and agricultural chem- progress (Merges and Nelson 1990; Mazzoleni and ical products when test data were submitted after Nelson 1998). Concerns have been expressed considerable effort (Kirk 1997). Others claim that regarding cumulative research and development this provision allows for the use of the test data by (Scotchmer 1991) in general and "platform" or authorities to approve subsequent equivalent prod- enabling technologies (such as gene sequences and ucts (Correa and Yusuf 1998). If the U.S.-Argentina telecommunication standards) in particular. The dispute of May 2000 (WT/D/22), which is at the debate has thus revolved around how to achieve a stage of bilateral consultations, is eventually balance between rewarding current efforts at inno- resolved by a panel or Appellate Body, the outcome vation and encouraging further innovation. Com- should throw more light on this issue. petition law that safeguards against anticompetitive practices has an important role in this regard. There is a substantial and growing body of litera- Economic Implications of IPRs ture on the economic impact of stronger IPR pro- IPRs can be broadly classified in two categories tection in developing countries, as measured according to their economic function. (a) Some through empirical studies (Maskus 2000a). Unfor- types of IPRs, notably patents and copyright, serve tunately, few studies have yielded clear-cut results. to bridge the gap between the social value and the Nevertheless there are indicators that stronger private value of innovation. (b) Others, such as patent regimes: trademarks and geographical indications, merely distinguish the origin and quality of goods and ser- · Could lead to increased global trade--specifically, vices. Economists have been more fascinated by the exports from OECD countries to developing first category. countries (Maskus and Penubarti 1995). If private innovators cannot appropriate the · Could attract more foreign direct investment returns to their innovation, they will not produce (FDI) for host countries, particularly in sectors the socially optimal level of innovative activity. But such as pharmaceuticals (Mansfield 1994). the social value of innovation lies in its widespread · Could lead to increased licensing of technologies diffusion and there is thus a tradeoff between the to, and possibly more local production through, incentives for creativity and innovation and those FDI in developing countries (Maskus 1998). for diffusion (Besen and Raskind 1991). Intellectual · May lead to pharmaceutical research and devel- property, however, is not the only means of appro- opment (R&D) more appropriate to the needs of priating the returns from innovation. Several stud- developing countries--although it is too early to ies have shown that IPRs are particularly important tell (Lanjouw and Cockburn 2000). for generating and protecting creativity and innova- · Contribute to higher growth rates. Stronger tion where considerable expenditure of time and patents in open economies have been shown to resources is required to generate products or raise growth rates by 0.66 percent, on average processes that, once produced, are quick, cheap, and (Gould and Gruben 1996). The strength of IPRs easy to imitate. (Pharmaceutical products, comput- may decline as incomes increase from very low er programs, music recordings, and films are exam- levels and then rise at the highest income levels. ples.) Other industries that are less susceptible to Trade openness and market freedom may affect imitation or are subject to rapidly changing tech- this trend positively (Maskus 2000a). 364 Implementing the TRIPS Agreement On the other hand, there are studies showing that: domestic creativity and innovation in developing countries. There is mixed descriptive evidence on · Stronger IPRs may not significantly reduce the the U.S. Patent and Trademark Office Website con- North-South technology gap, although utility cerning the increased filing of patents by developing models could be an important source of technical countries in the United States in recent years, sug- change and information diffusion (Park, 2000). gesting that greater local inventive activity in some · In India prices of patentable medicines could, countries is coinciding with changes in patent laws depending on assumptions, rise by as much as or strategies. As far as major diseases are concerned, 250 percent. Compulsory licenses could signifi- a primary challenge is to develop mechanisms for cantly reduce prices, although not to prepatent encouraging the required R&D (see Box 35.2). levels. Price controls, applying India's present pol- icy, could reduce patented drug prices by only Dispute Settlement to Date about 40 percent (Watal 2000b). · Product differentiation and the availability of The TRIPS Agreement was the outcome of difficult substitute medicines in the therapeutic category North-South and intra-North negotiations, reflect- do play an important role in reducing such price ing strong economic interests both of the rightsown- effects (Fink 2000). In India, even in the absence ers and of those benefiting from weaker levels of of patents, new drug markets are fairly concen- protection for IPRs. This conflict of interest was trated, and market share has been gained through partly resolved through "constructive ambiguity," brand loyalty established by trademarks (Watal with each side interpreting the agreement according 1995). to its own convenience. Interpretation of ambiguous clauses in certain ways in national laws may be one More empirical analysis is needed on the eco- means of asserting victory in past negotiating bat- nomic implications of the policy options available tles--the more so as there are no official records of to developing countries under the TRIPS agree- the negotiations and there are areas where the text is ment, such as compulsory licensing or parallel unclear and liable to differing interpretations. trade, to supplement recent work (Scherer and North-South disputes on the TRIPS agreement have Watal 2001). Some observers have long felt that only begun to surface in the WTO in many of these compulsory licenses at reasonable royalties do not controversial areas, largely because developing coun- necessarily impede technological progress or lower tries had up to the end of 1999 to implement most of the rate of innovation (Scherer 1977). Yet there the provisions of the Agreement. Nevertheless, some remains opposition, particularly by the research- North-South disputes already arose in the transition based pharmaceutical industry, to the use of this period. Some important ambiguities were clarified policy instrument (see ), as, for in the Doha Ministerial Declaration on the TRIPS example, in the case of HIV/AIDS medicines in agreement and public health, which clearly asserts South Africa and Thailand (see ; see members' right to use to the fullest extent the flexi- also Box 36.1 in this volume). The economic case bility available in the TRIPS agreement. for parallel trade is less clear, and the interests of Up to January 2002, WTO members had invoked various groups of developing countries are likely to the dispute settlement procedures 24 times, in 20 differ significantly. distinct cases. In the overwhelming majority of the There are few studies that document how changes complaints (16 of the 24), the United States was the induced specifically by the TRIPS agreement in complainant. The EC was the complainant in 6 developing countries would affect future exports, cases, Canada in 1, and Brazil in 1. Up to 1999, there whether of generic products after patent expiry or were 5 complaints against developing countries and of domestically owned IP products. Nor are existing in May 2000, 2 more were added. Of the 24 com- studies on technology transfer able to distinguish plaints, 9 (including 2 against a developing country) clearly between increases in licensing payments were settled by mutual agreement, 7 were decided stemming from higher costs and those resulting by panels (3 of them at the appellate level), and the quantitative or qualitative increases. More impor- remaining 8 are pending. Three of the 8 pending tant, there has been no empirical evidence on the complaints are more than three years old. Of the 20 relationship between stronger IPRs and the level of distinct cases, 10 relate primarily to patents or 365 T E C H N O L O G Y A N D I N T E L L E C T U A L P R O P E R T Y B O X 3 5 . 2 C O M B AT I N G D I S E A S E W O R L D W I D E : F O S T E R I N G T H E R E Q U I R E D R & D A critical task facing the global economy is to recognize or enforce patent protection for phar- develop mechanisms that encourage research maceutical products. Regarding the latter prob- aimed at developing treatments for diseases lem, the TRIPS agreement requires that which are common in poor countries and, at the developing WTO member countries provide same time, achieve widespread distribution of patents for new pharmaceutical products by those treatments at affordable prices. The issue 2005 at the latest (by 2016, for least-developed has become prominent because of the severe epi- countries). There is concern, however, that the demic of HIV/AIDS, in particular in Sub-Saharan provision of product patents in pharmaceutical Africa, South Asia, and Southeast Asia. But products could confer considerably greater mar- HIV/AIDS is not the only disease plaguing poor ket power on rightsholders by delaying the entry nations; malaria, tuberculosis, and other maladies of generic competitors for new products. Then are equally debilitating. In fact, HIV/AIDS is such firms might reduce sales or output in partic- unusual in that it affects both rich and poor coun- ular markets, supporting higher monopolistic tries. Pharmaceutical firms therefore have incen- prices in key medical therapies. tives to develop HIV/AIDS medicines for sufferers Considerable pressure has been exerted on in high-income economies, and what is being pharmaceutical companies to provide drugs to debated is how to transfer these medicines to poor countries at marginal production cost (or poor countries. In contrast, malaria and other dis- less). For example, Merck & Co. recently eases endemic to impoverished nations are "neg- announced that it would cut the prices of two lected" in that they attract little research and AIDS-controlling drugs in Africa by 40 to 55 per- development (R&D). For example, the World cent, adding to sharp price cuts announced a Health Organization (WHO 1996) has estimated year earlier. Abbott Laboratories offered to sell its that of the US$56 billion spent globally on med- two AIDS drugs, Norvir and Kaletra, at prices that ical R&D in 1994, less than 0.2 percent was spent would earn the company no profit. Many other on tuberculosis, diarrheal maladies, and pneumo- firms, including the Bristol-Myers Squibb Co. and nia, and virtually all of this research was carried GlaxoSmithKline PLC, have announced similar out by public agencies and military authorities. price cuts. These research-intensive firms have R&D on antimalarial vaccines and drugs is mea- three concerns about low-cost distribution pro- ger. Some research is going on under the aus- grams. First, provision at marginal cost adds pices of the Multilateral Initiative on Malaria, nothing to their ability to cover the costs of R&D. involving the United Nations Development Pro- Second, while they may be willing to supply their gramme (UNDP), the World Bank, and WHO, medicines cheaply, they wish to retain the exclu- and by the Medicines for Malaria Venture, a pub- sive distribution rights inherent in patents. lic­private sector cooperative initiative. Funding Indeed, this preference underlay the recent law- for the Multilateral Initiative comes to perhaps suit by several firms against the South African US$3 million per year, and Medicines for Malaria government, challenging the constitutionality of is soliciting support from foundations in the hope its 1997 Medicine and Related Substances Con- of raising US$30 million per year. These amounts trol Act. Third, drug manufacturers are con- are inadequate for the job, given the costs of cerned that the availability of far cheaper developing and testing new drugs. medicines in poor countries could erode their There are two main reasons for this low rate of ability to sustain higher prices in rich countries. R&D. Most important, the low purchasing power Under Article 68 of Brazil's Industrial Property in poor companies gives pharmaceutical compa- Law (Law 9.279/96), foreign firms must manu- nies insufficient incentives to introduce new facture patented drugs within Brazil before three drugs into those markets. A second reason is that years have elapsed from the grant of the patent. in the past many developing countries did not Failure to meet these "working requirements" 366 Implementing the TRIPS Agreement B O X 3 5 . 2 ( C O N T I N U E D ) could result in an order by the Brazilian authori- tribution of the drugs to designated countries at ties to local firms to manufacture generic substi- low cost, while preventing backflow of cheap tutes under compulsory license--a threat that medicines to higher-income nations. If such recently faced the makers of the AIDS drugs negotiations are unfeasible or ineffective, it may Efavirenz (Merck & Co.) and Nelfinavir (Roche). be advisable to establish a system of royalties This issue was raised by the United States at the under which countries could acquire licenses to WTO, but a bilateral settlement was arrived at, produce and distribute the drugs. For this system and the case was withdrawn. to be effective, small countries without produc- In economic terms, to address effectively the tion facilities may need to be given the right to diseases endemic to poor countries through import drugs from generic producers in third development of and access to new treatments countries. requires separation of the dynamic incentives for Ganslandt, Maskus, and Wong (2001) estimate R&D from the need for widespread distribution the annual cost of such an international strategy at low cost. Because paying for the required at between US$8.2 billion and US$12.1 billion. R&D is beyond the means of poor countries, any While this commitment would represent a sub- comprehensive solution to the problem requires stantial portion of current aid funding (which significant increases in assistance from industrial amounted to US$84.9 billion in 1999), it would countries and financial support from multilateral correspond to only 0.03 to 0.05 percent of the organizations and private donors. These monies OECD's 1998 GDP. Indeed, if the US$12.1 billion would be used for two purposes. An immediate were paid by the United States, the European task would be to build effective health care deliv- Union, and Japan it would come to only ery systems in poor countries, where health US$13.50 per person per year. For a final per- infrastructures are weak. The second task would spective, the US$12.1 billion may be compared be to provide incentives for firms to engage in with the anticipated loss in South African GDP, if R&D in new and effective vaccines and medi- the current epidemic continues unchecked, of cines. Most likely, these incentives would involve US$22 billion in 2010. purchase by governments or international public agencies of bulk amounts of targeted drugs from Source: Prepared by the volume editors, based on Gans- manufacturers at negotiated prices and the dis- landt, Maskus, and Wong (2001). EMRs (5 of these concern pharmaceuticals), 3 to Proposals and Prospects for Reform of the copyright, 3 to trademarks, and 4 predominantly to TRIPS Agreement enforcement. (See "Update of WTO Dispute Settle- ment Cases," available at .) In the preparations for the WTO ministerial meet- Some have suggested that retaliation by with- ings at Seattle in 1999 and Doha in 2001, there were drawal of concessions under the TRIPS agreement more proposals for reform of the TRIPS agreement could be more effective and beneficial than conven- by developing countries than by the original deman- tional trade retaliation (Subramanian and Watal deurs, certain industrial countries. This is a reflec- 2000). Indeed, Ecuador's request for such retaliation tion of the perception by many developing countries against the EU in the Bananas dispute was recently that, in both mercantilist and real terms, they were granted by the Dispute Settlement Body after WTO losers from this agreement. Thus, some developing arbitration (see WT/DS27/ARB/ECU, available at countries, backed by some important nongovern- ). mental organizations in industrial countries, have 367 T E C H N O L O G Y A N D I N T E L L E C T U A L P R O P E R T Y sought clarifications regarding their ability to take is free to determine the grounds on which compul- measures to protect public health and on the mean- sory licenses are granted and to establish its own ing and interpretation of specific provisions. Some regime on parallel imports. It extended the transi- of these countries have also sought extension of tion period for LDCs with respect to pharmaceuti- transition periods that expire by the end of 1999 or cal products by 10 years, to 2016. In addition, the 2005, depending on the category, and exclusions TRIPS Council was instructed to find an expedi- from the grant of patents for microorganisms and tious solution to the problems of countries with other biotechnological inventions, already conceded limited manufacturing capacities in the pharma- in the Uruguay Round. In addition, many have sup- ceutical sector in making effective use of compulso- ported the inclusion of subjects such as the protec- ry licensing. tion of genetic resources, traditional knowledge, and The Doha Ministerial Declaration that launched folklore as part of the review of Article 27.3(b) that the Doha Development Agenda includes a mandate began in 1998 and is currently under way in the for completing the negotiations on the system of TRIPS Council. The African group, led by Kenya and multilateral notification and registration of geo- Zimbabwe, was, for the first time, particularly active graphical indications for wines and spirits by the in this respect. Some developing countries wanted to next ministerial conference, to be held in 2003. It focus on the problems of technology transfer calls for further work on ongoing issues in the through a separate working group. Others formed TRIPS Council concerning the extension of addi- North-South alliances to call for increased protec- tional protection under geographical indications to tion of intellectual property by extending the higher products other than wines and spirits; review of the level of protection under geographical indications to provisions on biotechnological inventions; the rela- products other than wines and spirits. tionship between the Convention on Biological There were no industrial country proposals for Diversity and the TRIPS agreement; and traditional reform of the TRIPS agreement other than the Euro- knowledge and folklore. Other new develop- pean proposal for setting a deadline for the comple- ments--for instance, regarding the new WIPO tion of negotiations on the protection of copyright treaties--can be raised by members geographical indications for wines and spirits under the review of the TRIPS agreement. The through a multilateral notification and registration mandate on ongoing work overlaps with another system. However, in the course of discussions in the mandate, on outstanding implementation issues. TRIPS Council on biotechnological inventions, some Separately, ministers reaffirmed the mandatory pro- industrial countries expressed a desire to strengthen visions on the transfer of technology to LDCs and the provisions by deleting the current exclusions. In agreed that the TRIPS Council should ensure the past discussions industrial countries suggested the monitoring and full implementation of the obliga- inclusion under TRIPS of the copyright treaties con- tions in question. cluded in the WIPO at the end of 1996. The Doha ministerial conference led to the Notes acceptance of many of the demands of developing This chapter was written while the author was visiting at the Insti- countries. A separate declaration on the TRIPS tute for International Economics in Washington, D.C. agreement and public health emphasized that the agreement does not and should not prevent WTO 1 The U.S.-Brazil dispute on this issue was withdrawn in 2001 (see WT/DS/99/4, available at ). members from taking measures to protect public health. The declaration clarified that each member 2 The reports are available at . 368 36 K E I T H E . M A S K U S Benefiting from Technology Transfer Intellectual property rights can Intellectual significantly encourage the acquisition and dissemination of Property technical information. There are three channels through which Protection technology is transferred across borders: international trade in goods, foreign direct investment (FDI), and licensing of tech- nologies and trademarks to unaffiliated firms, subsidiaries, G and joint ventures. Economic overnments in developing nations theory indicates that transfers through each channel justify the adoption of stronger depend in part on local protection of IPRs, in com- intellectual property rights (IPRs) by claiming that plex and subtle ways (see Maskus 1998). such reform will result in more inward technology That imports serve to transfer technology is wide- transfer, more local innovation and cultural devel- ly accepted by economists. Imports of capital goods opment, and a faster route to closing the technology and technical inputs directly reduce production gap between themselves and rich countries. costs and raise productivity in the firms that employ Enhanced IPRs by themselves, however, are unlikely them. Evidence shows that international trade in to produce such effects. Expectations that stronger high-technology goods depends positively on the IPRs alone will bring about technical change and strength of patent regimes in large developing growth are likely to be frustrated. countries (Maskus 2000a). Moreover, one study According to the available evidence, claims that (Coe, Helpman, and Hoffmaister 1997) found that IPRs generate greater international economic activ- increases in the share in GDP of imports of machin- ity and domestic innovation are conditional. The ery and equipment from OECD countries tend to positive impacts of IPRs are stronger in countries raise total factor productivity (TFP) in developing with appropriate complementary endowments and countries. In this context, stronger patent rights in policies. Thus, the challenge of ensuring that new industrializing countries could raise long-run pro- IPR regimes become a positive tool for promoting ductivity growth significantly. beneficial technical change and development is It is clear from empirical work that the strength of multifaceted. This chapter considers the role of IPRs and the ability to enforce contracts have intellectual property protection in promoting tech- important effects on decisions by multinational nical change and discusses important supporting firms on where to invest and whether to transfer policies. advanced technologies through FDI (Maskus 369 T E C H N O L O G Y A N D I N T E L L E C T U A L P R O P E R T Y 2000b). FDI often embodies efficiency advantages and innovation were insufficiently developed to through superior technologies, management skills, warrant protection. But inadequate IPRs can stifle and marketing. The subsequent diffusion of this technical change even at low levels of economic knowledge into the broader economy is a complex development. This is because much innovation is process. Intellectual property rights could enhance aimed at local markets and may suffer infringement that diffusion by ensuring greater contract certainty that capitalizes on local familiarity. These invest- between enterprises and suppliers and by providing ments are costly and may only be made when risks more protection for commercializing technologies of unfair competition and trademark infringement in local markets. Furthermore, enterprises would are small. Adequate and enforceable IPRs also help experience stronger incentives to train managerial reward creativity and risk-taking by new enterprises and technical workers because workers would feel and entrepreneurs. more constrained against misappropriating trade In the overwhelming majority of cases, invention secrets. At the same time, IPRs raise imitation costs, in developing nations involves minor adaptations of thereby limiting diffusion of technologies, at least existing technologies, but the cumulative effect of temporarily. Learning by honest means could be these small inventions can be critical for growth in slowed if the system raised the costs of inventing knowledge and activity. Moreover, to absorb knowl- around patents. Finally, legal restrictions on defec- edge and know-how in advanced technologies tion of skilled workers would engender conflict requires considerable investment in such factors as between the objectives of training and diffusion. process control and product quality maintenance. That IPRs could play a positive role is suggested These investments tend to have high social returns by Park and Ginarte (1997), who focus on the rela- in developing economies because they are crucial tionship among patents, investment in capital and for raising productivity toward global norms (Even- in research and development (R&D), and growth. son and Westphal 1997). They found no direct correlation between patent An example of this process is that protection for strength and growth, but there was a strong and utility models, which are patents of short duration positive impact of patents on physical investment awarded to small, incremental inventions, has been and on R&D spending, which in turn raised growth shown to improve productivity in technology-fol- performance. Thus, it seems clear that IPRs and FDI lower countries. In Brazil utility models were work jointly to raise productivity and growth. important in permitting domestic producers to gain Turning to licensing, survey results point to the a significant share of the farm machinery market by importance of IPRs in persuading enterprise man- adapting foreign technologies to local conditions. agers to transfer their most advanced technologies Utility models in the Philippines encouraged suc- (Mansfield 1995). There is practical evidence from cessful adaptive invention of rice threshers. China to support these arguments (Maskus, In perhaps the most systematic study, Maskus and Dougherty, and Mertha 1998). When interviewed, McDaniel (1999) considered how the Japanese managers of many foreign enterprises expressed patent system affected postwar Japanese technical great reluctance to locate R&D facilities in China, progress, as measured by increases in TFP. The sys- citing fear of misappropriation and patent infringe- tem was designed to encourage incremental and ment. Nearly all reported that their enterprises adaptive innovation and diffusion of knowledge. transfer technologies that are at least five years Provisions included early disclosure of, and opposi- behind global standards or bring in technologies tion proceedings to, patent applications; an exten- that will be obsolete within a few years. Foreign sive system of utility models; and narrow claim enterprises are also reluctant to license advanced requirements in patent applications. The authors technologies to unrelated enterprises. found that this system promoted the development of large numbers of utility model applications for incremental inventions, which were based in part Domestic Innovation on laid-open prior applications for invention Traditionally, developing countries have established patents. In turn, utility models had a strongly posi- IPR systems that favor information diffusion tive impact on real TFP growth over the period. through low-cost imitation of foreign products and Innovation through product development and technologies, in the belief that domestic invention entry of new firms seems to be stifled by weak 370 Benefiting from Intellectual Property Protection trademark protection in poor nations. A recent sur- Building Markets and Improving Quality vey of trademark use in Lebanon provided evidence on this point (Maskus 2000c). Firms in the apparel Intellectual property rights not only promote R&D industry wish to design clothing of high quality and and product innovation; they also encourage the style aimed at Middle Eastern markets. Attempts to development of interregional and international dis- do so have been frustrated by trademark infringe- tribution and marketing networks that are impor- ment in Lebanon and in neighboring countries. In tant for achieving firm-level scale economies. Weak the food products sector, legitimate firms suffer IPRs limit incentives for such investments because from considerable misappropriation of their trade- rightsowners cannot prevent their marketing out- marks. Similar difficulties plague innovative pro- lets from debasing the quality of their products, nor ducers in the cosmetics, pharmaceuticals, and metal can they readily deter counterfeiting of their trade- products sectors. The essential point is that local marks. IPRs permit effective monitoring and product development and entry of new firms may enforcement of activities throughout the supply and be restrained by trademark infringement targeted distribution chains, giving both innovators and dis- largely at domestic enterprises. tributors an incentive to invest in marketing, ser- Similar problems exist in China (Maskus, vices, and quality guarantees. Dougherty, and Mertha 1998). According to anec- Quality assurance is critical for safeguarding the dotal information, trademark infringement signifi- interests of consumers. Widespread sale of counter- cantly and negatively affects innovative Chinese feit products can ruin reputations achieved at con- enterprises. Many examples were cited in interviews siderable cost, especially for new enterprises, and of the difficulties facing Chinese producers of their the problem can be overcome only at additional own brands of consumer goods, such as soft drinks, cost. In principle, effective trademark enforcement processed foods, and clothing. The establishment of both raises the average quality of products over time brand recognition in China requires costly invest- and provides a wider range of qualities from which ments in marketing and distribution channels. consumers may choose.1 This process is particularly Enterprises that achieve this recognition find their important in food products, beverages, cosmetics, trademarks applied to counterfeit products that are and medicines, where counterfeit products can be of lower quality and damage the reputation of the hazardous. Indeed, field research in China suggests legitimate enterprise. This situation probably has an that despite the advantages to poor consumers of important deterrent effect on enterprise develop- having access to low-cost product knockoffs and ment in China and effectively prevents interregional unauthorized copies of entertainment products, marketing, which would permit the attainment of these consumers are becoming resentful that market economies of scale. saturation by unauthorized goods diminishes the Copyright industries such as publishing, enter- range of legitimate goods available (Maskus, tainment, and software are likely to be dominated Dougherty, and Mertha 1998). by foreign enterprises (which can absorb temporary In a related vein, inadequate copyrights cannot sup- losses and can afford to deter infringers) and by port the complex contracts that allocate rights in mod- pirate firms. Thus, lower-quality copies are widely ern creative industries. Poor copyright enforcement is and cheaply available, but the economy's domestic thought to be a critical factor in the inability to create cultural and technological development is slowed. music industries in Sub-Saharan Africa, despite the For example, Lebanon has a small film and televi- abundance of musical talent, for it retards the estab- sion industry that believes it could successfully lishment of collection societies and recording facilities. export to neighboring economies if they had In contrast, India has long had a system of effective stronger copyright protection. In China the domes- copyright protection, which is thought by many tic software industry has grown rapidly in the area observers to have been important in developing and of particular business applications that do not suf- protecting its successful film and software industries. fer much from copying but has faced obstacles in developing program platforms. In short, domestic Complementary Policies commercial interests in stronger copyrights have emerged and are playing a role in promoting As suggested earlier, the benefits just listed are enforcement. unlikely to emerge to a significant degree unless 371 T E C H N O L O G Y A N D I N T E L L E C T U A L P R O P E R T Y other market and policy conditions complement be reduced significantly over time by trademark and the intellectual property system. Thus, policymak- copyright enforcement. The task of reallocating ers need to take a broad view of how to promote people currently engaged in such activity toward innovation, learning, and dynamic competition. legitimate business will be easier, the more flexible is The following collateral policy approaches are most the labor market in terms of internal migration and important in securing such gains. employment costs. Field evidence suggests that a significant share of counterfeit firms are able to continue producing similar goods legitimately Human Capital Development under licensing agreements after IPRs are enforced Perhaps the most important complementary factor (Maskus 2000c). In this sense, adjustment may be is a strong commitment to education, training, and less difficult than anticipated. Net job losses in for- skills development. The positive role of educational merly infringing firms could, however, be signifi- attainment in economic growth is well established cant in countries with extensive reliance on empirically. An economy with an abundance of counterfeiting. Countries may wish to establish skills will probably invest more in innovation and training and assistance programs for displaced product development, but such investment is more workers. likely where IPRs are protected. It is also important to foster flexibility in the mar- There are other arguments as well. IPRs have ket for technical and managerial personnel, which stronger stimulative effects in countries with an are important conduits for learning technologies adequate endowment of skills than in countries that and adapting them to new uses. In doing so, due are scarce in skills. One reason is that a nation with recognition of the role of appropriate nondisclosure a greater supply of technical and managerial skills is requirements in protecting trade secrets is in order. more capable of successfully adapting, and manag- The issue of capital markets may be more one of ing, foreign technology to local conditions. Teece scale than of flexibility. The ability of local entrepre- (1977, 1986) found that the costs of transferring neurs to undertake R&D and to commercialize new technology decline with increases in the local supply products is greatly lessened in an environment of of technical and professional workers. Moreover, limited capital. Countries may wish to liberalize strengthened IPRs reduce these transfer costs, as restrictions on capital flows, recognizing that foreign licensors and licensees operate in an environment of investors may be willing to take risks on new enter- freer information flows and greater certainty (Arora prises.2 Establishing venture capital markets may be 1996; Yang and Maskus 2001). appropriate in some circumstances. It is also advis- Finally, economies with stronger educational able to move toward market allocation of investment attainment and skill endowments are better able to and away from public direction of capital. diffuse technical information into competitive uses through honest means of discovery and competi- Technology Infrastructure tion. Nelson and Pack (1999) point to the impor- tance of learning and technical adaptation among While IPRs constitute an important stimulus for enterprises as critical in fostering structural change technology acquisition and adaptation, they may be in East Asian economies. While this may have hap- usefully supplemented by programs to promote pened in an environment of permissive imitation technical change. Industrial countries and many and copying, the abundant formation of human higher-income developing countries have extensive capital was an important factor underlying the systems of support in this area. Such programs process. With the advent of stronger IPRs it range from public assistance for basic R&D in uni- becomes yet more important to build a sound basis versities and research institutes to extension services of education and skills for competitive purposes. in agricultural science. They also provide incentives for commercializing the results of public research and encouraging collaborative research ventures Factor Market Flexibility among private firms and between private and pub- Tightened intellectual property protection is likely lic enterprises for the development of new tech- to raise pressures for structural adjustment in many nologies and products. Such models might be economies. Counterfeit production and piracy will usefully adopted in many developing countries if 372 Benefiting from Intellectual Property Protection tailored to specific circumstances and implemented eign competition. Thus, to strengthen IPRs while in a transparent and procompetitive manner. There maintaining closed markets is to work at cross-pur- is, however, an opportunity cost to allocating scarce poses. For example, a patent has more power in the budgetary resources to R&D programs. For exam- presence of an import quota on similar goods, ple, the social returns in the least-developed coun- which narrows consumer substitution choices. tries would probably be small in relation to those Competitive markets help limit the effective scope from further improvements in primary education of IPRs to their intended function--to foster invest- and in other pressing development areas. ment through competition between enterprises, but Technology development processes could benefit not to prevent fair entry. in many countries from the use of incentives to There are additional reasons why IPRs and open bring publicly sponsored inventions to the market- markets are complementary. Openness improves a place. According to survey evidence, public research country's access to available international technolo- institutes in developing countries often develop gies, intermediate inputs, and producer services, all useful inventions that fail to be commercialized items that can raise domestic productivity. The evi- (UNCTAD 1995b). This problem is common, for dence demonstrates that such flows are deterred by example, in China's state-run science academies weak patent rights and trade secrets (Maskus (Maskus, Dougherty, and Mertha 1998). Finding 2000a). In addition, a critical purpose of IPRs is to mechanisms under which public agencies and pri- encourage investment in improved product quality, vate enterprises can cooperate in such commercial- which is essential for breaking into export markets. ization could bring a number of new technologies Similarly, IPRs can support marketing investments to the market, with benefits for consumers. Intellec- that raise product demand and permit economies of tual property protection plays an important role in scale. sorting out the appropriate claims to the associated These observations support certain policy pre- economic returns. scriptions as countries strengthen their IPRs. First, Low levels of R&D spending may be associated it is important to continue efforts to liberalize with such factors as an inadequate environment for restrictions on trade, investment, and services. Sec- risk-taking, taxation systems that do not recognize ond, while authorities should remain vigilant about R&D as a business cost, and weak information the potential for licensing abuses, the common about technological opportunities. Policies could practice of inspecting all proposed licensing con- aim to remove such impediments. This observation tracts and requiring costly modifications and dis- would pertain especially to ensuring competitive closure clauses serves mainly to limit access to prospects for small and medium-size enterprises, advanced technologies. Thus, it seems advisable to which remain the source of much innovation in adopt a more open stance toward technology agree- both industrial and developing countries. ments and to replace technology-monitoring offices Interestingly, R&D activity by local enterprises is with reliance on competition rules. an important conditioning factor for effectively Industrial countries stand to gain considerably absorbing technologies transferred from abroad. For from a stronger global IPR regime. In turn, they example, Dougherty (1997) found that in Chinese should carry out their obligations to provide liberal manufacturing enterprises, TFP growth induced by access to their own markets. If IPRs are to support foreign licensing contracts was significantly higher more advanced production structures in developing where domestic enterprise partners were engaged in countries, those countries cannot be denied the R&D programs of their own. ability to compete abroad. In this context, countries adopting new regimes have a long-term interest in promoting free trade in goods in which their own Open Market Access emerging intellectual property advantages will sup- Economies that are more open to trade and FDI port exports. For example, developing countries experience a growth premium, relative to closed could build advantages in such goods as textiles and economies, from strengthening their IPRs (Gould apparel, handicrafts, local cultural products, and and Gruben 1996). One reason is that stronger processed foods. To ensure such gains, developing property rights create market power, which is more countries should push their richer counterparts to easily abused in economies that are not open to for- implement the agreement on phasing out the Multi- 373 T E C H N O L O G Y A N D I N T E L L E C T U A L P R O P E R T Y fibre Arrangement, to avoid the use of protectionist Rights (TRIPS) invites nations to consider the inti- technical product standards, to liberalize agricultur- mate linkages between intellectual property protec- al protection, and to exercise restraint in the use of tion and competition policy. It is instructive to set antidumping restrictions. out the major issues in order to understand the tradeoffs and complexities they pose.3 Competition Policy Regulating Monopoly Prices Competition rules are used to discipline anticom- petitive practices in the use of IPRs. The essence of The danger of monopoly pricing is rarely the focus IPRs is to define the boundaries within which an of competition policy per se and is more often the inventor enjoys exclusive rights to the use of her cre- subject of price regulation for purposes of public ation. To abuse an intellectual property right is to health. Box 36.1 provides examples of two cases in try to extend one's exploitation beyond the limita- which governments have taken action. Competition tions established. Claims that a rightsholder has policy tends to ignore the pricing decisions of firms engaged in anticompetitive activity are often com- protected by IPRs, since property rights permit plex and require significant judicial and legal firms to extract some portion of consumer surplus expertise in their interpretation. as the reward for innovation. Firms set prices that Several developing countries and countries in take account of market substitutes, which are rarely transition have recently upgraded or adopted com- absent in a competitive economy. In that context, petition regimes, but this area is open to consider- the proper role of policy is to ensure that products able reform. Article 40 of the Agreement on and technologies face effective competition within Trade-Related Aspects of Intellectual Property the relevant market. B O X 3 6 . 1 PHARMACEUTICAL POLICIES AND THE TRIPS AGREEMENT In response to the TRIPS agreement, South Africa domestic firms may meet demand. They also per- and Brazil recently introduced laws bearing mit parallel importation (imports of original or directly on their ability to react to price increases generic versions without the authorization of the that may emerge from patents. The greatest spur South African patent holder) of drugs and allow to these attempts to limit patent rights came the health minister to override regulatory deci- from a desire to procure AIDS drugs at affordable sions concerning the safety and registration of prices in order to manage an enormous health medicines. The law requires pharmacists to care crisis. Both laws are controversial, and the employ generic substitution (prescribe generic Brazilian legislation briefly became the subject of versions of patented drugs) unless the doctor or a WTO dispute. patient forbids it; sets limits on pharmacy markup rates; and bans in-kind inducements from drug South Africa's Medicines Law manufacturers to physicians. In November 1997 South Africa enacted signifi- Although it may be a heavy dose of regulation, cant amendments to its Medicine and Related South Africa's law is probably consistent with the Substances Control Act that permit the health TRIPS agreement (Abbott 2000). While some minister to revoke pharmaceutical patent rights legal scholars claim that patent rights necessarily in South Africa if the associated medicines are extend to an ability to preclude parallel imports, deemed too expensive. The amendments further the bulk of opinion is that TRIPS Article 6 provides empower the minister to order compulsory full latitude for each country to choose its own licensing if the patentee engages in abusive prac- policy on exhaustion of a patent--the point of tices, defined basically as failure to sell a drug in distribution (national, regional, or international) adequate amounts to meet demand or refusal to at which the rights of an IPR holder to control fur- license the product on reasonable terms so that ther sales are exhausted. Beyond this issue, TRIPS 374 Benefiting from Intellectual Property Protection B O X 3 6 . 1 ( C O N T I N U E D ) Article 31 provides ample grounds under which The legislation explicitly defines "failure to be compulsory licenses may be issued, subject to worked" as "failure to manufacture or incomplete certain conditions (Watal 2001). In particular, manufacture of the product" or "failure to make licensing may be compelled when a prospective full use of the patented process." Although the user has not been able to obtain a license from Brazilian industrial property law refers to all the patent holder on reasonable commercial patents, its most aggressive use is aimed at trans- terms within a reasonable period of time, so long ferring production of AIDS drugs to domestic as market-based compensation is paid. Compul- firms and government agencies in order to sory licenses may be issued without observing reduce their prices below those on the U.S. and even these constraints in cases of national emer- European markets. Media reports indicate that gency. Finally, the price-control provisions of the this active intervention has dramatically reduced South African amendments do not seem to be treatment costs in Brazil.* In combination with restrained by the TRIPS agreement, which does prevention programs and effective methods of not address domestic health regulation. distribution and clinical treatment, the country has limited AIDS mortality to far lower levels than Brazil's Industrial Property Law those in Sub-Saharan Africa. Law 9279, which came into force in 1997, updat- In early 2001 the United States issued a com- ed most aspects of Brazil's industrial property plaint at the WTO about Brazil's "working" regime to comply with the TRIPS agreement. It requirements. The case was suspended shortly provides patents for pharmaceutical products as thereafter, in part because of pressure brought to required, but it permits the issuance of compulso- bear on the U.S. government by advocates of ry licenses in cases where patent holders choose inexpensive access to essential medicines for to supply the market through imports rather than developing countries. through local production. That is, the law does not recognize imports as a method of meeting its *"Look at Brazil," New York Times (January 28, 2001). requirement of "working" in the Brazilian market. Source: World Bank (2001c). The South African and Brazilian cases reflect 14 years, as well as a statement that nothing in the widespread concern that the implementation of TRIPS agreement could be used to prevent those patent protection in poor countries could keep new countries from taking whatever steps are necessary medicines out of reach for those most in need. It to procure essential medicines at low cost in the should be noted that the TRIPS agreement cannot event of medical emergencies. Thus, the Doha require that countries patent drugs that were avail- agreement essentially permits the least-developed able on the market before patents were introduced, countries to ignore patent rights in drugs for the while pharmaceutical companies may choose not to foreseeable future. take out patents in poor economies, thereby provid- ing room for governments to procure generic ver- Interpreting Licensing Agreements sions. Developing countries, however, were sufficiently worried about the potential impact of Alleged abuses relate most often to selling practices patents on future drugs that they pushed for some and licensing restrictions. A vast literature exists on relief at the WTO ministerial meeting in Doha, the competitive effects of market power created by Qatar, in November 2001. An important outcome of patents, trademarks, and protected know-how that meeting was an agreement by WTO members (OECD 1989).4 There are few concrete guidelines in that the least-developed countries could put off the area because of the complicated nature of mar- implementing patent protection for an additional kets for information and technology. Vertical licens- 375 T E C H N O L O G Y A N D I N T E L L E C T U A L P R O P E R T Y ing agreements, for example, may ensure that dis- exclusive rights to competing technologies and tributors maintain downstream product quality, products. Such efforts effectively are horizontal which aids competition, but tie-in sales of unrelated mergers, which may be analyzed in terms of their products to technology purchasers could extend the effect on current and future market concentration. scope of the initial property right and harm compe- A final problem is nonprice predation, in which tition. IPRs may be used to bring bad-faith litigation and Among the potential competitive problems that opposition proceedings in order to exclude and arise from the exploitation of IPRs are, first, harass competitors. This may be particularly trou- cartelization of horizontal competitors through blesome where potential rivals are small and new licensing agreements that fix prices, limit output, or and so have insufficient resources to defend them- divide markets. Competitors may be licensees or selves against extensive litigation. Other forms of licensors, either in the market for the product or entry deterrence may be practiced as well, and the technology itself or in extended markets. For exam- burden of competition authorities is to distinguish ple, patent-pooling and cross-licensing agreements predation from legitimate enforcement of IPRs. For between competing licensors may reduce competi- example, firms may refuse to license technologies in tion in downstream product markets that use the particular markets or to certain firms, which could licensed technologies as key inputs, particularly be construed either as a legitimate business practice where the agreements set prices or restrict territo- or as unfair competition. ries and fields of use. Thus, there are complex relationships between Competition agencies find it difficult to set rules IPRs and their potential abuse. Competition covering such licensing agreements. Rather, the authorities must develop the capability to distin- focus has been on whether an agreement could guish various forms of behavior in terms of poten- cartelize a significant share of a market, requiring a tial impacts on static and dynamic competition. It is definition of "significant market" and identification therefore probably advisable for countries develop- of competing products and technologies. Concerns ing competition rules to adopt the U.S. "rule of rea- also arise over agreements that require resale price son" approach rather than attempt to codify rules maintenance of distributors' prices. Such agree- covering specific actions, as in the EU approach. ments could fix prices in ways that would not be Perhaps most important is the recognition that necessary for monitoring and enforcing quality. It is the anticompetitive effects of licensing and sales evident that such risks are greater the more regulat- agreements depend critically on market structure. ed is entry into distribution contracts--a common In many developing economies entry of new com- situation in many developing countries. petitors is made difficult by monopoly distributor A second concern is the exclusionary effects of laws, lack of parallel imports, general trade and license agreements. Such agreements could prevent investment protection, and inadequate financial other firms from competing by raising entry barri- markets. Thus, it is important to consider the wider ers. For instance, tie-in sales could grant a licensor a relationship of business regulation to the develop- dominant position in the market for the tied good. ment of stronger IPRs. (Potential competitors would be forced to enter the markets for both the technology and the tied good.) Treatment of Parallel Imports Similar problems emerge where licensees are required to use only the licensor's current and Parallel imports are goods brought into a country future technologies. A third concern arises where without the authorization of the patent, trademark, licensors, either individually or in patent pools, or copyright holder after those goods were placed hamper the development of competing technolo- legitimately in circulation elsewhere. Note that gies through exclusive grant-back provisions and these goods are legitimate copies, not pirated copies exclusivity arrangements in future technology pur- or knockoffs. chases. Again, competition policy must assess the Parallel imports are regulated by the territorial potential anticompetitive impacts of such arrange- exhaustion of IPRs. Under national exhaustion, ments if there is to be intervention. rights end on first sale within a nation, but IPR Another general class of problems relates to owners may prevent parallel trade with other coun- attempts to acquire market power by purchasing tries. Under international exhaustion, rights are 376 Benefiting from Intellectual Property Protection exhausted on first sale anywhere, and parallel it is better to rely on private enforcement of con- imports are permitted. A third option is regional tractual exclusive territories while permitting par- exhaustion within a group of countries. allel trade. The TRIPS agreement recognizes, in Article 6, the Among the arguments made in favor of control- prerogative of each country to set its own regula- ling parallel imports is that price discrimination tions covering parallel imports. This treatment was need not be harmful and, under certain circum- critical in securing the adherence to TRIPS of stances, can raise economic well-being (Varian numerous developing countries, which maintain 1985). Banning parallel trade would result in inter- the right to set specific exhaustion regimes. national price discrimination, or one price set per Exhaustion policies vary widely, even among market. By contrast, full parallel trade would force industrial economies. With few exceptions, the uniform pricing by the IPR holder, subject to differ- European Union (EU) adopts exhaustion in all ences in transport and marketing costs. Economies fields of intellectual property within the EU but with inelastic demand would face higher prices bars parallel imports coming from outside its terri- under price discrimination than under uniform tory. The United States maintains a "common-con- pricing, harming consumers. Countries with elastic trol exception" in the case of parallel imports of demand--typically, developing economies--would trademarked goods. This principle permits trade- enjoy lower prices under price discrimination. mark owners to block parallel imports except when Indeed, in the presence of parallel trade, foreign both the foreign and U.S. trademarks are owned by rightsholders may choose not to supply such coun- the same entity or when the foreign and U.S. trade- tries because local demand might be insufficient mark owners are in a parent-subsidiary relation- (Malueg and Schwartz 1994). Thus, restricting par- ship. In addition, blocking parallel imports requires allel imports could lower welfare in developing a demonstration that the imports are not identical economies through higher prices and reduced in quality to original products and may cause con- product availability. fusion among consumers. Owners of U.S. patents Most developing countries, however, oppose and copyrights are protected from parallel imports. restricting parallel trade (Abbott 1998b). In part, Australia deregulated parallel imports in copyright- this reflects a concern that domestic prices of phar- ed compact disks in 1998. maceuticals could actually be higher for imported Few developing countries restrict parallel trade. goods under price discrimination. Furthermore, Some nations substitute laws mandating a sole many nations see opportunities for being parallel national distributor for products imported under exporters and achieving export and industrial trademark or copyright, effectively banning parallel growth through that channel, seeing little likelihood imports. In other countries parallel imports are that their markets will go unserved. widely seen as a useful policing device against price Whether price discrimination harms or helps collusion arising from territorial restraints, and par- particular nations depends on circumstances. allel exports are viewed as a channel for penetrating Malueg and Schwartz (1994) argue for banning par- foreign markets. allel imports on the grounds that perfect price dis- This wide divergence in policies toward parallel crimination would result in a net expansion of imports suggests that there is no clear answer to global output and would increase global welfare, whether such imports are beneficial or harmful in while ensuring that goods are provided to low-price welfare terms. Three arguments are advanced in markets. Maskus and Chen (2000) point out that favor of permitting parallel trade: (a) Restrictions parallel imports may be most beneficial within a on parallel imports amount to nontariff barriers to regional trade agreement, where transport costs are goods that have legitimately escaped the control of low, but that they could be costly otherwise. Parallel IPR owners. (b) Parallel imports could play an imports might also be restrained on the ground that important policing role against abusive price dis- they free-ride on the investment, marketing, and crimination and collusive behavior; because the service costs of authorized distributors. If this is a colluding firms could be foreign, the loss to con- serious problem, markets may suffer slower rates of sumers from the firms' behavior is not balanced by product introduction. a gain in local profits. (c) Government enforce- A final point is that parallel imports could disrupt ment of territorial rights invites rent-seeking, and national price-control systems in pharmaceuticals 377 T E C H N O L O G Y A N D I N T E L L E C T U A L P R O P E R T Y (Box 36.2). In terms of current global social policy, the absence of restraints on parallel exports of these this issue arises most dramatically in efforts to treatments, pharmaceutical companies are reluctant transfer treatments for HIV/AIDS and other dis- to participate. Thus, an effective system of restric- eases to the least-developed nations at low prices. In tions is called for in this area. B O X 3 6 . 2 D I F F E R E N T I A L P R I C I N G Differential pricing, also called tiered pricing or sis, however, finds little evidence that prices of equity pricing, is a system in which prices of iden- medicines are related to per capita income; tical goods increase with ability to pay, usually indeed, they are often higher in developing than measured by per capita income. It is discussed in industrial countries (Scherer and Watal 2001). most frequently in the area of medical drugs and Four primary factors seem to explain the fact that vaccines. For example, under a differential pric- prices in developing nations are often higher ing scheme for HIV/AIDS drugs or malaria vac- (and prices in industrial countries lower) than cines, prices in poor countries would be set at far might be anticipated on the basis of tiered pric- lower levels than in rich countries. Differential ing. First, many developing countries retain high pricing often exists within countries as well, tariffs and taxes on medicines, and local distribu- through procurement mechanisms. Bulk pur- tion systems may be monopolized and inefficient. chases by governments or health care institutions Second, drug manufacturers may find it more can achieve considerable discounts for eligible profitable to sell low volumes of branded drugs at patients in relation to prices charged on open high prices to the relatively wealthy in developing markets to uncovered patients. (In the latter case countries than at low prices in high volumes to the price differences may not be related to levels the poorer segments. Third, because of price of patient income, if higher-income individuals controls and monopsony purchasers in higher- are more likely to be included in the purchasing income economies, prices may be effectively lim- programs.) For differential pricing to be sus- ited there. Fourth, despite the scope for tained, restraints against reselling the drugs must segmenting markets through transport costs, be in place, which is often difficult. restraints on parallel imports, and differences in Differential pricing is not the same as the stan- packaging and trademarks, countries may be dard economic concept of price discrimination, effectively integrated in other ways. Specifically, under which products are priced higher (or reference pricing systems (under which prices in lower) in countries or to consumers with less (or one country are controlled according to some more) elastic demand. Price discrimination is a average of prices in other countries) and con- profit-maximizing strategy that should emerge cerns that consumers in high-income economies naturally within segmented markets where reli- would demand similar price advantages provide able information about demand exists. Differen- an incentive for firms to refuse significant price tial pricing, by contrast, reflects a conscious cuts in poor countries. These processes likely decision by firms and governments to organize form a major impediment to differential pricing, prices for the purpose of distributing critical which would otherwise be in the interests of goods, particularly essential medicines, to poor pharmaceutical companies. consumers at low cost. To be sure, some differential pricing exists at Price discrimination would, in principle, sup- the international level. It is well documented that port a pricing structure that correlated highly this strategy works for vaccines, where large dif- with differential pricing under the following ferences in per-unit prices exist between industri- assumptions: demand becomes more inelastic as al countries and poor countries. Numerous per capita income rises; demand may be perfect- pharmaceutical companies now provide anti- ly revealed; firms may set prices without restraint; retroviral drugs to the poorest countries at steep and markets are fully segmented. Empirical analy- discounts (see Box 35.2, in this volume). To date, 378 Benefiting from Intellectual Property Protection B O X 3 6 . 2 ( C O N T I N U E D ) this system has relied on bilateral negotiations prices of program drugs in poor countries in their between particular countries and firms, so the reference lists for price controls. Finally, agree- coverage of the price cuts in poor countries ment has to be reached on a mechanism for remains small. procuring drugs from pharmaceutical firms for Many analysts believe that differential pricing cheap distribution. Most likely this would involve must be at the core of international attempts to bulk purchasing programs, which could be car- distribute medicines to the low-income countries ried out by procurement agencies for individual and that mechanisms must be found to establish governments, groups of governments, or interna- and sustain such pricing tiers (Maskus 2001). tional organizations, in order to negotiate price Accomplishing this will require action on a broad cuts. Where the target drugs are patented, such international front. A consensus must be reached purchasing programs could be complemented by on which drugs are to be included in a system of a licensing regime in which royalties would be tiered pricing and which countries would qualify paid to the patent holder to permit local produc- for steep discounts under the program. Develop- tion and distribution in poor countries. These ing countries need to relax import constraints on licenses would in many cases have to permit medicines, improve their distribution and health imports from licensed producers in third markets care systems, and establish guarantees that funds because small nations could not efficiently pro- allocated for drug purchases will be managed duce the drugs themselves. efficiently and that program drugs will not be A system of this kind could not be financed diverted outside targeted markets. Higher- without considerable assistance from the rich income and middle-income developing countries and international development institu- economies need to prevent parallel imports of tions. Thus, it would require a substantive the program drugs, which means that extensive increase in aid budgets, perhaps buttressed by monitoring and labeling systems are required. tax incentives for firms that donate drugs to poor Industrial countries need to forgo including countries. Facilitating Technology Transfer gies for the management of important public health and environmental problems could choose to use Governments in many developing countries remain TRIPS to support highly restrictive licensing doubtful that the intended benefits of the TRIPS arrangements or not to license the technologies at agreement, especially in terms of additional tech- all (Watal 2000a). nology transfer on reasonable terms, will be forth- This omission could induce an effort to roll back coming. TRIPS Articles 66 and 67 commit some of the TRIPS standards. To forestall this, an industrial nations to use best efforts to identify important initiative for enterprises and agencies in measures they could take to encourage such trans- industrial countries to undertake in the near term fers, in particular to the least-developed countries, would be a program to make the technology transfer and to promote mechanisms to build a sound and commitments more effective. Such a program could viable technological base in the recipient countries. remove any impediments to outward transfers that To date, those best efforts have been nil, generating persist in the industrial economies. It could also concerns that technology exporters do not intend to envision a fund for providing considerably more employ TRIPS in a manner that would be seen as technical and financial assistance to poor countries equitable by technology importers. Rather, concerns in the implementation and administration of IPRs. are mounting that firms owning critical technolo- Competition authorities in the industrial countries 379 T E C H N O L O G Y A N D I N T E L L E C T U A L P R O P E R T Y could assist officials in developing countries in many other policy areas and priorities that will determining the most appropriate guidelines for determine the magnitude of the costs and benefits dealing with anticompetitive aspects of international of multilateral agreements and disciplines in this technology licensing. Finally, in recognition of the area. To mention one example that has attracted spillover benefits from environmental protection, much attention in recent years--the impact of IPRs multilateral negotiations could be held to define the on the price of vital medicines such as those for terms under which compensated licenses of right HIV/AIDS--it is well recognized that something would be issued in critical technologies. that is at least as important as the drugs them- Although issues related to protection of intellec- selves--even if they were provided at less than tual property are important for developing coun- cost--is the medical infrastructure needed to deliv- tries, care must be taken not to lose sight of the er the drugs (Box 36.3). B O X 3 6 . 3 H E A LT H C A R E S Y S T E M S A N D E S S E N T I A L M E D I C I N E S While the strengthening of patent rights in medi- quate for the size of the task. cines may have important impacts on the afford- Second, governments may choose not to place ability of drugs in poor countries, the issue a high priority on funding the development of recedes in significance in comparison with the medical human capital and health care facilities. acute needs for improved health care systems. In part, the inability to pay for health programs is Investment in nutrition, health services, and the result of chronically limited budgets, con- health delivery is vital for economic development. strained by other fiscal needs, but in part, it There are, however, significant obstacles that reflects a policy failure; government officials may often go far beyond the health sector itself. Suc- place relatively little value on social programs, cessful health outcomes may require changes in including health care. It may be difficult to personal behavior (regarding tobacco use, sexual achieve effective representation for the interests practices, and so on) and in social norms (for of patients and health facilities within the political example, discrimination against girls and women) process. Extremely poor governance in some and improvements in public administration. countries plagues health service delivery and Many poor nations have a chronic shortage of probably wastes what little resources are allocat- clinics, hospitals, medical personnel, and means ed to health. The result is inadequate provision of for transporting sufferers to the facilities that do all forms of health care, and very little funding for exist. A further problem is unequal distribution of drug procurement. The WHO World Health Report available services, with rural access being far 2000 recommended that governments purchase lower than urban, and public health care being "priority interventions," including essential inferior to private health care. Existing facilities drugs, for the entire population, but few low- are overwhelmed by the crush of patients and income economies come close to meeting this have limited ability to provide treatment. There standard. This funding shortfall not only results in are many underlying reasons for this situation. weak delivery and care systems but also means First, weak education systems contribute to the that the state of legislation and regulation may shortage of medical professionals. Doctors and lag the acute need to deal with endemic disease. nurses often find better employment alternatives For example, insufficient staffing and expertise in abroad. It is estimated that in 20 African coun- health agencies may imply approval periods for tries, more than 35 percent of nationals with a drugs entering the market that are longer than university education are now living abroad. necessary. Simultaneously, drug safety and quali- Donations of time and effort by foreign medical ty may be compromised by inadequate testing personnel are obviously beneficial but are inade- and monitoring programs. 380 Benefiting from Intellectual Property Protection B O X 3 6 . 3 ( C O N T I N U E D ) Third, conditions in the poorest countries, Saharan African countries other than South Africa especially in difficult locations such as Sub-Saha- average US$5 per person per year in public out- ran Africa, pose severe ecological obstacles to lays for health, in the face of massive epidemic health and economic development. A serious disease and plummeting life expectancies. India attack on malaria would require a rigorous, sci- is currently averaging around US$4 per person ence-based, and well-financed revival of mosqui- per year in public health outlays. China's public to control efforts. The social barriers are also real. sector health spending is US$5 per person per African countries looking to reduce the spread of year. By contrast, the rich countries spend more HIV/AIDS must confront seriously the low status than US$1,500 per person per year in public sec- of women in sexual relations and the high resort tor outlays for health. The meager spending in to commercial sex workers by migrant male the poorest countries is fundamentally a reflec- workers. Islamic countries must be especially tion of the poverty of those countries. Private out- attuned to providing education to girls, since the lays cannot make up the difference because poor gaps in male and female literacy tend to be high- households are too poor to pay for their own est in Muslim societies. Similar problems confront health services and because public health mea- low-status groups and ethnic minorities. sures are vitally needed. Widespread poverty lies at the root of such Development assistance aimed at health pro- problems. Impoverished families may forgo med- grams is meager. Donor support for health in ical treatment in favor of other needs. Extensive 1999 amounted to just US$1 per African per year, informal employment limits the development of US$1 per Indian, and US$1 per Chinese. It is likely private or public insurance markets that can pool that the shortfall of donor financing was an health risks across large patient volumes. Poverty important reason why the world community did makes it difficult for governments to establish an not prevent the AIDS epidemic from exploding in adequate and equitable tax base for funding the past decade, or halt the resurgence of malaria, public health programs. tuberculosis, and vaccine-preventable diseases. In One stark measure of the shortfall is the order to achieve even minimally effective health extraordinarily low level of resources flowing to systems in the poor countries, levels of support health in the world's low-income countries. Sub- would need to increase considerably. Notes This chapter draws on material in Maskus (2000a). 1 While this statement is widely accepted by economists and business scholars and finds extensive anecdotal support, I have found no systematic econometric study of its applicability in developing countries. 2 Field research found anecdotal evidence that foreign venture capitalists are actively seeking new projects in China, Hong Kong (China), and Taiwan (China) to the extent that they are allowed to take equity positions (Maskus, Dougherty, and Mertha 1998). 3 The papers in Anderson and Gallini (1998) provide an excel- lent and comprehensive overview. 4 The OECD also publishes reviews of competition policies in its member countries, which are useful sources of information on how competition authorities define and deal with IPR abuses. 381 37 A R V I N D S U B R A M A N I A N Proprietary Second, and following from the first point, the North has a Protection of strong economic and ecological interest in ensuring the preser- vation of genetic resources. Genetic Developments in biotechnology increase the prospects of com- Resources and panies in the North that are interested in better harnessing Traditional the South's gene pool. A number of research and other institu- Knowledge tions in industrial countries have entered into contracts with developing countries for search- ing out genetic resources ("bio- T he Agreement on Trade-Related prospecting"). The best-known example is the Aspects of Intellectual Property contract signed in 1991 between Merck & Co. and Rights (TRIPS) brought many aspects of intellectu- Costa Rica's Instituto Nacional de Biodiversidad al property (IP) into prominence in the developing (INBio) under which Merck paid INBio US$1 mil- world. This increased consciousness has prompted lion to provide a limited number of documented developing countries to explore the use of IP to pro- plant samples for isolation, as well as plant and tect the vast repositories of genetic resources that insect extracts for use in the drug discovery they house and to exploit the potentially significant process.2 economic benefits of these resources. A number of Meanwhile, serious concerns have emerged about circumstances in the 1990s contributed to high- the loss of global biodiversity, which could have cat- lighting the possible importance of such protection. astrophic consequences.3 Although the estimates First, the basic physical reality confers an advan- vary greatly, a consensus seems to be emerging that tage on developing countries. These countries, it is large-scale destruction of biodiversity is occurring, estimated, are home to about 90 percent of the with potentially serious consequences for world's genetic resources and traditional knowledge humankind. For example, it is estimated that the (Wilson 1992).1 More than 90 percent of the world's biodiversity-rich tropical rain forests and moist research and development activity, however, takes forests are being cleared at the rate of 1.8 percent of place in the industrial countries (Sachs 2001). The their area per year, which translates into the loss picture of a gene-rich, technology-poor South and a each year of an area of forest the size of the state of technology-rich, gene-deficient North creates the Florida (FAO 1995). Wilson (1992) estimates that potential for mutually beneficial bargains between deforestation is leading to a loss of about 2.7 per- the two groups. cent of the species in these forests every decade, and 382 Proprietary Protection of Genetic Resources and Traditional Knowledge with the rate of deforestation accelerating, the bio- mechanism for international cooperation in rela- diversity loss could be enormous. A median esti- tion to genetic resources--or, to put it differently, mate is that there are approximately 10 million what is the market failure that warrants corrective species in the world, and the current extinction rate social action? What are the main considerations that of species is about 5 percent per decade (Raven and should underlie a system of proprietary protection McNeely 1998). At that rate, about 50,000 species for traditional knowledge and genetic resources, could be lost each year, of which only 70,000 have and what are the practical difficulties associated been recognized and named, and about two-thirds with implementing a cooperative scheme? What are of living species would be lost over the course of the the main options facing developing countries in the 21st century. Wilson (1992: 280) has warned, period ahead as they consider international cooper- "Clearly we are in the midst of one of the great ation on this issue? What, realistically, are likely to extinction spasms of geological history." be the economic payoffs to developing countries for In a curious inversion of the accusations of piracy their genetic resources? Finally, what future actions of conventional forms of intellectual property rights should developing countries be considering to real- (IPRs) leveled against developing countries in the ize their objectives? 1980s and 1990s, the gene-rich South had similar complaints against the gene-importing North. In a The Subject Matter and Its Value number of prominent incidents companies in the North turned out to be involved in the use--with- Developing countries are seeking to protect two out remuneration--of plants or resources found in related but distinct resources: traditional or indige- developing countries. The term "biopiracy" began nous knowledge, and genetic resources, which to acquire resonance. In some instances the practice include seeds, endoplasm, rare animal and plant led to the development of a patented product, fuel- species, and parts of plants and animals. The former ing the perception that developing countries were refers typically to practices in farming and agricul- doubly disadvantaged: not only were their resources ture that have been devised and refined over long being used without remuneration, but the resulting periods of time and can be clearly attributed to end product came with a higher price tag attached human actions.6 The latter, by contrast, are not usu- because of the ensuing monopoly.4 ally the product of human invention or creativity Finally, in the preparations for the Seattle minis- but are typically found in nature. The real impor- terial meeting of trade ministers in late 1999, a tance of genetic resources lies in the encoded genet- number of developing countries raised the issue of ic information that is proving to be valuable in protection of traditional knowledge. Bolivia, developing medicines and pharmaceutical products Colombia, Ecuador, and Nicaragua specifically pro- to cure human diseases and for raising agricultural posed that the Seattle conference establish a man- productivity. date for the next round of trade negotiations to (a) Why is it important to protect indigenous knowl- carry out studies, in collaboration with other rele- edge and genetic resources? First, there is the eco- vant international organizations, in order to make nomic benefit. Plants and other organisms are recommendations on the most appropriate means natural biochemical factories and yield many prod- of recognizing and protecting traditional knowl- ucts that enhance human welfare. Leaving aside the edge as the subject matter of intellectual property fact that a large proportion of the world's popula- rights; (b) on the basis of these recommendations, tion--in China, India, and Brazil, for example-- initiate negotiations with a view to establishing a depends on plants for medicine, the modern drug multilateral legal framework that would grant effec- industry is founded on genetic resources. It is esti- tive protection to expressions and manifestations of mated that more than 100 drugs in international traditional knowledge; and (c) complete this legal commerce are derived from plants. For instance, of framework in time for its inclusion as part of the the top 20 drugs sold (with a market value of US$6 results of the current round of trade negotiations billion) in the United States in 1988, 2 were taken (WTO, WT/GS/W/362).5 directly from natural resources, 3 were semisynthet- This chapter addresses the following questions: ic, 8 were synthetics with chemical structures mod- What is the subject matter under discussion, and eled on natural compounds, and 7 had their what are its potential benefits? Why do we need a pharmacological activity defined as a result of 383 T E C H N O L O G Y A N D I N T E L L E C T U A L P R O P E R T Y research on natural products (Raven and McNeely What, then, is the case for government intervention 1998).7 Advances in biotechnology, in particular the at national and international levels? possibility of using genetic information and of Traditional knowledge and genetic resources have transferring genes from one species to another, con- the characteristics of public goods. First, they are siderably enhance the potential economic value of nonrival in consumption: once created (as in the genetic resources.8 case of traditional knowledge) or preserved (as in Second, biodiversity has important ecological the case of genetic resources) they can be used by functions that sustain plant and human life. It pro- any number of people. One person's enjoyment of vides a variety of services: protecting watersheds, traditional knowledge and genetic resources does regulating local climates, maintaining atmospheric not diminish another's. Second, they are nonexclu- quality, absorbing pollutants, and generating and sionary: once they are created or preserved, it is dif- maintaining soils, among others. A properly func- ficult to prevent their being used by everyone. In the tioning ecosystem is necessary to maintain the basic case of genetic resources, access to a very small photosynthetic processes that sustain life on Earth. (even infinitesimally small) quantity may be In addition to these instrumental uses, genetic enough for enjoyment of the benefits they yield resources--plants and animals--may have an (Sedjo 1992; Heal 1998). intrinsic value, in and of themselves, and indepen- It is well known that public goods such as knowl- dent of their usefulness to humankind. Many ecolo- edge and information give rise to a market failure. gists and environmentalists maintain that human This happens because the creator of the public good beings have an absolute moral responsibility to pro- is not remunerated for his or her efforts; the bene- tect plants and animals. As Paul Ehrlich and E. O fits of knowledge and information, once created, Wilson (1991) put it, "Human responsibility in this can be appropriated by everyone else. Ex ante, respect is deep, beyond measure, beyond conven- therefore, the incentive to expend effort to create tional science for the moment but urgent nonethe- socially useful goods is blunted, leading to an less." Finally, biodiversity may have cultural and underprovision of these goods, from the perspective aesthetic value. In certain societies, plants and ani- of society. mals are revered and have symbolic value--ele- An important distinction needs to be made phants in Hinduism, the bald eagle in the United between traditional knowledge and genetic States, the lily in France, and so on. Ecotourism is a resources. Traditional knowledge is the product of manifestation of the aesthetic value of biodiversity. human invention and creation. Genetic resources Although the existence of these noninstrumental are ultimately given to us by nature. The common- values of biodiversity is clear, it is extremely difficult ality with traditional knowledge (or R&D) is that to quantify them. It is important to bear in mind the stock of genetic resources available to that, given the uncertain state of science, a large part humankind--and the growth or decline in those of the economic and ecological value of biodiversity resources--is also the result of human efforts. may be an "option value." That is, a resource may The preservation of genetic resources entails con- have no known value today, but if it is preserved siderable costs, mainly associated with the alterna- now, better information in the future, as science tive uses to which the habitat that houses genetic progresses, will allow more informed decisions to be resources can be put. As the returns to agriculture made about it. If future information suggests that or forestry rise, it becomes more profitable in a pri- the resource has no value, it can be destroyed; if the vate sense to deforest or cultivate than to maintain resource turns out to have a lot of value, it can be forests and preserve the genetic resources that they exploited. But the latter option will be precluded if contain. A partial way around this is to protect the resource is destroyed today; hence the notion of resources ex situ, outside their habitat. Although the option value of preservation. this has the advantage of lower costs, it is feasible only for a small fraction of species--known species. For unknown species, which are in the overwhelm- Why Public Policy Intervention? ing majority, ex situ preservation is by definition The fact that traditional knowledge and genetic impossible. Moreover, the ex situ approach pre- resources have social value does not by itself create a serves only species, whereas long-term species sur- case for government intervention to protect them. vival requires preservation of entire habitats. 384 Proprietary Protection of Genetic Resources and Traditional Knowledge Thus, there is a serious market failure associated National and International Actions with the creation of indigenous knowledge and the protection of genetic resources, necessitating some In practice, the first of the two steps outlined above form of government intervention. The next step is has begun to be implemented. Resource-rich coun- to establish that there is a case for concerted inter- tries such as the Andean Pact countries (Bolivia, national cooperation rather than intervention at the Colombia, Ecuador, Peru, and Venezuela), Australia, national level only. Brazil, Costa Rica, Fiji, India, the Philippines, and The need for international action follows from Thailand have either passed legislation conferring the fact that significant cross-border externalities some form of property rights for traditional knowl- are involved in the preservation of genetic edge and genetic rights or are in the process of resources. That is, the economic value of protecting doing so. genetic resources is not confined to the country These national actions have in turn been sanc- where these resources are found. The economic tioned by a significant international initiative that benefits from medical uses, ecotourism, and preser- has recognized the need for protection of genetic vation of habitats are felt across boundaries. If a resources. The Convention on Biological Diversity cancer cure is found using a plant in Cameroon, the (CBD), signed in 1992, made the leap and moved benefits will be felt all over the world. Hence, all away from the position held previously, that genetic countries have a stake in ensuring that Cameroon resources were the common heritage of preserves resources of benefit to humanity. Since humankind.9 The convention now acknowledges genetic resources are predominantly located in that states have sovereign rights over their natural developing countries and the benefits they provide resources (Art. 15.1); that states also have the are most likely to be enjoyed by consumers in authority to determine access to these resources industrial countries, the cross-border dimension is through national legislation (Art. 15.1); and, cru- significant, strengthening the case for a collective cially, that such access is subject to the prior international effort to preserve the resources. informed consent of the country housing the resource (Art. 15.5). Although it requires countries to take measures to protect genetic resources (for Proprietary Protection of Traditional example, in Article 8), it does not explicitly call on Knowledge and Genetic Resources host countries to institute systems of proprietary Given the public good nature of traditional knowl- protection for genetic resources. It does refer to the edge and traditional resources, it seems that an obvi- need to seek the approval of the holders of the ous solution to remedying the market failure is to knowledge, and, in Article 8(j), it encourages the grant some form of proprietary rights to this subject equitable sharing of benefits arising from the uti- matter. (This is similar to the logic underlying the lization of such knowledge. Significantly, it makes grant of traditional IPRs.) Under this approach, no reference to the need for such systems to be countries that have genetic resources should institute internationally recognized. a regime of property right protection. Given the international externalities and in order to prevent Problems free-riding by other countries, it also seems that any such proprietary rights should be internationally rec- A number of conceptual and practical difficulties ognized. Thus, if a protected species in, say, Madagas- are associated with the institution of a new propri- car is unlawfully appropriated by a company in the etary right for traditional knowledge and genetic United States, it should be possible for the rights- resources (Subramanian 1992; Cottier 1998). holder to seek redress under U.S. law. This means that there should be an internationally recognized system 1.What, exactly, is the subject matter that should be of property rights in traditional knowledge and protected--does it include seed varieties, farming genetic resources (Sedjo 1992; Subramanian 1992; methods, rare species, folklore, traditional knowl- Cottier 1998). Rights acquired in the country of ori- edge? Should the form of protection be different for gin alone are likely to be of limited value if they are traditional knowledge and for genetic resources? not extended to the major markets where derived 2.Who would be the beneficiary of the right--indi- and patented products will be marketed. viduals or groups? Should the right be confined to 385 T E C H N O L O G Y A N D I N T E L L E C T U A L P R O P E R T Y indigenous communities? Conceptually, the right parate actors--individuals, local communities, should be awarded to those who have had a key nongovernmental organizations, governments, role in accumulating knowledge or preserving international institutions, research centers, and resources. How would this be determined, given pharmaceutical, agricultural, and other firms--will that the current state of science may not enable need to be brought together in the search for ways germ plasm to be assigned to any particular site? and means of achieving the common goal of pro- How are rights to be assigned if new varieties tecting biodiversity. draw from a large number of sources of germ The issue in a trade context is a narrower one: plasm? whether there needs to be international coopera- 3.How would the right be acquired? Broadly, there tion, possibly in the form of an international agree- are two options. Under the first option, they ment such as the TRIPS agreement, to accord could be acquired, as in the case of copyright, as proprietary protection for traditional knowledge soon as certain legal conditions are met, with and genetic resources. Clearly, the role of such an minimum formalities. This approach offers agreement would be to complement the other vital administrative ease and simplicity. The second efforts described above. Before embarking on such a option, similar to that taken in the main areas of path, developing countries may need to consider the property rights protection, such as patents and options for international cooperation. There are trademarks, is to require some kind of formal reg- three possible variants, which are discussed next. istration before the right is acquired. This approach would be more complicated but does Option I: Negotiation of a Full-Fledged provide clarity and legal and business security. International System of IPR Protection Formal registration delineates the private and public domain in a practical, clear manner. The most ambitious path would be to seek a full- 4.There is likely to be a thorny transitional issue of fledged international agreement on the protection what to do with traditional knowledge already in of genetic resources. Theoretically, this is the most the public domain and with genetic resources appealing, but there are significant practical diffi- already acquired under the previous regime that culties of a legal and political nature. The legal did not require any form of compensation.10 problems, outlined above, have to do with designing a system of sui generis protection that is easy to One of the main points of opposition to the grant implement. The political problems are related to the of proprietary rights for genetic resources in indus- resistance to be expected from industrial countries trial countries is likely to be the long-standing prin- that are reluctant to embrace any new system of ciple in IP law that such rights cannot be granted for protection, especially one that goes against the grain things found in nature; rights are granted only for of entrenched legal precepts that see only creation, human creations or inventions. In general, the not preservation, as susceptible to proprietary pro- essential common prerequisite for granting propri- tection. etary protection is novelty or distinctiveness, under a doctrine of nonobviousness or inventive step (as Option II: Using and Strengthening the Existing in patents) or one of sufficient level of creativity International Framework (copyright and industrial design) or distinctive fea- ture (trademarks). Thus, protection depends on the A less ambitious approach would be to work within newness of the knowledge and information gener- the parameters set by the CBD and to rely on con- ated (Cottier 1998).11 tractual arrangements between industrial country companies and developing countries to regulate access to the use of genetic resources. The advantage Options for International Cooperation of this approach is its simplicity and the fact that it It is clear that the protection of genetic resources has precedents in practice. Its disadvantages are will require international cooperation on a number twofold. of fronts: resource management, science, technolo- First, voluntary cooperation, although welcome, gy transfer, finance, and so on (see Reid 1998).12 It is cannot be guaranteed in all instances, and the terms also generally understood that a number of dis- of such cooperation will necessarily be influenced 386 Proprietary Protection of Genetic Resources and Traditional Knowledge by whether prior rights to such resources exist. If registers, under strict obligations of confidentiality. these rights are internationally recognized, and if It could also assist in negotiations between compa- infringements are credibly punishable, the reward nies and groups over the use of genetic resources, for maintaining the resources will be higher than it set standards for such contracts, and provide a dis- otherwise would be. In the absence of such rights, pute resolution mechanism. bargaining power is asymmetric, and this could The advantage of this proposal is that it would skew any bioprospecting contracts in favor of the avoid the need to negotiate an international treaty companies. on intellectual property rights for genetic resources. Second, and more seriously, the contract cannot It could also create an incentive for pharmaceutical bind third parties (those that are not parties to the companies to join because the transactions costs of contract). Thus, if a company in a country other dealing with the GBS would be lower than those than the host country makes unremunerated use of associated with national bureaucracies administer- genetic resources, the owner of the resource will ing national laws. have no legal recourse, and biopiracy will prevail. Moving along the spectrum of the pure contrac- How Much Will the Effort Be Worth? tual route, but in the direction of strengthening it, would be the suggestions made by the World Wide A key question is, what economic benefits will, real- Fund for Nature. These would require that patent istically, flow to developing countries for the genetic applications in the field of biotechnology (a) dis- resources that they possess? The initial expectations close the country of origin of biological samples sparked by the early contracts between INBio and used in the research leading to the invention that is Merck have been to some extent belied. There is the subject of the application and (b) include a skepticism about the likely economic benefits, stem- statement of compliance with all national laws ming both from theoretical considerations and relating to access to genetic resources and to any from experiences with bioprospecting contracts. bioprospecting arrangements entered into by the Simpson, Sedjo, and Reid (1996) argue strongly prospective patentee. Furthermore, countries of ori- that the marginal value of genetic resources is likely gin and local communities would have an opportu- to be small because of the real possibility of redun- nity to oppose the grant of a patent where there is a dancy in discoveries and because identical drugs or perception of noncompliance with national laws. drugs with similar clinical properties can be isolated One could go even further and require that the from different species. The number of presently TRIPS agreement incorporate an obligation on all untested species is high, and no one tree or genetic countries that the prospective patentee obtain the resource is likely to be the unique repository of the prior informed consent of the country from which valuable code: if one tree holds the potential for an genetic resources are obtained as a precondition for important cure, another tree that evolved in the grant of the patent (or at least that such consent response to the same environmental stresses is likely be obtained before the commercialization of the to have similar properties. patent).13 Stone (1998) asserts that the contentiousness over the equitable sharing of the fruits of genetic resources is almost certainly disproportionate to the Option III: A Global Biocollection Society likely booty. Estimates of the value of an average Yet another form of international cooperation that species vary widely, from US$44.0 to US$24 million. is a halfway house between the pure contractual Simpson, Sedjo, and Reid (1996) calculate that the approach and the full-fledged proprietary protec- median value of the world's top 18 biodiversity hot tion route is a scheme proposed by Drahos (2000) spots is about US$2 per hectare, although for some to establish a global biocollection society (GBS), areas of western Ecuador it might be US$20 per possibly under the auspices of the World Bank. hectare. Others, using a different methodology, con- Membership would be open on a voluntary basis to clude that the world's 3 billion hectares of tropical companies and to groups with claims to traditional forest are worth, on average, US$0.9 to US$1.32 per knowledge and genetic resources. The GBS would hectare as pharmaceutical mines (Mendelsohn and act as the repository for community registers of Balick 1995). In the case of the well-publicized indigenous knowledge and as the custodian of these agreement between Merck and INBio, the contract 387 T E C H N O L O G Y A N D I N T E L L E C T U A L P R O P E R T Y involved the payment of a fixed sum of US$1 million of supporting the conservation and sustainable use plus a contingent fee--a royalty--estimated at of genetic resources, thereby benefiting farming and between 2 and 3 percent of the value of any eventual other rural communities.15 discovery. There has been no real rush of pharma- Developing countries must demonstrate that ceutical firms to mimic the Merck-INBio agreement, their actions at the national level to protect genetic which is not known to have generated anything for resources and traditional knowledge are workable Merck to date. The jury thus seems to be out on the and that they do indeed lead to flows of resources to practical value that developing countries can expect individuals and communities that serve to increase to appropriate from possessing genetic resources. the incentives to protect these resources. This will Estimates of the likely benefits will clearly have to strengthen developing countries' case for seeking to condition developing countries' stance in any mul- replicate internationally their systems of domestic tilateral negotiations on protection of genetic protection. resources. In any future bargaining a hard-headed Finally, work needs to be intensified in national assessment of the likely gains in this area should and international fora, including the World Intellec- inform what developing countries are willing to tual Property Organization, on resolving the diffi- give up in such areas as other IPR issues. For exam- cult legal issues concerning the creation of a ple, Cottier (1998) implicitly argues that industrial proprietary system for protection of traditional countries' willingness to accept developing country knowledge and genetic resources. demands on genetic resources should in turn induce developing countries to reconsider their Notes opposition to extending patent protection to 1 South America is reported to have the largest concentration of biotechnological inventions.14 Whether this is a plant species: Colombia, Peru, and Ecuador between them bargain worth making remains to be seen, and more have 40,000 species on just 2 percent of the Earth's surface. research on the economic value of genetic resources 2 The contract also included a contingency or royalty payment needs to underpin policy positions. that would be triggered in the event of commercially prof- itable exploitation of the resources provided. There have been similar contracts between the National Cancer Institute in the The Agenda for Developing Countries United States and organizations in Madagascar, the Philip- Irrespective of the form that international coopera- pines, Tanzania, and Zimbabwe for provision of samples. INBio has also signed agreements with a major U.K. technology tion takes, developing countries that are hosts to licensing firm, the British Technology Group, to isolate and test genetic resources have to strengthen their own legal a compound that could potentially eliminate a nematode and institutional frameworks for protecting tradi- which plagues banana and other crops, and with Cornell Uni- tional knowledge and genetic resources. (The CBD versity and the Bristol-Myers Squibb Co. to test and identify requires countries to protect biodiversity.) The first chemical compounds developed from insect extracts for use in treating malaria. step, to enact legislation, is being undertaken in a number of countries (see ). The 3 Throughout this chapter the terms "genetic resources" and next step would be to create a registration system "biodiversity" are used interchangeably, although the two are not formally identical. Biodiversity is a measure of the genetic for compiling all forms of knowledge and resources variation contained within the Earth's biological resources that could, in principle, be protected. (plants, animals, fish, bacteria, insects, and so on). It increases This compilation could proceed at both the with the number of distinct species. Many species have critical national and international levels. INBio has estab- masses or thresholds: once their population falls below a criti- lished an elaborate biodiversity inventory--through cal number, they are doomed to extinction. Biodiversity is thus the development and management of biological, not related simply to the quantity of biological resources but also to their distribution in relation to the threshold levels. ecological, taxonomic, and related systematic infor- Beyond these levels, additional quantities of resources con- mation on living species and systems found in Costa tribute little to biodiversity (Heal 1998). Rica (Sittenfield and Lovejoy 1998). This informa- 4 The neem tree, traditionally used in India for pest control, tion lists not only the species but also where they medical therapies, and other purposes, was the subject of a may be collected without damaging ecosystems. patent by a chemical company. Cottier (1998) describes the India has pioneered legislation for the registration case of the collection in Kenya by Western companies of bark of traditional knowledge and the establishment of a and leaves from the Acacia nilotica tree and the subsequent use National Community Gene Fund that has the goal of these materials in the development of patentable products. 388 Proprietary Protection of Genetic Resources and Traditional Knowledge 5 The African group of countries also proposed the inclusion of voluntarily contribute. To date, not much by way of contribution this issue in the next round of trade negotiations to this fund has been forthcoming. (WT/GC/W/302). 10 The CBD grandfathers seed collections that were already in 6 Several examples of socially useful practices have been cata- existence before 1994. logued in the literature. Among them is the practice of cutting 30-to-40-day-old sorghum or calatropis plants and placing 11 The one exception is in the field of undisclosed information or them in irrigation channels to control termite attack in light, trade secrets, where requirement of novelty is not a precondi- dry soils (Gupta 1998). tion for obtaining protection. A common approach in the past has been for environmental nongovernmental organizations 7 Artemisin, the only drug effective against all strains of malaria, and governments of rich countries to persuade developing was discovered because the Chinese use it to treat fever. Taxol, country governments to protect habitats, including tropical derived from western yew, has anticancer properties. Michel- forests, that are rich in genetic resources. There has been some lamine B, a novel compound from a vine found in Cameroon, success--governments have established parks and preserves-- has promise as an anticancer, anti-HIV drug. but this effort has been inadequate, and the protection is hap- 8 The ability of modern biotechnology to transfer genes across hazard (Sedjo 1992). organisms renders feasible the development of a wide array of 12 For example, the pilot phase of the Global Environment Facility genetically engineered plants and animals that have significant provided financial support for conservation in developing economic and therapeutic value. Thus, the returns to main- countries in return for actions by them to protect biodiversity taining a rich and biodiverse system are probably large that had national and global benefits. because improvements in biotechnology and natural genetic information will be complements in production. 13 The pharmaceutical industry is increasingly involved in the transfer of chemical extracts rather than samples of plants and 9 More than 130 countries have signed the convention. A animals. A strict reading of the CBD suggests that the require- notable exception is the United States, which had concerns ment of prior informed consent (in Art. 15.5) applies to genet- about the possible dilution of traditional intellectual property ic resources and not to chemical extracts, although the rights protection as an outcome of the CBD. convention allows countries to take measures to manage their The principle that genetic resources are the common heritage biological resources that would also apply to such extracts. of humankind was a central tenet of the International Undertak- ing on Plant Genetic Resources and of the Food and Agriculture 14 Article 27.3 of the TRIPS agreement allows countries to Commission on Plant Genetic Resources, both established in the exclude most biotechnological inventions from patent protec- 1980s. Under this regime, countries that restricted access to tion. This article is currently under review. Some industrial such resources were criticized, as happened in the case of countries are keen to see the exception narrowed, which Ethiopia, which in 1977 placed an embargo on the export of its would force developing countries to extend the scope of pro- coffee germ plasm. Although the intent of these initiatives was tection for most biotechnological inventions. to ensure widespread availability of genetic resources for agricul- ture and industry, it gave the source country little benefit for its 15 At the international level, a Clearing House Mechanism under use and hence provided little incentive for conservation. These the CBD and the System-wide Information Network on Genet- international agreements on plant genetic resources embodied ic Resources within the centers of the Consultative Group on the notion of farmers' rights arising from their past, present, and International Agricultural Research provide a framework for future efforts to conserve and improve plant genetic resources. data collection and exchange of information. In principle, They envisaged the right to equitable compensation for their these networks could become nuclei for the international reg- use that would come out of a fund to which countries would istration of plant genetic rights. 389 38 F R A N K J . P E N N A C O E N R A A D J . V I S S E R Cultural often yielded only limited finan- cial benefits for creative artists in developing countries. For exam- Industries and ple, arguably the most popular song ever to emerge from Africa Intellectual was "The Lion Sleeps Tonight." Composed and recorded in 1939 Property Rights by Solomon Linda, a Zulu singer who had hunted lions as a boy, the song was an attempt to translate into English a tradi- tional lion-hunting song, "Mbube." On his third take in a T recording studio, Linda came up he economic argument underlying with the memorable lines "In the Jungle, the mighty this chapter is that various cultural jungle, the lion sleeps tonight . . ." The studio industries--music, crafts, and ethnobotanicals, for obtained copyright in the song and sold it to a vari- example--can be sources of wealth for developing ety of record companies. In the 1960s a folk group countries, just like extractive industries such as recorded the song as "Wimoweh." Most recently, the gold, diamonds, and oil. (Even the United States has song was used in the Disney film The Lion King. entertainment as its largest export, with defense Through various arbitration hearings, lawyers for exports its second largest.) Moreover, cultural the studios, the record companies, and the Disney industries distribute such wealth more equitably film producer have estimated the total royalty and are less prone to capture by elites. This chapter income of the song over the last 60 years to be describes some of these cultural industries and dis- between US$10 million and US$20 million. Yet cusses related intellectual property rights issues that Solomon Linda was paid 1 pound in cash and had a must be addressed in order for developing countries menial job of stacking records for the rest of his life. to assert their interests. The discussion draws on the He died in 1962, but only in 1980 was his wife able experience of some of the projects in this area that to afford a headstone for his grave (Malan 2000). are managed by the Policy Sciences Center. In spite of the significant unrealized market potential for music exports from the developing world, the initial experience of the Policy Sciences The Music Industry Center in least-developed African countries sug- Worldwide sales of music recordings are estimated gests that there often seems to be an "iron triangle" at US$40 billion annually. "World Music" makes up composed of local elite musicians, the managers of 1 percent of such sales, of which about half is collection societies, and (foreign) multinational accounted for by African popular music (Collins record companies. The way in which this triangle 2000).1 This significant international market has operates mirrors Solomon Linda's experience. 390 Cultural Industries and Intellectual Property Rights Local, poor composers and performers are paid for authorize any reproduction of the work, the right to their creative input on a "work for hire" basis, usual- broadcast the work or to perform it in public, and ly receiving low flat fees. They sell their composi- the right to make an adaptation of the work (which tions and performances to elite musicians in their includes translating it). Two moral rights are pro- own country who have their own recording studios. tected: the right to claim authorship of the work (the These musicians then sell the songs on the interna- "paternity" right), and the right to object to any dis- tional market through either their own record labels tortion, mutilation, or other modification of, or or foreign multinational record companies. The col- other derogatory action in relation to, the work that lection societies that are supposed to collect royal- would prejudice the author's honor or reputation ties for composers barely perform this function. (the "integrity" right). Royalties for performers are rarely collected, unlike With musical expressions of folklore, the position the case in Europe or the United States. is more complex. Many developing countries regu- Some of the antidotes to this situation include late the use of expressions of folklore within the educating poor composers and musicians about framework of their copyright laws (Ficsor 1997). their rights (including their right to royalties) and These countries do so by taking advantage, express- about standard-form contracts and contract terms, ly or by implication, of a special provision in the fair pricing means, and the need to form unions--in Berne Convention stating that, for unpublished short, to build "countervailing constituencies of the works where the identity of an author is unknown poor." In addition, lawyers can be trained in intellec- but there is "every ground" to presume that he or tual property law in developing countries, and legal she is a national of a particular country, legislation assistance can be provided to poor musicians. in that country may designate the competent Intellectual property issues in the music industry authority to represent the author and to protect and have to do with copyright and related rights, espe- enforce his or her rights (Art. 15.4.a).2 Thus, expres- cially those of performers. A crucial distinction sions of folklore are assimilated to original literary should be made between recordings of perfor- and artistic works, so that the economic rights in mances of original music, on the one hand, and per- respect of such expressions can be exercised by the formances of musical expressions of folklore, on the designated authority.3 other. Performances of original music bring copy- Expressions of folklore, however, fit uncomfort- right and related rights (those of performers, pro- ably into the copyright paradigm. For one thing, ducers of sound recordings, and broadcasters) into they are often the result of a continuing and slow play, whereas performances of musical expressions process of creative activity exercised by a certain of folklore are almost completely lacking in protec- community by consecutive imitation, whereas tion in terms of both content and performance, works protected by copyright should show some especially at the international level. individual originality. In short, copyright is author- centric, while with expressions of folklore, any notion of an author in the copyright sense is gener- Copyright ally absent (Ficsor 1997).4 Moreover, as the term of An original musical composition is covered by copy- copyright protection is usually determined with ref- right, without any need for registration or compli- erence to an identifiable author, the lack of such an ance with any formality. Such informal acquisition author in the folklore context again makes folklore a of copyright is mandated by Article 5.2 of the Berne square peg in the copyright round hole.5 Folklore Convention for the Protection of Literary and Artis- has evolved over centuries and continues to do so. tic Works, most of the substantive provisions of Any notion of a fixed term of protection in respect which have been incorporated by reference into the of folklore denies this essential feature. TRIPS agreement (Art. 9.1). Where the musical composition consists of words and music, copyright Folklore: Paying Public Domain in the words, as a literary work, exists separately from copyright in the musical work (the notes). The Adolf Dietz has proposed the payment of remuner- Berne Convention extends, to authors of original ation for the use of works and performances in the works, economic (exploitation) and moral rights. public domain through the creation of a communi- The economic rights comprise the exclusive right to ty right of authors and performers (Dietz 2000). 391 T E C H N O L O G Y A N D I N T E L L E C T U A L P R O P E R T Y The underlying notion is that the community of liv- These issues are addressed in the WIPO Copy- ing authors and performers should benefit from the right Treaty (WCT), adopted on December 20, use of works and performers of their predecessors 1996, at the Diplomatic Conference on Certain that are no longer protected because the term of Copyright and Neighboring Rights Questions protection has expired. Such a remuneration right under the auspices of the World Intellectual Proper- can be established by legislation in favor of an ty Organization (WIPO). The WCT confirms an authors' and performers' fund administered by a author's exclusive right, in the digital environment, foundation or nonprofit corporation. The entity to reproduce his or her work in any manner or form should be largely managed and administered by the (Agreed Statement Concerning Article 1.4).8 It is authors' and performers' organizations themselves. also understood that the storage of a protected work Collecting societies, where they exist, can collect the in digital form in an electronic medium constitutes remuneration in the same way as they do for the use a reproduction of that work. The WCT grants an of protected works and performances. The money exclusive right to authors to authorize making their would not be distributed according to the individ- works available through interactive, on-demand ual distribution schemes but would be forwarded to services (Art. 8). The relevant act of exploitation is the foundation or corporation. For developing making the work available to the public: the act countries, this is an attractive proposal.6 Of course, "commences, and is completed by providing public nothing prevents the extension of this proposal to access to the work" (Hugenholtz 2000). For the first include expressions of folklore, which would then time in an international instrument, the WCT rec- attract a similar right of remuneration. ognizes that in a digital environment any new rights with respect to digital uses of works would, for the rights to be effective, require the framework support The Digital Environment: of provisions dealing with technical measures of The WIPO Copyright Treaty protection and electronic rights management infor- The emergence of global information networks like mation. To this end, the WCT obliges contracting the Internet and electronic commerce raises a num- parties to provide adequate legal protection and ber of key issues in the field of copyright: effective remedies against the circumvention of measures protecting the rights of authors (Art. 11), · The use of computers requires that works be and to provide, under certain conditions, adequate transformed from their traditional material form remedies against the removal or alteration of elec- into digital form. Digitization has two main tronic rights management information (Art. 12).9 advantages: transmission of a digitized work occurs without any degradation (every copy is Related Rights: Performers perfect), and copies of such a work can be made quickly and cheaply. Unfortunately, these advan- Since 1961, performers of musical works have been tages also mean that copyright may be infringed protected within the framework of the International with ease and on a scale previously unknown. Convention for the Protection of Performers, Pro- · Material stored or made available for access on ducers of Phonograms and Broadcasting Organiza- hosts or transmitted through the Internet may be tions (the Rome Convention). For historical reasons, the subject of copyright owned by a third party this protection was weak; performers do not have who has not consented to these activities. exclusive rights but should merely be able to prevent · To protect their works against these first two risks, certain acts from being performed in respect to their authors have often resorted to technical protec- recorded performances or should be able to prevent tion measures. These measures usually operate at their live performances from being recorded or one of two levels: access control and copy con- broadcast. The TRIPS agreement similarly states that trol.7 They remain effective, of course, only if performers should "have the possibility of prevent- their unauthorized circumvention is prohibited. ing" a limited number of acts (Art. 14.1). · With works in digital form, it is easy to remove Like the WCT, the WIPO Performers and Phono- any rights management information. If this is grams Treaty (WPPT) addresses issues relating to the done, it may become difficult to prove copyright protection of performers' rights in the digital context, ownership. and it greatly enhances their position. For the first 392 Cultural Industries and Intellectual Property Rights time, it grants performers certain exclusive rights of has estimated that it generates US$30 billion annu- authorization in respect to their live and recorded ally worldwide, and some analysts have estimated performances. For their live performances, perform- the amount as double that. A study commissioned ers have the exclusive right to authorize the broadcast- by the Policy Sciences Center of the crafts industry ing and communication to the public of such in India, the largest such industry in the world, esti- performances and their recording (fixation) (Art. 6). mated the annual revenue of this industry at US$5.6 For their recorded performances, performers have the billion. It employs 9 million to 10 million artisans exclusive right to authorize the reproduction of these (Liebl and Roy 2000). performances (Art. 7), their distribution (Art. 8), their As in the music industry, the absence of effective rental (Art. 9), and their being made available so that enforcement of property rights creates significant "members of the public may access them from a place potential opportunity costs for developing coun- and at a time individually chosen by them" (Art. tries and indigenous peoples. The experience of the 10).10 Performers are also entitled to remuneration Kwakiutl Indians of western Canada is an excellent for broadcasting and communication to the public of example (Chartrand 1996). For over a century, commercial recordings of their performances (Art. Kwakiutl women have knitted sweaters with a tradi- 15) and (another first) are given moral rights on sim- tional thunderbird motif. In the mid-1980s two ilar terms to those extended to authors (Art. 5)--that Japanese businessmen visited the Indian reserva- is, paternity and integrity rights. tion, purchased some sweaters, and mass-produced Developing countries seeking international protec- copies for sale in Asia. Apparently, US$100 million tion of expressions of folklore can benefit from an worth of sweaters was sold. The Kwakiutl were out- important step forward in the WPPT. Unlike the raged; all they had received was payment for a hand- Rome Convention, which limits the definition of ful of sweaters. This incident fueled a movement for "performers" to those who perform "literary or artis- the protection of the rights of indigenous peoples in tic works" (Art. 3), the WPPT extends the definition Canada. to apply also to those who perform "expressions of A preferable alternative scenario would have been folklore"(Art. 2). Certain expressions of folklore, such for the Kwakiutl to have registered the thunderbird as folk tales, folk poetry, folk songs, instrumental folk design and negotiated a licensing agreement with music, folk dances, and folk plays, live through per- the Japanese businessmen. Under such a license, formance. To the extent that these performances are they could have been paid royalties that probably protected against unauthorized recording and broad- would have been calculated as 20 percent of gross casting and communication to the public, the expres- income, or US$20 million for the tribe. This is a sions of folklore being performed are indirectly good example of how, by asserting the intellectual protected. This is a fairly efficient means for their property rights of indigenous peoples, multination- indirect protection (Ficsor 1997). al corporations can become "partners" instead of The WPPT grants coextensive rights to the pro- "predators." ducers of sound recordings (Arts. 11­15). Obviously, There are limitations to copyright protection in they do not enjoy any moral rights in respect of their this context, as is illustrated by two Australian cases. recordings. If the balance between authors (including In Yumbulul v. Reserve Bank of Australia (1991, 2 composers), performers, and producers of sound I.P.R. 481), representatives of the Galpu clan unsuc- recordings in the WCT and the WPPT is maintained cessfully sought to prevent the reproduction by the and is not disturbed by the existence of "iron trian- Reserve Bank of the design of a Morning Star pole gles" of the sort described above, they could be used on a commemorative banknote. A member of the to make multinational record companies into "part- clan had created the pole, having obtained the nec- ners and not predators,"helping foreign direct invest- essary knowledge and authority through initiation ment, as well as domestic investment by local and revelatory ceremonies. The clan claimed that he musicians, to contribute toward alleviating poverty. owed the clan a communal obligation to prevent the design of the pole from being used in a manner that was culturally offensive. The trial judge ruled that The Crafts Industry the artist had successfully disposed of his intellectu- The importance of the crafts industry is often trivi- al property right through a binding agreement. The alized, but the U.S. International Trade Commission judge did lament that "Australia's copyright law 393 T E C H N O L O G Y A N D I N T E L L E C T U A L P R O P E R T Y does not provide adequate recognition of Aborigi- will be referred to as a certified Indigenous creator. nal community claims to regulate the reproduction The second mark--the collaboration mark--will be and use of works which are essentially communal in applied to "products or services derived from a work origin" (p. 490). of art which has been created by an Aboriginal or Milpurrurru v. Indofurn Pty Ltd (1995, A.I.P.C. 91) Torres Strait person or people who satisfy the defini- concerned the importation by a company based in tion of `authenticity'" (NIAAA 1997). This mark rec- Perth of carpets made in Vietnam. The carpets ognizes that products and services are often reproduced designs of certain Aboriginal artists, a produced, reproduced, or manufactured under portfolio of whose work had been produced by the licensing agreements with indigenous people. The Australian National Gallery. The court awarded collaboration mark will be applied to such products substantial damages for copyright infringement to and services, provided that the licensing arrange- the Aboriginal artists whose designs had been ments are "fair and legitimate."13 reproduced on the carpets, but it did not find itself Although such labels of authenticity will raise the able to compensate the communities whose images profile of indigenous artists and help to make sure were used in culturally inappropriate ways. The rul- that they are properly remunerated, they provide ing noted that "the statutory remedies do not recog- only limited protection to these artists. It is unlikely nize the infringement of ownership rights of the that, by themselves, they will prevent the production, kind which reside under Aboriginal law in the tradi- import, or export of forgeries (Wiseman 2001). Fur- tional owners of the dreaming stories" (p. 39081). thermore, since the marks are registered in a national Thus, copyright law does not recognize the com- registry, their effect is limited to that national territo- munal harm that may result from the unauthorized ry except in the rare situation in which they qualify reproduction of indigenous designs and, according- for protection as well-known marks.14 ly, does not compensate communal harm (Blakeney 1998, 2000). More satisfactory answers, albeit to a Trade Dress limited extent only, can be given to the question as to whether legal protection can be established in At a recent workshop hosted by the World Bank on respect of statements such as "handmade in India," the crafts industry in India, Jerome Reichman sug- "Kente weaving from Ghana," or "Mola designs gested using trade dress protection for indigenous from Panama." crafts (Reichman 2001). This type of protection relates to product packaging (Dinwoodie 1997): if such packaging is inherently distinctive, it qualifies Certification for trademark protection, (potentially) forever. In Australia the preferred legal technique for protec- Reichman counseled: tion against nonindigenous people who manufacture and sell indigenous artifacts at the expense of the Make a fancy package. Make not just the India indigenous artistic community is the use of certifica- mark, but the way that it comes in, per compa- tion marks that serve as labels of authenticity (Wise- ny and per product and per region. Those are man 2001).11 The National Indigenous Arts strongly protectable in national law and in Advocacy Association (NIAAA) recently registered international law under the TRIPS Agreement, the first of two proposed national indigenous labels which requires all sorts of international trade- of authenticity as certification marks in Australia. mark protections and is very strong now. The labels, to be applied to goods and services of Aboriginal or Torres Strait Islander origin, will make Again, the problem is that the protection is terri- it more difficult for non-Aboriginal people to pass off torial only, based on national legislation, with the their works as authentically Aboriginal.12 The first only possible exception being for well-known mark--the label of authenticity--will be applied to marks. "products or services that are derived from a work of art created by, and reproduced or manufactured by Unfair Competition: Misleading the Public Aboriginal or Torres Strait Islander people who satis- fy the definition of `authenticity'" (NIAAA 1997). An Using misleading indications of origin may consti- artist who has successfully applied to use this label tute unfair competition. The Paris Convention, for 394 Cultural Industries and Intellectual Property Rights example, requires countries to grant protection can take action against the use of a false indication against "indications or allegations the use of which (Art. 102). in the course of trade is liable to mislead the public The Madrid Agreement for the Repression of as to the nature, the manufacturing process, the False or Deceptive Indications of Source on Goods characteristics, the suitability for their purpose, or extends this protection to "deceptive" indications of the quantity, of the goods" (Art. 10 bis [3]3). The source. Although a deceptive indication may literal- WIPO Model Provisions on Protection against ly be true, it is still misleading. For example, when Unfair Competition are more explicit; they state two geographic areas in different countries have the that any act or practice "in the course of industrial same name but only one of them has been identi- or commercial activities, that misleads, or is likely to fied with the source of certain products, and when mislead, the public with respect to an enterprise or the name is then used for goods originating in the its activities, in particular, the products or services other area in a way that leads members of the public offered by such enterprise" constitutes unfair com- to believe that they originate in the first area, such petition (Art. 4.1). "Misleading may arise out of use is deceptive; the public believes that the prod- advertising or promotion and may, in particular, ucts originate in the geographic area with respect to occur with respect to . . . the geographical origin of which the indication has traditionally been used products or services" (Art. 4.2). (Baeumer 1997). Here, too, the main problem with protection The Lisbon Agreement for the Protection of against unfair competition is that it is confined to Appellations of Origin and Their International Reg- the national level and thus differs from country to istration provides strong protection for certain geo- country. Although this kind of protection may pro- graphical indications, called "appellations of origin." tect indigenous artists against misappropriation That term connotes "the geographical name of a within their national states, it offers no protection at country, region, or locality, which serves to designate the international level. a product originating therein, the quality or charac- teristics of which are due exclusively and essentially to the geographical environment, including natural Geographical Indications and human factors" (Art. 2.1). This protection is A major intellectual property issue for developing based on an international registration of an appella- countries is the unauthorized use of a geographical tion of origin with the WIPO. The main factor limit- indication for noncompeting products, such as "Taj ing the scope of application of this agreement is the Mahal" for a hotel in Nevada. This is a complex requirement that an appellation of origin be protect- issue. It "is not necessarily a question of misleading ed as such in its country of origin before it can be consumers" but "also a question of the reputation of registered as such with the WIPO (Art. 1.2). the geographical indication"(Baeumer 1989). At the The WTO TRIPS agreement defines "geographi- same time, the protection should not overreach; it cal indications" as "indications which identify a has been argued that although the protection of good as originating in the territory of a Member [of geographical indications against unauthorized use the WTO], or a region or locality in that territory, on all types of goods would be excessive, protection where a given quality, reputation or other character- should be given against "a blatant misuse of reputa- istic of the good is essentially attributable to its geo- tion" (Bienaymé 1989). graphical origin" (Art. 22.1).16 The scope of The Paris Convention states that goods for which protection expressly extends to deceptive indica- a false indication of source is used should be seized tions within the meaning of that term in the Madrid on importation (Art. 9.1) or seized in the country Agreement (Art. 22.4). Protection should be avail- into which they are imported, if the false indication able against misleading use of a geographical indi- had been applied in that country (Art. 9.2); should cation and against acts of unfair competition (Art. be barred from importation (Art. 9.5); or should be 22.2) and against the registration of a trademark subject to such other actions and remedies as are that contains, or consists of, a geographical indica- available to nationals of the country in question tion relating to goods not originating in the territo- (Art. 9.6).15 Any producer or manufacturer ry indicated, if use of such a trademark is of such a engaged in the production or manufacture of the nature as to mislead the public about the true place goods to which the geographical indication refers of origin (Art 22.3). Geographical indications in 395 T E C H N O L O G Y A N D I N T E L L E C T U A L P R O P E R T Y respect to wines and spirits enjoy additional protec- industrial property rights of indigenous peoples tion (Art. 23). Members of the WTO have agreed to with respect to their traditional knowledge should enter into negotiations to raise the level of protec- involve, first, protection against acquisition of tion for individual geographical indications (Art. industrial property rights by "outsiders" as the 24). Some developing countries have argued in the result of an appropriation of traditional knowledge TRIPS context that the work mandated in respect to and, second, protection under industrial property of the establishment of a notification and registration the traditional knowledge of indigenous peoples for system of geographical indications for wines should their own benefit. be extended to other products recognizable by their geographical origins, such as handicrafts and agro- Protection against Exploitation of food products.17 Industrial Property Where protection against exploitation of industrial Ethnobotanicals property is concerned, two complementary Genetic resources and traditional knowledge are approaches can be taken. The first is to establish a another important potential source of revenue for notification requirement for patentability. For developing countries and indigenous peoples, and example, at the third session of the Standing Com- this is an area in which intellectual property protec- mittee on the Law of Patents (SCP) of the WIPO, tion can play an important role. Two examples are the Colombian delegation proposed the inclusion illustrative. of a provision to this effect in the (then draft) Patent Law Treaty. The proposal stated that where it Neem. For hundreds of years, people in India have appears that an invention which forms the subject made use of a natural fungicide in the bark of the of a patent application is based on genetic resources neem tree. Recently, a pharmaceutical company that are part of a country's "biological and genetic went to India, identified the active ingredient in the heritage," a copy of the contract affording access to bark, patented that active ingredient, and then the genetic resources in the country of origin offered it for sale to Indians. This caused an uproar should be filed (SCP/3/10). Unfortunately, the pro- in the country. The registration of the patent for posal was rejected for relating to a substantive neem in the European Patent Office was challenged, requirement of patentability and so not falling and the Technical Board of Appeal eventually found within the scope of the treaty, which is concerned in favor of the applicants for revocation of the with formal requirements (procedures and docu- patent. The board ruled that since the properties of mentation) only. The prospects of a similar propos- neem tree bark had been known for many years in al succeeding at an international trade negotiation India, the patented invention did not satisfy the round are slim, given the strong opposition to such absolute novelty requirement of the European a requirement by pharmaceutical companies and Patent Convention.18 the governments of countries with strong pharma- ceutical industries. Lycopene. For centuries, the Amazonian Indians in The second approach is to prevent the unautho- Ecuador have used tamate, a small, cylindrical rized (improper) acquisition of industrial property tomato found in the jungle, for its cancer-fighting rights (especially patents) over traditional knowl- properties. A multinational pharmaceutical compa- edge by documenting and publishing traditional ny went to Ecuador, isolated the active ingredient, knowledge as searchable prior art, should the hold- lycopene, patented it, and now sells it as one of the ers of the traditional knowledge concerned want cutting-edge products in cancer treatment. this. (An example of such an approach is the World In neither case did the country or its indigenous Bank's Indigenous Knowledge Program.) Once such peoples receive any benefit from what should have knowledge becomes part of the prior art, that mere been their industrial property rights to these items fact destroys the novelty of any invention based on of traditional knowledge. (The phrase "industrial such knowledge. Even if a patent is obtained, it may property" includes patents, utility models, "petty be revoked on this ground. This procedure may patents," trade secrets, and the like.) A closer look at involve an application launched by the holders con- the two examples shows that any protection of the cerned (which would involve substantial legal costs) 396 Cultural Industries and Intellectual Property Rights or by a rival pharmaceutical company that wants to other things, for an undertaking of confidentiality exploit the knowledge for its own gain, and at its and for remuneration (usually, a royalty). The Poli- own cost. (See the further discussion in the next cy Sciences Center is experimenting with a trade section.) secret approach for communities to use so that they can derive revenue from ethnobotanical knowledge. For example, the center has made a grant to the Protection for Exploitation of Industrial Property nongovernmental organization Otro Futuro in Economically, the main aim of the second aspect of Venezuela to assist it in helping the Dhekuana Indi- protection is to secure revenue for indigenous peo- ans develop an archive and atlas and to protect their ples through the exploitation of their ethnobotanical intellectual property rights in resources ranging knowledge. Again, two approaches are possible: from myths, stories, legends, and music that can be through patents and through transfer of technology. copyrighted to ethnobotanicals that can be patent- ed. The problem is that up until now the Dhekuana Patents. Holders of traditional knowledge can be had perceived property in all these categories as given access to the industrial property system to being communally owned by 12 tribes inhabiting enable them to obtain patents (or utility models or about 2 million square acres. The experiment "petty patents" where provision is made for these), involves establishing a community foundation in where appropriate. A basic problem with doing so is which the tribes would be represented on the board that a patent protects active ingredients that have of directors and in which the property rights for been isolated and tested. Such isolation and testing copyright and patent could be vested. (Another may cost hundreds of millions of dollars and are out option is to establish a for-profit corporation; see of reach for most developing countries, let alone Norchi 2000.) Ethnobotanical knowledge would be their indigenous peoples. Another problem is that it protected by being treated as a trade secret by the may not be possible to obtain a patent because the community foundation, not to be disclosed to a novelty of the invention may have been destroyed pharmaceutical company or others unless such by the local community's own prior use of the "outsiders" agree to pay royalties to the foundation. invention.19 This approach is not free from pitfalls, either. Trade secret protection usually depends on the legal Transfer of Technology. A variety of approaches to rules of each country, and international attempts at transfer of technology can be considered. For these harmonization have not been very successful. The to be effective, there must be an organized body of TRIPS agreement, for example, simply states, "Nat- knowledge and an identifiable entity to administer ural and legal persons shall have the possibility of transfers. One approach is illustrated by the contract preventing information lawfully within their control signed in 1991 between Merck & Co. and the non- from being disclosed to, acquired by, or used by oth- profit organization Instituto Nacional de Biodiversi- ers without their consent in a manner contrary to dad (INBio) in Costa Rica. Under the terms of the honest commercial practices" (Art. 39.2).20 The pro- agreement, Merck received about 10,000 plant sam- tected information should be secret in the sense that ples over a two-year period and was supplied with it is not, as a body or in the precise configuration and information about their traditional use. Merck paid assembly of its components, generally known a reported US$1.35 million to INBio for the samples among or readily accessible to persons within the and agreed to pay a royalty ranging between 2 and 3 circles that normally deal with the kind of informa- percent (Blakeney 2000). If one of the samples tion in question; that it has commercial value becomes a billion dollar drug, this implies US$20 because it is secret; and that the person lawfully in million to US$30 million in royalties. control of the information has taken reasonable A problem with this approach is that if the royal- steps, under the circumstances, to keep it secret. A ties are paid to an official body, they may disappear problem with ethnobotanical knowledge is that into the general state revenue account and not ben- often the steps to keep the information secret may efit the relevant communities or individuals. An not be sufficient under existing common or civil law alternative approach relies on the law relating to the rules; secrecy often flows only from the fact that, protection of trade secrets: the trade secret is dis- because of customary law and practices, few people closed (licensed) to someone in exchange, among have access to the information (Gervais 2001). 397 T E C H N O L O G Y A N D I N T E L L E C T U A L P R O P E R T Y The WIPO Model Provisions on Protection barriers to entry into insuperable legal barriers to against Unfair Competition do not take the matter entry (Reichman and Samuelson 1997). Economic much further. A proposal relating to unfair competi- efficiency, by contrast, demands very low prices for tion in respect to secret information simply states, such use and favors minimum incentives to provide "Any act or practice, in the course of industrial or the needed investment and services. commercial activities, that results in the disclosure, An alternative approach favors a weak intellectual acquisition or use by others of secret information property right to overcome the risk of market failure without the consent of the person lawfully in control without creating legal barriers to entry.23 A modest of that information . . . and in a manner contrary to version of this approach calls for a misappropriation honest commercial practices shall constitute an act model based on simple unfair-competition princi- of unfair competition" (Art. 6.1).21 Examples of ples. Courts could use market-oriented factors to unfair competition include secret information determine whether there had been an "unfair extrac- acquired in breach of contract or of confidence (Art. tion" from a database. Useful factors include the 6.2). The same conditions as in the TRIPS agree- extent of the data appropriation by the user; the ment have to be satisfied in order for information to nature of the data appropriated; the purpose for qualify as "secret" (Art 6.3). In determining whether which the user appropriated the data; the degree of reasonable steps have been taken to keep the infor- investment initially required to bring the data into mation secret, account is to be taken of the amount being; the degree of dependence, or independence, of effort and money spent by the rightful holder on of the user's own development effort and how sub- developing the secret information; the value of the stantial the user's own investment in such effort has information to that person and to the person's com- been; the degree of similarity between the contents petitors; the extent of the measures taken by the of the database and a product developed by the user; rightful holder to keep the information secret; and the proximity or remoteness of the markets in which the ease or difficulty with which it could be lawfully the database maker and the user are operating; and acquired by others (Art. 6, note 6.20).22 Also, the how quickly the user was able to come to the market secret information has to be identifiable (for exam- with his or her own product compared with the time ple, in documents or through storage in a database). required to develop the original database (Reichman Thus, to protect ethnobotanical knowledge not and Samuelson 1997). Obviously, any such protec- only in the country of origin but also in foreign tion has to be balanced by limitations and exceptions countries, the legal rules relating to trade secret pro- favoring science and education. tection may have to be reviewed and strengthened Of these two approaches, the second is of more internationally. The fact that the secret information benefit to developing countries. Although it allows has to be identifiable (in this context, usually in a makers of databases to be protected against the risk database such as the archive and atlas of the of market failure, it does not create real barriers to Dhekuana Indians) raises a further issue: the pro- market entry at the expense of the scientific and tection of nonoriginal compilations of data. Essen- education sectors. tially, there are two approaches to such protection. The first grants the maker of the database strong sui Compensatory Liability generis intellectual property protection in the form of exclusive rights. For example, under the terms of In a different context, Reichman has suggested a the European Database Directive (1996 O.J. [L 77]), "third intellectual property paradigm" loosely the maker of a database obtains an exclusive "right derived from classical trade secret law and from to prevent extraction and/or reutilization of the antitrust principles applicable to two-party trans- whole or of a substantial part . . . of the contents of fers of unpatented industrial know-how. The pro- that database" (Art. 7.1). This approach usually posed regime results in a rights regime of almost unlimited dura- tion, subject to few, if any, public policy limitations. aims to avoid market failure without introduc- For this reason it has been argued persuasively that ing the market distortions characteristic of this type of protection jeopardizes basic research, intellectual property rights and without forfeit- eliminates competition in the markets for value- ing the pro-competitive social benefits that added products and services, and converts existing result from trade secret laws under optimum 398 Cultural Industries and Intellectual Property Rights conditions. It solves the free-rider problem fac- excessive monopoly costs of intellectual property ing growing numbers of investors in applied rights, whereas weak protection gives rise to the know-how by directly linking the prospects for problem of excessive free-riding and to underin- short-term returns on investment to the stipu- vestment in innovation (Drahos 2001). The key is to lation of a standard, multi-party set of default find the correct balance. This observation also holds rules applicable to eligible forms of innovation. true on the international level: (Reichman 1994) Most states are in the position of being net The proposed "compensatory liability" regime is importers of intellectual property rights. Cer- inspired by the Italian neighboring right that pro- tainly all developing countries fall into this cat- tects engineering projects. Article 99 of the Italian egory. For countries that are importers of Copyright Law of 1941 gives authors of engineering intellectual property the temptation is not to projects, or other analogous productions, who con- recognize the intellectual property rights of for- tribute novel (but not obvious) solutions to techni- eigners, thereby allowing for the possibility that cal problems the right to a reasonable royalty from their nationals will be able to free-ride on the third parties who commercially exploit their techni- research and development activities of foreign- cal contributions without authorization. This right ers. For exporters of intellectual property rights to "equitable compensation" subsists for 20 years the aim is to extend the length and breadth of from registration. An appropriate notice must intellectual property rights in order to gain the appear on copies of the plans. maximum return from the trade in intellectual Reichman has argued that this regime could solve property rights and the goods and services to some pressing needs of developing countries: which they relate. (Drahos 2001) As with small-scale innovations, the goal is to Drahos asserts that democratic bargaining among reward both first comers (in this case, the rele- states will increase the likelihood of setting efficient vant indigenous community), and second com- international intellectual property norms. Democ- ers (those who build on the community's ratic bargaining requires representation, full infor- cultural heritage), without impeding access to mation, and nondomination. Representation the public domain or the flow of new products. requires that developing countries' interests be rep- With small amounts of tinkering, a compensa- resented at the negotiating table. It also implies that tory liability regime could be adapted to the representatives have some continuity of voice in encourage use of traditional knowledge with- the process, which in turn implies that exclusion out denying the relevant indigenous communi- should not be practiced. As far as full information is ties the right to a fair share of the proceeds. concerned, Drahos notes that during the TRIPS (Reichman 2000) negotiations developing countries were often not party to bilateral talks between the United States This regime could best be extended to ethno- and the European Union and so did not have access botanical knowledge. Legislation can allow "second to the same level of information. Moreover, all states comers" to exploit such knowledge commercially were ignorant about the likely effects of the TRIPS without prior authorization, subject to an obliga- agreement on information markets; the real-world tion to pay a reasonable royalty to a designated per- costs of extending intellectual property rights and son or institution. At the international level, the their effects on barriers to entry into markets were legal framework for this regime can be established not at all clear. As for nondomination, Drahos notes either by an express provision in a future trade the use by the United States of the "301 process" instrument or by incorporation within Article 10 prior to and during the Uruguay Round, as well as bis of the Paris Convention. bilateral agreements with developing countries to bring them closer to the U.S. position.24 In this sense, "TRIPs was less a negotiation and more a Possible Issues for Multilateral Negotiations `convergence of processes'" (Drahos 2001). Peter Drahos has argued that overly strong intellec- This chapter has addressed elements of the sec- tual property protection gives rise to the problem of ond requirement for democratic bargaining on 399 T E C H N O L O G Y A N D I N T E L L E C T U A L P R O P E R T Y intellectual property­related issues--full informa- tory liability" regime. Disclosure of access and tion. The preceding discussion suggests the follow- utilization of such resources can be mandated, ing policy options that might be pursued at the with disclosure triggering an obligation to pay international level by developing countries. "reasonable compensation." Copyright and Related Rights Unfair Competition · In the interest of developing their music indus- · As an alternative to an exclusive-rights regime for tries, developing countries should seek the incor- the protection of nonoriginal databases, develop- poration of the WIPO "Internet treaties" into a ing countries can consider arguing for the adop- future trade agreement, in the same way that the tion of an expropriation model of protection. TRIPS agreement incorporates the substantive · As an alternative to extending the protection of provisions of the Berne Convention. geographical indications, developing countries · Similarly, the incorporation of the WIPO Per- can seek strengthened protection against mislead- formers and Phonograms Treaty (WPPT) will ing the public as a form of unfair competition. establish an international regime for the indirect · To protect ethnobotanical knowledge against protection of expressions of folklore. unauthorized misappropriation, trade secret pro- · Although the sui generis protection of nonoriginal tection at the international level needs to be databases (those that do not meet the requirement strengthened. of originality in copyright law) is important for · As an alternative to the protection of ethnobotanical the protection and exploitation of ethnobotanical knowledge by a notification requirement of knowledge, one should be mindful of the possible patentability, a compensatory liability regime can be negative effects of a strong exclusive-rights regime introduced, bolstered by a disclosure requirement. on the scientific and education sectors. The Inter- national Bureau of the WIPO has commissioned a Notes study of such impacts, and the findings from this study should be considered carefully. This chapter was supported by the Policy Sciences Center, a non- governmental organization founded by Harold D. Laswell, Myres S. McDougal, and George Dession at the Yale Law School in 1948; Trademarks and Geographical Indications by the World Bank's Development Grant Facility for Culture; by the government of the Netherlands; and by the Swedish Interna- · The protection of geographical indications could tional Development Agency. be strengthened. Developing countries can argue 1 The sector of the international music recording trade known as strongly for extending a notification and registra- World Music emerged as a distinct category in the 1980s tion system of geographical indications for wines (Collins 2000). It began with the growing interest in the popu- to other products that can be recognized by their lar music of Africa, such as the jugu music, maringa, and geographical origins and that are economically Afrobeat of anglophone West Africa, the mbalax and electro- and culturally important to these countries, such griot music of francophone West Africa, the soukous of Central as handicrafts and agrofood products. Africa, the mbawanga and chimurenga music of southern Africa, and rai music from Algeria and Morocco. Today World Music includes, in addition to its popular African music com- Patents ponent, Cuban, Latin American, Arabic, Indian, Eastern Euro- pean, and Asian popular and folk styles. · A good case can be made for the introduction of 2 Article 15.4 forms part of the Stockholm (1967) and Paris an authorized access requirement. Where it (1971) Acts of the convention. According to the intention of appears that an invention which forms the subject the revision conference, this article implies the possibility of of a patent application is based on genetic granting protection to expressions of folklore (Ficsor 1997). resources that are part of a developing country's 3 Sometimes, of course, works that appear to be expressions of "biological and genetic heritage," a copy of the folklore can be traced back to their original authors, allowing contract affording access to the genetic resources them or their successors in title to recover royalties. This in the country of origin should be filed. occurred with "El Condor Pasa," a famous song of the 1970s. · This requirement can be transformed into a (less- 4 An interesting exception to this general observation has been er) notification requirement to fit a "compensa- noted in respect to the traditional Onge people of the 400 Cultural Industries and Intellectual Property Rights Andaman Islands in the Bay of Bengal (Norchi 2000). Songs 14 The special protection of well-known marks is based on article are composed for certain occasions, and their performance 6 bis of the Paris Convention, as extended by Articles 16.2 and may be requested again later. Only the original composer is 16.3 of the TRIPS agreement. This protection will be further then allowed to sing the song. Should anyone else try to do so extended by the adoption of the Joint Recommendation con- without the composer's permission, that act is treated as theft. cerning Provisions on the Protection of Well-Known Marks by the Assembly of the Paris Union for the Protection of Industrial 5 An example is the "life plus 70" rule (copyright ceases to sub- Property and the General Assembly of the WIPO, September sist 70 years after the death of the author) introduced for 20­29, 1999. The recommendation clearly raises the level of Europe by Article 1(1) of Council Directive 93/98 of October protection beyond that of the TRIPS agreement and extends 29, 1993, harmonizing the terms of protection of copyright the scope of the substantive subject matter by also dealing, for and certain related rights (1993 O.J. L290/9). example, with business identifiers and domain names (Kur 6 Some may argue that this proposal amounts to a new form of 2000). (indirect) taxation, which may make its adoption politically dif- 15 A "false indication" is one that does not correspond to the ficult in countries such as the United States (Gervais 2001). facts; it is an indication relating to a geographic area for goods 7 Examples of access control are the use of passwords and not originating in that area. An indication is false only where encryption; an example of copy control is the use of software this association/connotation is understood by the public in the which limits the number of copies that can be made of a digi- country where the indication is used for such products tal work. (Baeumer 1997). 8 The inclusion of temporary ("ephemeral") copies within the 16 Although this definition is based on that of an "appellation of reproduction right had the potential of defeating the entire origin" in the Lisbon Agreement, it is broader in one respect: treaty (Vinje 1997). Even this agreed statement, unlike the the TRIPS agreement protects goods that derive a reputation other agreed statements, was adopted not by consensus but from their place of origin without their having a quality or by majority vote. other characteristic that is due to that place (Baeumer 1997). 9 The general obligation in Article 11 gives contracting parties 17 See, for example, the communication submitted by Kenya, on an important degree of flexibility, when drafting their domes- behalf of the African Group, to the General Council of the tic legislation, to choose the types of technological measure WTO in preparation for the 1999 ministerial meeting in Seattle that should be protected, the types of sanctions that should be (WT/GC/W/302, August 6, 1999). imposed, and the actual activities that should be targeted. The 18 Essentially, the novelty of an invention is destroyed if it has provision in Article 12 reflects general agreement on the need been disclosed anywhere and in any manner or form (Art. 52). to protect certain types of information attached to works in order to provide some security for their identification and 19 Whether such prior use would actually destroy the novelty of tracking in open information networks (Reinbothe, Martin-Prat the invention for which a patent is sought depends on the and Von Lewinsky 1997). patent law of the country in which protection is sought. Under the U.K. Patents Act of 1977, for example, use can be the basis 10 This right is subject to the impairment test also found in of a challenge to novelty if it effects a public release equivalent respect to the limited (copyright) rental right in the TRIPS to publication (Cornish 1999). The question is whether a agreement (Art. 11). skilled worker, by observation or analysis, could discover and 11 A certification mark, in terms of the Trade Marks Act 1995 reproduce the invention; see Stahlwerk Becker's Patent (1919) (Cth), is a sign used, or intended to be used, to distinguish 36 R.P.C. 13 (HL). goods or services dealt with or provided in the course of trade 20 The phrase "a manner contrary to honest commercial prac- and certified in relation to quality, accuracy, or some character- tices" connotes "at least practices such as breach of contract, istic (such as origin, material, or mode of manufacture) from breach of confidence and inducement to breach, and includes other goods or services dealt with or provided in the course of the acquisition of undisclosed information by third parties who trade, but not so certified (sec. 169). Certification marks sym- knew, or were grossly negligent in failing to know, that such bolize and promote the collective interests of certain groups of practices were involved in the acquisition." traders; by preventing traders whose goods do not comply with the certification process from using the mark, the integri- 21 The use of the expression "secret information" as opposed to ty of those traders whose goods are certified is maintained "undisclosed information" in the TRIPS agreement does not (Wiseman 2001). imply any difference in substance; it merely indicates that the rightful holder of the information must take certain measures 12 Goods may include a wide range of items, such as fabrics, or must behave in a certain way to keep the information boomerangs, coolamons, nets, traps, seed and shell necklaces, unknown to third parties (Art. 6, note 6.01). didgeridoos, musical recordings, sticks, and sculptures. Ser- vices may include activities such as theatre, dance, concerts, 22 The "rightful holder" of secret information is the natural or and educational and tourism programs (Wiseman 2001). legal person who is lawfully in control of such information (Art. 6, note 6.03). 13 The Policy Sciences Center is assisting the Indian commissioner for handicrafts to implement a certification system for prod- 23 It has been argued that traditional intellectual property mod- ucts labeled "handmade in India." els, as supplemented by classic trade secret law, often fail to 401 T E C H N O L O G Y A N D I N T E L L E C T U A L P R O P E R T Y afford those who produce the current most commercially valu- able information goods enough lead time to recoup their investments. The risk of market failure inherent in this "state of chronic under-protection tends to keep the production of information goods at suboptimal level" (Reichman 1994). 24 Section 301 of the U.S. Trade Act of 1974, as amended, allows (and sometimes requires) the U.S. government to take unilat- eral retaliatory actions against alleged unfair trading practices of partner countries. Since 1988, the U.S. Trade Representative has been required to identify foreign countries that deny ade- quate and effective protection of intellectual property rights or deny fair and equitable access to U.S. intellectual property holders. 402 39 C A R S T E N F I N K B E ATA K . S M A R Z Y N S K A Trademarks, viduals in special depositories created by governments. Regis- tration prevents the coexistence Geographical of confusingly similar marks and serves as proof of ownership Indications, and in the case of, for instance, legal disputes. Although trademark Developing laws and registration offices can be found in virtually every Countries country, national regimes often differ markedly regarding the scope of protection, the require- ments for names and symbols T that can be protected, guidelines rademarks are words, signs, or sym- for avoiding confusing marks, registration costs, the bols that identify a brand with a legal means available for fighting infringement, and certain product or company. A consumer buying a other important details. Since trademark laws and can with the Coke or Pepsi logo knows that when regulations differ from country to country, a firm she opens the can she is going to taste a carbonated entering a foreign market may find that its trade- soft drink with a certain cola flavor. Similarly, a trav- mark is not protected abroad. To address this issue, eler booking a flight on Lufthansa usually has a intergovernmental treaties have been negotiated to good idea about the safety record of the airline, the protect the trademark rights of foreign firms. Most comfort of the seating area, the in-flight service, the prominently, the WTO Agreement on Trade-Relat- punctuality of arrivals and departures, and so on. In ed Aspects of Intellectual Property Rights (TRIPS) short, trademarks offer customers the assurance of requires nondiscrimination--national treatment purchasing what they intend to purchase. Geo- and most-favored-nation treatment--with regard graphical indications--or appellations of origin-- to intellectual property protection and defines min- fulfill a very similar purpose. They identify a imum standards of protection that the trademark product or service with a particular location. A tea laws of member countries have to meet. lover ordering a cup of Darjeeling tea anywhere in Since its inception, the TRIPS agreement has been the world wants to be assured that the leaves used to criticized for being detrimental to the interests of make the tea are indeed from the Darjeeling region developing countries in terms of transferring rents in India. to intellectual property­owning firms in the indus- Trademarks and geographical indications belong trial world, restricting access to essential products to the wider family of intellectual property rights such as life-saving drugs, and creating burdensome (IPRs) and as such receive legal protection against implementation requirements. Most of the contro- unauthorized use by third parties. Trademarks, like versy around TRIPS, however, has focused on inno- other IPRs, can be registered by businesses and indi- vation-type IPRs, notably patents, copyright, and 403 T E C H N O L O G Y A N D I N T E L L E C T U A L P R O P E R T Y plant breeders' rights, rather than on trademarks or holder's good reputation by offering products with geographical indications.1 inferior characteristics bearing the same brand This chapter reviews the general economic name, consumers would be willing to pay less for rationale for trademark protection and its role in brand-name products because of the risk of receiv- developing countries, assesses the obligations of ing fakes, and the trademark holder would, from a developing countries under the TRIPS agreement in social point of view, underinvest in developing this area, and discusses the interests of developing products with desirable characteristics such as countries in future multilateral negotiations. We quality and safety. conclude that trademark protection fosters efficient There is a key difference between protection of allocation of resources, promotes business develop- trademarks and protection of IPRs that seek to ment, and may enhance export performance. The stimulate creative and inventive activity--patents, potential negative effects of TRIPS-consistent copyright, plant breeders' rights, and so on. Higher trademark regimes are likely to be small, especially quality induced by trademark protection is clearly a compared with some other forms of intellectual private good. Better physical or functional features property covered by the agreement. Although the of a product or a greater extent of ancillary pre- and development of a business-friendly trademark after-sale services are rival in consumption. By con- regime is primarily a challenge for domestic reform, trast, one person's consumption of a new technolo- compliance with the TRIPS standards of protection gy or artistic expression induced by patent or can benefit developing economies. The discussion copyright protection does not diminish another mainly focuses on trademarks, but most of the person's consumption of the same technology or arguments are also applicable to geographical indi- artistic expression; thus, these products can be cations. We deal explicitly with geographical indica- thought of as public goods (Arrow 1962). From a tions only when there are considerations specific to social welfare perspective, preventing free-riding-- this type of intellectual property. as all forms of IPRs do--is clearly desirable when it concerns reputation for quality. It is only second best, however, in the case of a public good such as The Economics of Trademark Protection the formula for a new drug or the lyrics of a new The basic economic rationale for trademark protec- song, which, ideally, should be made available at tion goes back to Akerlof's (1970) seminal insight marginal cost to the public at large. This fundamen- regarding the failure of markets to provide for an tal difference is reflected in the fact that patents and efficient allocation of resources if consumers are copyright receive protection for only a limited time unable to assess the quality of products offered to period (20­50 years), whereas trademarks can them. In this situation, information asymmetries endure forever, provided they remain in use.2 between sellers and buyers prevent some transac- A closely related difference between the two forms tions in high-quality goods from occurring, leading of IPRs is that patents and copyright expressly grant to inefficiencies. Trademarks offer a way around this monopolies--albeit limited in scope--whereas dilemma. As producers of goods and services devel- trademarks can, in theory, coexist with perfectly op a reputation for quality over time, consumers competitive markets. Trademarks alone do not pre- can use brand names to distinguish between a pre- vent firms from imitating a competing good or ser- mium-quality product and a low-end product. vice, be it a low-end or a premium-quality product, Alternatively, brand names allow for easy identi- as long as it is sold under a different brand name. In fication of goods that are examined and compared this regard, geographical indications are somewhat by governments or private consumer groups with different because they do prevent firms from out- the aim of disseminating information about prod- side a given region from producing a perfect substi- uct reliability, safety records, and other quality tute if production relies on immobile, aspects. Consumers may be willing to pay a premi- region-specific inputs (such as favorable climatic um for branded goods, regardless of their quality, conditions, in the case of agricultural produce). In in order to minimize search costs. For this system practice, however, aside from selected luxury prod- to work, consumers must be able to clearly associ- ucts (for example, Roquefort or Parmesan cheese), ate the good or service with its producer. If com- geographical indications typically do not pose a sig- petitors were able to free-ride on the trademark nificant obstacle to interregional competition. 404 Trademarks, Geographical Indications, and Developing Countries Advertising-intensive consumer products (so- of scale. Second, trademarks contribute to greater called status goods) constitute a special group within product safety--crucial in pharmaceutical and food the category of products bearing trademarks. In the industries--since the goods can easily be traced case of status goods, the mere use or display of a par- back to producers that can be held responsible; ticular branded product confers prestige on its own- branded products are thus more likely to meet ers, apart from any utility derived from the product's mandatory health and safety standards. Legitimate function and physical characteristics (Grossman and producers may also have a greater incentive to fight Shapiro 1988a). Since in this case the brand name black-market producers if the latter free-ride on plays a central role in firms' product differentiation established brands. Third, trademarks may encour- strategies, it is no surprise to find that owners of age innovative activity. A firm that has developed a well-known brands often register up to 40 or more solid reputation for high-quality goods is likely to different trademarks to deter competing firms from invest continuously in upgrading its product port- entering their "brand space." Market research folio to maintain its competitive edge. Whether pro- reports regularly put the value of well-known brands tection of trademarks alone provides a sufficient at billions of dollars. For instance, according to incentive for engaging in research and development Interbrand, a consultancy, the Mercedes brand is (R&D) activities depends, however, on the extent to worth about US$22 billion. Brand affiliation is also which firms can appropriate their innovative efforts surprisingly stable; of the 25 top-selling consumer through trade secrets or simply by being the first in goods brands in the 1960s, 16 are still among the top the market. For many industries, other forms of 25 today (Evans and Wurster 2000: 150). intellectual property protection are likely to be nec- In the case of status goods, brand ownership can essary if the firm is to appropriate R&D efforts, and, confer substantial market power on producers, but indeed, companies that own a valuable brand name this does not necessarily imply that firms generate usually also show a strong portfolio of patents and supernormal profits. Since the potential supply of other types of IPRs.3 trademarks is limitless, as long as entry is otherwise A concern that frequently arises in the context of unrestricted the resulting market structure for intellectual property ownership is the potential for many status goods industries can be characterized anticompetitive practices by titleholders. As men- as monopolistically competitive--witness the tioned above, this concern is arguably less severe for intense competition between a relatively small trademarks than for patents and copyright because number of brand-name sports shoe producers trademarks do not expressly grant market power. around the world. Nonetheless, owners of strong brands can create A further twist in the case of status goods is that market entry barriers, most prominently through their perceived value often stems from their exclu- vertical restraints in licensing agreements with dis- sivity--in other words, from the fact that only a tributors and retailers--witness the government select group of consumers enjoys them. If trade- argument in the prominent U.S. antitrust case mark protection were not enforced, imitation against Microsoft. Appropriate competition policy would allow consumers with less discerning tastes remedies are necessary to counter such practices if to enter the "exclusive club." Hence, illegitimate pro- indeed they are found to be detrimental to con- ducers can impose an externality on trademark sumers.4 holders by preventing legitimate firms from offer- ing their customers the prestige associated with a Trademarks in Open Economies small, select network of users. This externality may dilute the market power of premium-good produc- The basic proposition that trademark protection ers and lower their incentives to invest in reputation improves economic efficiency by helping to resolve (Grossman and Shapiro 1988b). information asymmetries is strongest in internation- While the basic role of trademark protection is to al markets because asymmetries of information are provide incentives for firms to invest in quality and likely to be more pronounced when producers and reputation, there are other benefits associated with consumers are located in different countries. Eco- secure trademark rights. First, in the absence of nomic research has demonstrated the importance of counterfeiting, legitimate producers can more easily reputation in determining trade patterns between expand sales, allowing them to achieve economies countries (see, for example, Rauch 1999). Developing 405 T E C H N O L O G Y A N D I N T E L L E C T U A L P R O P E R T Y a brand name in a new market requires substantial price. On the other hand, prices can fall if a firm that investments in marketing and in establishing a distri- in the absence of protection targeted only high- bution network, and firms are unlikely to make the income, price-insensitive consumers now faces a investment if they fear misappropriation of their broader demand with potentially more price-sensi- trademark. tive buyers. These issues are very relevant to developing It should be noted that trademark protection, by economies. For instance, a recent survey conducted itself, does not curtail poor people's access to essen- in Lebanon showed that although local apparel pro- tial goods such as life-saving drugs. This is arguably ducers have a strong interest in designing high- the most controversial aspect in the debate over quality clothing for Middle Eastern markets, their patent rights in least-developed countries. Pharma- attempts to do so were hindered by trademark ceutical companies that have invested considerable infringement by smaller firms in Lebanon and resources in developing a new drug may charge neighboring countries (Maskus 2000c). Creating prices that keep their products out of reach of the conditions conducive to trademark development sick in developing countries. Enforcing patent pro- can have a positive impact on exports from develop- tection prevents other firms from producing a ing countries. Although there is no empirical evi- patented drug and selling it at a lower price. This is dence on this point, trademark protection might not a trademark protection issue; any firm could also encourage multinational corporations to set up produce and sell a good equivalent in physical char- facilities to produce goods for both domestic and acteristics to that produced by the trademark holder foreign markets, promoting the transfer of foreign as long as the protected brand name was not used. technologies and know-how. Anecdotal evidence Thus, consumers should always be able to purchase indicates that stronger trademark protection can a good of the same quality. encourage local business development through A dimension of trademark law that becomes rele- international franchising arrangements. vant in the international context is the exhaustion A question relevant from the perspective of devel- doctrine, which defines the territorial rights of oping countries is whether the enforcement of for- trademark holders after goods have been first put on eign trademarks leads to greater profits for foreign the market. Under a system of national exhaustion, a titleholders and a transfer of rents. As noted above, titleholder can prevent parallel importation of the the market power of trademark holders is most like- trademarked product from a foreign country where ly to arise in the case of status goods, implying that it is sold either by the trademark owner or by an economically significant rent transfers are likely to authorized dealer. Trademark holders, provided they be confined to this group of products. Although no have some market power, can then price their goods empirical estimates of the potential size of such according to market-specific demand and thereby transfers are available, there are several reasons to generate higher global profits. In contrast, if rights believe that their potential size is limited, especially exhaust internationally, the titleholder loses his or when compared with the enforcement of patent her exclusive privilege after the first distribution of rights. First, it is unlikely that all operating profits the product, thus allowing parallel imports from for- will be transferred to the foreign titleholder's home eign territories. Leaving aside transport and other base; firms may decide to reinvest part of the profits trade costs, arbitrage between markets leads to a uni- in marketing, reputation building, and ancillary form price across markets of the trademark owner's sales services abroad. Second, the existence of close product. Trademark exhaustion regimes differ substitute products is likely to increase the price sen- markedly across countries. Industrial countries tend sitivity of consumers and limit the pricing power of to be more closed to parallel imports of trademarked the foreign titleholders (Fink 1999). Trademark goods than developing countries, but there are many enforcement by local authorities will give consumers exceptions and legal details that qualify this general the option to switch to generic (previously, counter- rule. (See Abbott 1998b for an introduction to the feit) products. The effect of trademark protection on legal treatment of parallel imports.) the trademark holder's price is theoretically ambigu- No consensus exists among economists as to ous. On the one hand, displacement of counterfeit which exhaustion doctrine constitutes the "better" producers increases the market power of the trade- policy. (See Fink 1999 for a survey of the pros and mark-holding firm, which tends to increase the cons of freeing parallel imports.) One justification 406 Trademarks, Geographical Indications, and Developing Countries for restricting parallel trade is based on the possibil- than patents and may therefore be more suitable to ity of benign price discrimination, of which poorer developing country environments.6 Much lower countries are thought to be the main beneficiaries.5 outlays are required to administer a trademark reg- Another argument in favor of national exhaustion istrar than a patent office. Evaluating patent appli- extends the argument of beneficial vertical cations requires specialized knowledge of many restraints to the international level. Specifically, par- technical areas, which is not necessary in the case of allel importers are sometimes accused of free-riding trademark approvals. Therefore, it may be more on country-specific investments in marketing or viable to finance trademark offices through regis- sales services by the trademark holder or the official tration fees without imposing high costs on appli- local distributor. cants. Trademarks are also more accessible from the The main argument in favor of allowing parallel perspective of firms in developing countries, espe- importation is to stimulate what is sometimes called cially small and medium-size enterprises. The costs intrabrand competition. Especially in countries of creating a trademark are small compared with with small markets and undeveloped competition the resources required to develop a patentable inno- policies, parallel imports can increase competition vation. Furthermore, trademark applications are in the distribution and retailing of products. As a significantly less complex than patent applications, rule of thumb, therefore, many developing coun- and applying for a trademark is less costly.7 This tries are likely to be better off under a system of helps explain why local residents in developing international trademark exhaustion. In practice, countries control a larger share of trademarks than however, firms often can prevent parallel importa- patents in their countries, as shown in Figure 39.1. tion through private contractual means. Parallel Firms seeking protection for their company or imports are also curtailed if firms own other forms brand names need to file one or more registrations of intellectual property rights on a particular good at designated national offices. Statistics on the num- and these rights exhaust nationally (see Fink 1999). ber of new trademark registrations, which national Hence, an international exhaustion regime is no offices regularly report to the World Intellectual substitute for an efficiency-oriented competition Property Organization (WIPO), provide insights on policy that scrutinizes vertical restraints in terms of the use of the trademark system by domestic and the benefits and costs to consumers. foreign companies.8 Figure 39.1 depicts the global "demand"for trademarks and, for purposes of com- parison, patents as well, in 1981­82 and 1996­97. Can Trademarks Work in Developing Countries? The number of trademarks granted by governments Information asymmetries in goods and services around the world increased more than threefold markets are at least as pervasive in developing coun- between those periods, to about 1.4 million a year. tries as in the industrial world. Thus, developing The number of patents also rose, but only by 2.3 economies, in principle, can benefit from establish- times, to 750 million. ing an effective framework for the registration and Several forces can explain this surge in the use of protection of trademarks and geographical indica- trademarks and patents. For one thing, the expan- tions. Although most developing countries do have sion of international trade and other forms of eco- trademark laws on the books, the effectiveness of nomic integration may have led titleholders to seek the laws is often held back by inadequate adminis- broader geographic coverage for their IPRs. But this tration and enforcement procedures. These inade- globalization effect only partly accounts for the quacies can be the result of corrupt and inflated growing demand for protection. As revealed in Fig- bureaucracies, weaknesses in the legal system at ure 39.1, the number of grants to domestic residents large, or the lack of a broader competition policy only has also risen. This purely domestic effect in framework. Hence, a weak overall governance struc- the case of trademarks is likely attributable to ture--often associated with low income levels-- changes in firms' strategies toward brand naming, poses a major challenge in harnessing the positive the evolving competitive landscape across indus- contribution trademark protection can make tries, and legal reforms in selected countries. toward a business-friendly investment climate. In 1996­97 about 58 percent of all trademarks Notwithstanding these constraints, it can be were granted to domestic residents (Figure 39.1). argued that trademarks are less resource intensive This share had fallen somewhat since 1981­82, 407 T E C H N O L O G Y A N D I N T E L L E C T U A L P R O P E R T Y 39.1 Global Demand for Trademarks and Patents Number of worldwide grants (thousands) 1,600 1,400 1,200 1,000 800 600 400 200 0 1981­82 1996­97 1981­82 1996­97 Trademarks Patents Nonresidents Residents Note: For various reasons, numbers should be considered orders of magnitude only. First, the number of countries included in the two dif- ferent periods differed because of political changes. Second, the availability of protection, the strength of protection, and the quality of the administrative system for intellectual property rights differ significantly across countries. In some countries, for example, firms may be deterred from filing a trademark or patent because enforcement is weak or because it takes several years to complete the application process. Finally, in some developing countries the collection of industrial property statistics is unreliable, and reporting practices to the World Intellectual Property Organization (WIPO) are not consistent across countries. Sources: Braga, Fink, and Sepulveda (2000); authors' estimates from WIPO data. when it stood at around 64 percent, but it was still minimum standards of protection and priority fil- larger than for patents, where little more than 40 ing rights, which offer trademark holders a time percent of grants went to domestic residents. This period of six months to file their titles abroad once pattern is also borne out in Figure 39.2, which com- they have registered them domestically. Firms resid- pares trademark and patent grants in 1997 in low- ing in countries that are members of the 1891 and middle-income countries with grants in high- Madrid Convention, currently encompassing 68 income countries. Remarkably, more than half of states, can substantially reduce the transactions trademark grants in developing countries went to costs involved in registering trademarks in multiple domestic residents. Although this share is somewhat countries by filing a single international application lower than in the high-income group, it suggests and designating the countries where they seek pro- that developing country firms do not use the trade- tection. Similarly, the Lisbon Treaty of 1958, with 20 mark system much less intensively than their indus- signatories, provides for an international registra- trial country counterparts. tion mechanism for geographical indications. The WTO TRIPS agreement goes beyond the pre- existing IPR conventions. It sets more stringent International Agreements on Trademarks international standards for trademark protection in Nations have long protected the trademark rights of selected areas and, for the first time in an interna- foreigners and have cooperated to facilitate the tional treaty, establishes enforcement principles that cross-border registration of trademarks. Most WTO members are required to meet. The specific prominently, the Paris Convention of 1883, to TRIPS provisions on trademarks require members which 162 countries are signatories, establishes to establish registration procedures that are trans- 408 Trademarks, Geographical Indications, and Developing Countries 39.2 Nationality of Trademark and Patent Title Holders by Country Income Group, 1997 Percentage of grants 100 75 50 25 0 Low and middle High income Low and middle High income income income Trademarks Patents Note: The definition of income groups follows the World Bank classification. In the case of patents, countries of the former Soviet Union were excluded because the reported data appear to include grants for utility models. Source: Authors' estimates from WIPO data. parent and independent of the characteristics of formal deadline for completing the negotiations on goods and services for which protection is request- the above-mentioned system of registration by ed. Countries must also extend protection to inter- 2003, when the Fifth Ministerial Meeting is to be nationally recognized trademarks in order to held. WTO members still disagree, however, on the prevent their speculative registration and fraudu- issue of extending to other products the higher level lent use. It is important to recognize, however, that of protection currently granted to wine and spirits, in many countries the trademark standards negoti- as called for under TRIPS.9 ated under TRIPS were already part of pre-TRIPS With regard to enforcement, TRIPS sets general law and jurisprudence (Watal 2001). standards on, among other things, enforcement On geographical indications, the TRIPS agree- procedures, the treatment of evidence, injunctive ment calls for protection against use that would relief, damages, and provisional and border mea- mislead the public or constitute an act of unfair sures. At the same time, Article 41.5 makes clear that competition. A stronger form of protection is pro- countries do not need to "put in place a judicial sys- vided for wines and spirits, for which the use of tem for the enforcement of intellectual property translations or expressions such as "kind" or "type" rights distinct from that for the enforcement of law is prohibited. (For example, wine sold as "Chablis- in general, nor does it affect the capacity of mem- type wine, made in Chile" would be in violation of bers to enforce their law in general." the agreement.) The agreement calls for the estab- lishment of a multilateral system of notification and The Benefits and Costs of Trademarks under registration of geographical indications for wines TRIPS for Developing Countries and spirits, although little progress has been made in this area to date. The potential benefits associated with legally bind- The Final Declaration of the WTO's Doha minis- ing disciplines on trade-related measures under the terial meeting reaffirms the agenda on geographical WTO have to do with overcoming domestic political indications built into the TRIPS agreement. It sets a economy constraints, enhancing the credibility of 409 T E C H N O L O G Y A N D I N T E L L E C T U A L P R O P E R T Y domestic regimes, and enjoying access to independ- What are the costs of implementing the TRIPS ent arbitration of disputes. How significant are these obligations on trademarks? As described above, benefits in the case of TRIPS trademark obligations? only minor adjustments, if any, are necessary to In the first place, some governments may see the make trademark laws in most developing countries benefits of trademark protection but be politically compatible with the TRIPS requirements. More- unable to introduce such protection because of over, trademark offices in most developing coun- strong pressure from producers of counterfeit tries are funded through user fees and sometimes goods. Multilateral negotiations can play a useful even generate a surplus for the government. An role in overcoming such pressure, as countries increased volume of applications that could result would gain in other areas being negotiated, leading from TRIPS compliance is therefore unlikely to to a dynamic similar to that commonly associated impose a heavy burden on government budgets. with reciprocal trade negotiations. It is important to The most significant adjustment cost is in the area recognize, however, that counterfeit activities in of IPR enforcement. Watal (2001) concludes that most developing countries rely mostly on weak developing countries following common law sys- enforcement of existing laws rather than on the tems were largely in compliance with the TRIPS absence of laws per se. Since the general enforce- obligations in regard to legal enforcement mea- ment standards and procedures are unlikely to sub- sures. Some changes to laws and procedures were or stantially weaken political ties between counterfeit are necessary in several Southeast Asian and Latin producers and the government, the potential bene- American countries in areas such as border mea- fits of TRIPS in this context are limited. sures, preventive measures, or evidentiary require- Second, compliance with the TRIPS trademark ments. As mentioned, TRIPS does not require standards can enhance the credibility of national countries to enforce intellectual property rights trademark regimes by eliminating the risk of policy more stringently than other aspects of law, and it reversal. It can signal to domestic and foreign can therefore be argued that developing countries investors that the country is "open for business." are not required to devote "excessive" resources Again, however, it is important to realize that the solely to the enforcement of trademark rights. TRIPS standards of protection had already been largely satisfied in most countries and that coun- Moving Forward tries had little incentive to unilaterally weaken pro- tection. Thus, the extent to which TRIPS enhances Geographical indications are one area in which credibility depends largely on the degree to which potential extensions of the agreement could benefit the agreement results in strengthened trademark developing countries. Developing economies have enforcement. an interest in extending the higher level of protec- Third, the WTO provides a forum for countries to tion currently being granted wines and spirits under settle disputes on trademarks on "objective" legal TRIPS to other kinds of foodstuffs and agricultural grounds and offers independent arbitration, which products and even to handicrafts. Examples of can contribute to reducing potential political back- developing country products that may benefit from lashes arising from intergovernmental disputes. In such protection include Darjeeling tea, Kenyan cof- view of the territorial approach to intellectual prop- fee, and Kente cloth from Ghana. Developing coun- erty, such disputes are bound to emerge in an tries that have expressed interest in broadening the increasingly integrated world economy. Interesting- scope of protection include Cuba, Egypt, India, ly, the only two disputes on trademarks and geo- Mexico, Morocco, Nigeria, Turkey, and Venezuela. graphical indications have been between the Recently, Mexico demanded protection for "Tequi- European Union and the United States, both of la," while Central European countries expressed which provide strong protection for intellectual interest in expanding geographical indications to property.10 Although developing countries are gen- cover cheese, chocolate, beer, and embroidery (Watal erally disadvantaged in the dispute settlement 2001: 267). In this context, progress on establishing a process (since they can commit fewer resources to multilateral system of geographical indications defending their interests at the WTO), they are like- could further the interests of developing countries. ly to prefer multilateral arbitration to unilateral One difficulty in the area of geographical indica- measures by large industrial countries. tions concerns terms considered generic or semi- 410 Trademarks, Geographical Indications, and Developing Countries generic and therefore not eligible for protection. In economies of scales, and promoting exports. The low future negotiations, developing countries should cost of creating and registering a trademark makes ensure that their geographical indications do not the benefits of protection accessible to local firms, become generic terms. This may require bilateral including small and medium-size enterprises. agreements with industrial country governments, as Although many of these benefits can be obtained well as the filing of court cases and challenges to by establishing or strengthening national trademark registration of offending trademarks in other coun- regimes, developing countries can reap benefits, in tries. Both kinds of action may require substantial the form of enhanced credibility and access to inter- financial and human resources. national arbitration, from adopting TRIPS provi- The emergence of the Internet and the resulting sions with regard to trademarks, One has to keep in new issues related to trademark protection and mind, however, that many of the deficiencies of international cooperation constitute another area trademark systems in the developing world stem for future negotiations. In cyberspace, borders from broader institutional weaknesses, which com- between nations are in many ways meaningless, pliance with the TRIPS obligations will not address. challenging the territorial approach to intellectual Specifically, trademark protection is unlikely to property protection. A noteworthy development in have many positive effects unless it is accompanied this context was the creation of the WIPO dispute by improvements in the overall quality of the legal resolution body for domain names. This provides system. In addition, when providing trademark holders of trademark rights with an administrative protection it is crucial to develop a competition pol- mechanism for challenging the bad-faith registra- icy in order to prevent anticompetitive practices tion of Internet domain names that correspond to resulting from strong brand ownership. those trademarks. The arbitration of disputes between private parties by an intergovernmental Notes organization such as the WIPO arguably represents the deepest form of international cooperation on The authors are grateful to Keith Maskus for his helpful com- intellectual property. To date, this mechanism has ments. worked well; in 2000 the body received more than 1 A patent is a legal title granting the owner the exclusive right 3,200 cases.11 In the future, demands may well arise to make commercial use of an invention. Copyright protects to address trademark issues related to Internet the expression of an artistic or literary creation, such as a song, domain names and e-commerce in the TRIPS con- book, software code, or motion picture, but allows for free copying of the ideas or methods embodied in it. Plant breed- text, although it is not yet clear what the focal point ers' rights protect new plant varieties that are distinct from the of such demands would be. It is important that existing ones and that are uniform and stable. developing countries which are likely to become 2 Nordhaus (1969) framed the task for public policy of defining active players in e-commerce defend their interests the length of patent protection as a maximization problem in during discussions on the evolving international which the objective function is the net welfare gain to society architecture for IPR-related Internet issues. brought about by the invention, constrained by the require- ment that the inventor's return be sufficient to encourage inventive activity. Conclusion 3 In some cases, trademarks may extend the benefits of innova- In contrast to the other areas of the TRIPS agree- tion beyond the life of a patent. For instance, consumers may ment, protection of trademarks is unlikely to be continue to purchase a brand-name drug even though gener- detrimental to low-income countries. Since trade- ics are offered, simply because they have confidence in the brand. mark protection does not prevent development of substitute goods (as long as those do not bear pro- 4 As is well established in the literature on vertical controls, tected brand names), it is unlikely to be associated restrictive licenses may actually be in the interest of con- with large rent transfers abroad, nor are large outlays sumers--for example, to ensure downstream product quality on the part of local vendors. required to administer the system. Indeed, trademark protection can have a positive impact on developing 5 For formal treatments of this argument, see Hausman and countries by increasing the amount of information MacKie-Mason (1988) and Malueg and Schwartz (1994). available to consumers, encouraging firm investment 6 At the same time, we do not want to dismiss the potential in quality and reputation, bringing about firm-level benefits of patent protection in developing countries. 411 T E C H N O L O G Y A N D I N T E L L E C T U A L P R O P E R T Y 7 It must be kept in mind that while obtaining a trademark can be relatively inexpensive, marketing it may require significant resources. 8 Several caveats need to be pointed out. First, new registrations are a flow figure, which may not necessarily correspond to the stock of trademarks in force. Second, the number of grants is unlikely to be proportional to the "economic value" of the underlying intellectual property titles. Finally, although the sta- tistics consider all domestically incorporated firms as domestic residents, it is possible that these firms are partially or fully owned by foreigners. 9 So far, the discussions have been marked by divisions between the EU (the trade bloc that hosts the largest number of geo- graphical indications) and "new world" countries such as Argentina, Australia, Canada, Chile, and the United States, which are mostly "consumers" of such indications and thus favor relatively lax standards of protection. 10 In one dispute, the European Union challenged the TRIPS-con- sistency of a U.S. law regarding the use of trademarks in con- nection with businesses or assets confiscated in Cuba. The other dispute was brought by the United States against the European Union's alleged lack of protection of trademarks and geographical indications for agricultural products and food- stuffs (see ). 11 Its success can be attributed to two factors. First, the preven- tion of "cybersquatting" is in the interest of all countries and cannot readily be remedied by purely domestic enforcement. Second, the Internet is a radically new phenomenon, with few entrenched interests, the promise of large economic benefits, and a user community that is inherently "global" and there- fore likely to respect international arbitration. 412 VI " B E H I N D - T H E - B O R D E R " A N D R E G U L AT O R Y I S S U E S A s tariffs and quotas became less important The chapters in this part discuss the principal in OECD countries as the result of rounds "behind-the-border" regulatory issues that have of multilateral trade negotiations, differences in nation- come up in the WTO context and provide an intro- al regulatory regimes became more apparent. Policies duction to the key questions that are likely to be that were put on the table for possible negotiation of prominent on the multilateral negotiating agenda rules included government procurement, product in the coming decades. Simon J. Evenett, in Chapter standards, inward foreign investment, competition 40, discusses issues surrounding government pro- law, labor standards, and environmental norms. The curement. Product standards and related regulation WTO now includes disciplines in some of these areas-- are the topic of Chapter 41, by John S. Wilson. in procurement (albeit only for signatories of the rele- Bernard Hoekman and Kamal Saggi, in Chapter 42, vant agreement) and in product standards. In others review the pros and cons, from developing coun- there are as yet no disciplines, although all have advo- tries' standpoint, of multilateral rules for investment cates. Calls for multilateral action range from coordi- policies. Similarly, Chapter 43, by Peter Holmes, nated application of national policies to the looks at the case for rules on competition policies harmonization of domestic regulatory regimes. and anticompetitive behavior. In Chapter 44, Simon 413 " B E H I N D - T H E - B O R D E R " A N D R E G U L AT O R Y I S S U E S J. Evenett examines the role and effectiveness of countries to take actions that restrict trade if this is national anticartel policies in an era of globalization. necessary for the realization of domestic social and Finally, in Chapter 45 Simon Tay reviews North- other objectives. Often, however, such policies will South positions on labor standards, and in Chapter not be efficient; a standard principle of economics 46 Veena Jha examines debates and landmark dis- holds that distortions should be dealt with at the putes concerning environmental regulation and source, and in most cases trade is not the cause of trade agreements. the distortion. All these areas are relevant for development; the A major factor driving WTO talks on regulatory issue is to determine what makes sense from a issues is not so much that policies explicitly discrim- national point of view. As far as the WTO is con- inate against foreign firms as that differences in cerned, a key concern is to identify the rationale for, national regimes impose additional costs on firms and the objectives behind, proposals to address a that operate globally. Efforts to impose common specific issue in the WTO and to determine what standards that at first glance appear to be sensible type of cooperation will help attain domestic regula- measures for reducing compliance costs may well tory objectives. The Doha Ministerial Declaration be inappropriate because differences in regulatory called for the initiation of negotiations in a number regimes should reflect differences in national cir- of regulatory policy areas, including investment and cumstances and preferences. Thus, in many areas competition policies, trade facilitation, transparency harmonization may make little sense. This does not in government procurement, and the relationship imply that multilateral rules cannot be beneficial. between multilateral trade and environmental agree- For example, rules requiring transparency and dis- ments. Negotiations in the first four areas (the "Sin- closure can be very beneficial in efforts to combat gapore issues," so called because working groups on corruption and rent-seeking. Rules of thumb can be these issues were created at the 1996 WTO ministe- identified with respect to the design of regulatory rial meeting in Singapore) are to be launched in regimes and implementing institutions. Many of the 2003, at the Fifth WTO Ministerial Conference, if chapters that follow identify such rules. How to consensus exists on the modalities of negotiations. apply them specifically is something that must be Developing countries must ensure that the global determined by national authorities. rules negotiated in multilateral bodies support the In many of the areas discussed in this part, strong development process. One message that emerges vested interests may oppose efficient regulation or strongly from recent research is that the traditional measures that are needed to attain social or other "one size fits all" WTO approach is not necessarily noneconomic objectives. Here, as is the case with appropriate in areas where countries lack both the traditional trade policies, it is important that various experience and the critical institutional develop- groups outside government have the ability to ment needed to apply proposed rules. In the undertake analysis to identify the impact of status absence of country-level experience, it is difficult to quo policies--or the absence of policies. This is par- determine what types of cooperation (rules of the ticularly important in areas such as government pro- game) will support the development process. Devel- curement (discussed in Chapter 40) and protection oping and supporting research that helps to identify of the environment (Chapter 46), two areas in and distill such country experience is vital. which bad policies can have huge costs for the The behind-the-border agenda is essentially a economy and society. development agenda; at the national level it In some cases regulatory bodies and instruments involves determining what types of rules should will have to be upgraded, and this will require apply and what types of institutions are required to resources. For example, modernization of develop- implement them. While some regulatory policies ing country product standards systems, including may have implications for the ability of foreign firms institutions and infrastructure related to goods pro- to contest markets, this should not be the major fac- duction and sanitary and phytosanitary standards, is tor driving efforts for multilateral rule-making. Regu- a critical part of the new global trade environment lation should address domestic or global market (see Chapter 41). Efficient and market-driven stan- failures and do so efficiently. This may imply impos- dards systems can aid export competitiveness and ing restrictions on trade--which explains why there product quality. Meeting international standards for is a general exceptions article in the GATT allowing quality and for health and safety is increasingly a 414 "Behind-the-Border" and Regulatory Issues precondition for contesting international markets. behind-the-border issues, as well as an introduction Many low-income countries are not adequately to the economics of policy in these areas. equipped to deal with evolving product standard Many of the contributions in Bernard Hoekman requirements and need substantial assistance to sat- and Petros C. Mavroidis (eds.), Law and Policy in Pub- isfy export requirements. This is not necessarily an lic Purchasing: The WTO Agreement on Government issue for negotiation at the WTO; what is required Procurement (Ann Arbor: University of Michigan first and foremost is financial and technical assis- Press, 1997), discuss the genesis and operation of tance on a scale commensurate with the problem. the Government Procurement Agreement (GPA) and An important issue that arises for many of the sub- its relevance to countries at differing levels of devel- jects discussed in this part of the Handbook is to opment. Alan Sykes, Product Standards for Interna- determine the sequencing of reforms and to identi- tionally Integrated Goods Markets (Washington, D.C.: fy the preconditions for pursuing policy actions in a Brookings Institution Press, 1995), discusses stan- given area. For example, antitrust regulations make dards and standardization in the context of trade a lot of sense from a consumer welfare point of and trade policy. UNCTAD's annual World Investment view; what could be objectionable about instru- Report is the standard source of global data on FDI ments designed to ensure that firms do not exploit trends. F. M. Scherer, Competition Policies for an Inte- market power to raise prices? Policymakers, howev- grated World Economy (Washington, D.C.: Brookings er, must determine the social cost-benefit ratio of Institution Press, 1994), provides an introduction to designing and implementing antitrust legislation the linkages between competition and trade policies. (see Chapter 44). One question to ask is whether Kym Anderson and Richard Blackhurst (eds.), The this is the most effective and efficient instrument Greening of World Trade Issues (London: Harvester available to ensure that markets are contestable. If Wheatsheaf, 1992), brings together papers explor- there are many state-owned companies or public ing the linkages between trade and the environ- sector monopolies, antitrust legislation will have lit- ment. Pietro Nivola (ed.), Comparative tle effect, as such firms are generally excluded from Disadvantages? Social Regulations and the Global its reach. More fundamentally, a government may Economy (Washington, D.C.: Brookings Institution get a much bigger "competition" return by reduc- Press, 1997), is an in-depth discussion of the ten- ing regulatory barriers to entry that confront new sions that arise between national regulatory regimes firms. Another consideration is whether the sup- and "competitiveness." David Vogel, Trading Up: porting infrastructure to make antitrust work is in Consumer and Environmental Regulation in a Global place. Is there an effective legal and judicial system Economy (Cambridge, Mass.: Harvard University with competent and honest judges? Does the Press, 1995), is an excellent treatment of the rela- expertise exist to undertake the required investiga- tionship between trade, trade policy, and regulatory tions? Such questions arise in many areas of regula- regimes that discusses a number of cases in depth. tion and must be carefully considered by national On the general theme of dealing with domestic pol- authorities. icy differences, see Jagdish Bhagwati and Robert Hudec (eds.), Fair Trade and Harmonization: Prerequi- site for Free Trade? (Cambridge, Mass.: MIT Press, Further Reading 1996), and Gary Burtless and others, Globaphobia: The CD-ROM that accompanies this Handbook con- Confronting Fears about Open Trade (Washington, tains a number of useful readings dealing with D.C.: Brookings Institution Press, 1998). 415 blank 40 S I M O N J . E V E N E T T Multilateral (national treatment) does not apply to procurement. It was not until the completion of the Disciplines and Tokyo Round of multilateral trade negotiations, in 1979, that Government a code of conduct for central government procurement was Procurement introduced into the GATT. The code bound only its signatories, and most GATT contracting parties did not join. During the Uruguay Round, the coverage of the agreement was expanded to T include services and additional he growth of total spending by cen- government entities. Membership, however, tral and local governments was one remains limited mostly to OECD countries. of the most profound economic changes of the 20th A number of factors explain the greater accept- century (Tanzi and Schuknecht 2000). Although a ance by OECD nations of international rules on large portion of that growth was devoted to higher national procurement practices. Since the mid- transfer payments, governments and state-owned 1970s, government budgets in these countries have entities of all types spent considerable sums on come under increased pressure. Spending by welfare goods and services produced by the private sector. states on health, education, and pensions has grown Throughout much of the post­World War II era, considerably. Faced with the choice of raising taxes discrimination against foreign suppliers was rife. It or cutting nonwelfare spending, governments came was driven in part by the prevailing Keynesian to place greater emphasis on cuts. One way to macroeconomic orthodoxy, which emphasized that reduce nonwelfare spending is to stimulate compe- the smaller is the share of each dollar of government tition between firms that bid for government con- expenditure spent on goods produced abroad tracts, and allowing foreign suppliers to bid helps (imports), the larger is the increase in national achieve this goal. income caused by a rise in government expenditure. A second, related factor has been the widespread Governments could reduce this share--what econ- privatization of state-owned enterprises. Without omists call the marginal propensity to import--by governments standing behind them, these enter- refusing to buy goods from abroad. More important prises know that poor management decisions will rationales are nationalism and outright protection- not be "bailed out" by the state; that is, they face ism: claims that "our money" should be spent on what economists refer to as a "hardening of the "our goods" to secure "jobs at home." Because of budget constraint." But the owners of newly priva- such considerations, government procurement was tized firms also know that their profits will not be initially excluded from the GATT; Article III entirely taxed away by the state. These two changes 417 " B E H I N D - T H E - B O R D E R " A N D R E G U L AT O R Y I S S U E S provide the privatized firm with an incentive to such aid is tied to purchases from firms in a donor lower costs, including expenditures on goods and country, this diminishes the size of the national pro- services bought from other suppliers, and this may curement market and arguably also comes at some open up greater opportunities for cost-effective for- cost to the recipient nation. eign vendors. Estimates of the margin of preference--the price A third factor has been "export politics." Domes- wedge that would generate the observed reluctance tic firms see profitable opportunities in supplying of governments to purchase from abroad--suggest foreign governments and press their own govern- that there is considerable discrimination against ments to negotiate access to those overseas procure- foreign suppliers in national procurement markets. ment markets. Reciprocity considerations then Estimates of this margin by Francois, Nelson, and induce nations--often in the same region--to Palmeter (1997) range from 13 to 50 percent for simultaneously increase access to their procurement leading OECD nations across a wide range of gov- markets by firms from the same region. Reciprocity ernment purchases. Taken together, this evidence ensures that contracts lost by domestic firms to for- suggests that there is considerable scope for expand- eign suppliers are compensated for by an increase in ing competition in national procurement markets. contracts won in neighboring countries--although the firms that lose on the domestic market are Targets and Instruments unlikely to be the same as those that win more exports. Jan Tinbergen, a macroeconomist and Nobel laure- This chapter explores the reasons why govern- ate, introduced the distinction between a govern- ments engage in procurement discrimination, ana- ment's objectives (targets) and the policies lyzes the economic effects of discriminatory (instruments) that can be used to attain them. Tin- policies, and discusses the implications of this bergen argued that each target generally requires a analysis for the design of multilateral disciplines on specific instrument. In the case of procurement procurement. We start with a brief presentation of regimes, governments tend to have multiple targets the data on what governments procure. and numerous instruments. This reality compli- cates analysis of procurement policies and suggests the need for a multifaceted approach to internation- The Government Procurement Market al disciplines on the use of such policies. It is important to appreciate how large the world Perhaps the most common objective of procure- government procurement market is. Although pre- ment policies is to obtain value for money. This cise estimates are hard to come by, a recent OECD straightforward objective has many aspects. It could analysis indicates that total central government mean the purchase of goods that meet certain quali- expenditures by OECD members, excluding mili- ty levels, at minimum cost. Alternatively, it could tary spending and compensation of state employ- mean choosing the good with the highest quality ees, was almost US$2 trillion in 1998. The among a set of similarly priced goods. Governments comparable figure for 106 non-OECD members, also use procurement policies to favor certain which account for 90 percent of the developing groups, firms, regions, and industries. Such world's GDP, was just under US$0.3 trillion. Prior favoritism is widespread in countries as different as studies have estimated that local and provincial the United States and India, where the central gov- governments tend to spend approximately half as ernments give preference to small firms that bid for much again on goods and services, bringing the size contracts. Finally, purchases of military equipment of the world's procurement market to well over have long been subject to different rules, as nations US$3 trillion. have felt the need to maintain a broad range of mili- A large fraction of this spending is on services, tary production facilities for national defense pur- with the proportion of services higher in industrial poses. Such noneconomic objectives of procurement nations than in developing countries. In low- policy often involve the adoption of measures that income developing nations, a sizable proportion may not maximize economic efficiency. (10­20 percent is not uncommon) of the nonde- The manner in which governments procure fense government budget is funded by aid, loans, or goods and services can be complex. Our main focus grants (Hoekman 1998b: table 4). To the extent that here is on favoritism toward domestic suppliers-- 418 Multilateral Disciplines and Government Procurement that is, discrimination against foreign suppliers, When evaluating tenders, procuring entities must both suppliers located overseas and those located assess not only the price of the tender but also non- domestically that are partially or fully foreign price factors such as the quality of the good, after- owned. The latter include subsidiaries of multina- sales service, the reliability of the supplier, and so tional corporations with headquarters located on. Many of these factors are hard to quantify, and abroad. The question of transparency in procure- some subjective judgements are inevitable. The role ment is addressed at the end of this chapter. of intangible factors that enter into the contract The procurement of goods can be divided into award can cause controversy. Firms that do not win four distinct stages: tendering, evaluation, choice, the contract may feel that they were discriminated and challenge. At the tendering stage the govern- against, that the evaluation was conducted unfairly ment notifies suppliers of its desire to procure cer- or incorrectly, or that irrelevant factors were taken tain types of goods. The government can specify into account. the type of goods it wishes to buy, including, in Some governments have established procedures principle, the manner in which those goods are that enable firms to challenge the decisions of pro- produced. Three types of tendering procedures are curement officials. The specific details of these pro- distinguished in most legislation: open, selective, cedures determine, in large part, their effect on the and limited. In open procedures an interested sup- tendering and award steps of the procurement plier, located at home or abroad, can submit a bid process. If aggrieved parties can only recover the (or, as it is often called, a tender). Under selective costs of preparing the bid and of mounting a chal- procedures only prequalified firms may submit a lenge, a firm will be disinclined to object to a pro- tender. Prequalification is a mechanism to limit curement decision unless there is an expectation bidding to firms that meet certain performance that similar tenders will be issued in the future and criteria. Limited tendering procedures are the that the challenge will alter the subsequent behavior most restrictive, as the procuring agency invites of procuring officials. (Of course, the firm might only certain suppliers to submit tenders. Although also be concerned about retaliation by these officials there are circumstances in which limited tendering in future tenders.) If the challenge procedure offers is appropriate (for example, ensuring quick deliv- the potential to win substantial damages, this may ery in a time of national emergency), it frustrates well reduce inappropriate behavior by procuring one of the principal objectives of a procurement officials, but it may also lead to frivolous challenges regime--to stimulate competition among suppli- and bureaucratic measures by officials to reduce the ers, which is likely to result in lower prices for the likelihood of successful challenges. Finally, to the government. extent that foreign firms believe they are unlikely to Both selective and limited tendering create possi- receive a fair and objective hearing in a domestic bilities for effective discrimination against foreign challenge procedure, another form of discrimina- suppliers. In general, there are three forms of dis- tion is introduced into a nation's procurement sys- crimination against foreign suppliers: policies that tem. This increases the likelihood that the directly or indirectly reduce the number of bidders entry-based, cost-based, and price-based forms of (referred to here as entry discrimination); policies discrimination identified above will occur, again to reduce price competition for domestic bidders reducing the benefits that accrue to governments (price discrimination); and discrimination that from competition between potential suppliers. increases the costs of foreign bidders (cost discrimi- nation). A limited tender that deliberately involves Some Special Dimensions of Procurement few, or no, foreign firms is an example of entry dis- crimination. Establishing a "price preference" for If procurement involves services, cross-border trade domestic bidders by adding a specified percentage may not be feasible (see Chapter 32, by Sauvé, in to foreign firms' actual bids before they are com- this volume). A "commercial presence" will often be pared with domestic firms' actual bids is an example a prerequisite for bidding for a contract, leading a of price discrimination. The use of technical specifi- foreign firm to set up a subsidiary or a joint venture cations that must be met by bidders and that tend to with a domestic firm. The ability to establish a pres- be more costly for foreign firms to meet illustrates ence, and therefore to bid for government contracts, cost discrimination. is highly contingent on government policies toward 419 " B E H I N D - T H E - B O R D E R " A N D R E G U L AT O R Y I S S U E S foreign direct investment (FDI), joint ventures, and surveys). Thus, procurement rules--whether foreign mergers and acquisitions of domestic firms. national or domestic--should be designed to take The provision of many services also requires the account of the danger of corruption. Much can be movement of people across national boundaries. To done to force public entities to make their decisions effectively supply some services, such as accounting, in an open fashion, including banning firms found auditing, consulting, and legal services, only the to have engaged in bribery from bidding for govern- temporary movement of personnel may be ment contracts for a number of years (in addition to required. The ability of a foreign firm to contest the other financial and criminal penalties), protecting domestic service market--whether or not the buyer whistle-blowers, and conducting random and pub- is a government--depends critically on the nation's lished ex post audits of procurement decisions that policies toward temporary work visas and longer- explicitly take into account comparisons with simi- term work permits. lar transactions in the private sector. Cost Overruns and "Renegotiation" Economic Analyses of Procurement Discrimination There is one caveat to the presumption that a lower Discrimination is used for a variety of objectives, bid increases the chances of securing a state con- both economic (value for money) and noneconom- tract: a bid may be too low to be commercially ic (e.g., national security, infant industry promo- viable. While this may seem implausible, it is impor- tion). In assessing the effect of procurement tant to remember that firms often supply a good discrimination, the questions to be asked are: What well after the contract is awarded. If a firm is build- effect does the policy have on market outcomes-- ing a bridge for a government and the firm runs out does it work? Does it work in an efficient manner? of money (because its bid was below the cost of What is the effect on national welfare, which aggre- constructing the bridge), the procuring entity is in a gates the benefits that purchasers (including the bind. Does it begin the procurement process afresh, government) and suppliers derive from market or is it cheaper to pay the existing contractor addi- exchange? tional sums to complete the bridge--to bail it out? Most attempts to answer these questions have The possibility that suppliers may be bailed out will focused on the effects of procurement discrimina- affect the prices that firms bid in the first place. If tion in a single market or auction. Actions in one domestic firms are more likely to be bailed out than market, however, can have knock-on effects in other foreign firms, this confers an advantage on the for- markets; in procurement, as in other contexts, gen- mer. Procurement procedures should be designed eral equilibrium effects can be important. Two to take such factors into account and to ensure that examples highlight the importance of considering low bidders have the capacity and the incentive to the overall effect. complete the contract. First, suppose that a government decides to no longer buy a good from abroad. This will raise the relative price of that good because the additional Corruption and Collusion resources that domestic firms need to use to satisfy Given that both officials and firms may be motivat- the government's higher demand for their products ed by a desire to use the procurement process as a must be diverted from other productive resources. vehicle for creating rents, disciplining the scope for A procurement ban can therefore affect the prices of corruption and collusion is an important feature of production factors such as land, labor, and capital. government procurement. Firms may decide to The factor that gains (or that loses) from the impo- reduce the competition for contracts by "rigging" sition of a procurement ban is the one which is used the prices they submit to the government, with the most (or least) intensively in the production of the expectation that firms will takes turns in winning a good that the government is now buying solely contract or will divide the spoils. Firms could from domestic firms.1 Second, a ban may affect pro- attempt to bribe or otherwise induce government ductivity. Suppose that labor productivity (output officials to view their tenders more favorably; such per worker) and capital productivity (output per corruption has been extensively researched (see machine) rise when a firm's production level Bardhan 1997 and Rose-Ackerman 1999 for recent increases--that is, there are economies of scale. A 420 Multilateral Disciplines and Government Procurement ban on purchases of foreign steel shifts government sumers at lower prices. In this case, domestic con- demand to domestic steel producers, which increase sumers gain from the government ban. their output. This raises productivity, lowering the If government demand for cars does not exceed amount of labor and capital used to produce addi- domestic private sector supply at the time the pro- tional units of output, and increases the incentive curement ban is imposed, the imposition of a pro- for domestic steel production to expand exports as curement ban will have absolutely no effect. Assume well, since the incremental cost of production for that before the ban government demand is 45 cars, any market--including exports--has fallen. domestic private sector supply is 60 cars, and every- thing else in the example is as before. A ban will then merely reshuffle demand between foreign and Studies of Procurement Discrimination in Markets domestic producers. To see why, suppose that the The first influential analyses of procurement dis- government initially bought five cars from abroad. crimination examined the effects on outcomes in a After the ban, those cars must be supplied by domes- single market where there are two sources of supply tic firms, which in turn have five fewer cars to supply (domestic and foreign) and two sources of domestic to domestic consumers, but foreign firms have five demand (the government and private consumers). unsold cars that they can now supply to domestic The principal finding is that a procurement ban on consumers, making up any apparent shortfall on the foreign purchases by the government will affect out- supply side. Therefore, when government demand is put levels, imports, and prices only if government small in relation to domestic supply, a procurement demand exceeded domestic supply at the moment ban has no effect on overall sales, imports, output, or the ban was imposed. Suppose that in the market prices. (This striking result was first pointed out in for cars, before a procurement ban is imposed, gov- Baldwin and Richardson 1972.) ernment demand is 60 units, private sector demand Unfortunately, this clear-cut prediction--that a is 40, the domestic industry supplies 45 cars, and 55 future procurement ban will affect market out- cars are imported. The government must import at comes only if the level of government demand least 15 (60 ­ 45) cars. If a ban is now imposed, before the ban exceeds domestic private sector sup- there is excess demand for domestic cars on the part ply--does not necessarily provide information of the government, and domestic supply must about whether an existing procurement ban restricts expand to meet government demand. The excess market access to foreign firms. If government demand drives up the price paid by the government demand under a ban is less than domestic private for cars. Assume that at the higher price govern- sector supply, the ban is likely to have had no effect ment demand equals domestic supply at, say, 52 on market outcomes, including imports. If, howev- cars. In this case government demand is lower by 8 er, observed government purchases equal the quan- cars, domestic production has risen by 7 cars, and tity supplied by the domestic private sector, we do imports have fallen by at least 15 cars. not know whether the ban has distorted market The effect of discrimination on private con- outcomes. This equality could have prevailed before sumers depends critically on whether the initial the ban, and in that case the imposition of the ban price before the ban was determined by world prices would have had no effect on imports. Thus, the sim- or by domestic market forces. If the world price ple test does not always provide a reliable guide determined the price before the ban, domestic con- about the market access­reducing effect of an exist- sumers are unaffected by the procurement ban, ing procurement ban. since they do not need to pay the higher postban Evenett and Hoekman (2000) note that when price charged to the domestic government: they can procurement discrimination distorts market out- buy all the cars they want from world markets at the comes and domestic firms expand output and see established world price. In contrast, if foreign pro- their profits rise, new firms will enter the market, ducers customize their cars to the domestic market, driving the price paid by the government down the procurement ban leaves foreign producers with until the incentive to enter has abated. Whether an excess supply of at least 15 cars that they used to entry occurs depends on the prevalence of natural sell to the domestic government. Since these cars and policy-induced barriers to entry. Entry by were customized for a specific market, foreign sup- domestic entrepreneurs may be frustrated by rules, pliers must offer additional cars to domestic con- regulations, and licensing. The ability of foreign 421 " B E H I N D - T H E - B O R D E R " A N D R E G U L AT O R Y I S S U E S entrepreneurs to set up production facilities in the preference will be used to inflate those bids. They nation with the procurement ban depends on poli- respond to high price preferences by lowering their cies toward FDI. Therefore, the long-term conse- bids and reducing their profit margins. If the proba- quences of a procurement ban are a function of bility of a foreign firm being the low bidder is high, domestic competition and FDI policies. raising price preferences might actually reduce pro- curement costs. If the foreign firms have, on aver- age, lower costs than domestic firms, an increase in Procurement Discrimination in Auctions price preferences from zero will always reduce pro- Many government purchases are transacted not curement costs because the probability that a for- through markets but through bidding by suppliers eign firm will still win the contract is so high. for a contract. Each firm typically bids a price for Both McAfee and McMillan (1989) and Deltas completing the contract, given what it thinks other and Evenett (1997) found in simulations of auc- firms are bidding. Economists have found that out- tions that the possible cost reductions resulting right bans on foreign bids unsurprisingly reduce the from price preferences were very small (often less competition faced by each domestic bidder, which than 1 percent). Worse still, Deltas and Evenett then all submit higher bids. The government then found that if the government accidentally chose the finds itself choosing from a range of higher bids, wrong rate of price preference, these cost reductions and the overall cost of the contract rises. Thus, there quickly became cost increases--even with small is a strong presumption against forms of discrimi- errors. This is particularly important because poli- nation that exclude foreign bidders (McAfee and cymakers do not typically have access to all the rele- McMillan 1989). Similarly, any measure that vant information (including the actual probability increases the cost of one group of bidders results in distributions of the costs of the domestic and for- that group raising its bids (or perhaps not bidding eign firms) needed for choosing the optimal price at all). The other bidders, facing less competition, preference. Even if they did, political factors would raise their bids, too. Again, government's procure- probably "encourage" them to choose a higher- ment costs rise. In simulations of auctions, Deltas than-optimal price preference. Deltas and Evenett's and Evenett (1997) found that such cost discrimina- simulations imply considerable gains in the expect- tion against foreign firms can raise procurement ed profitability of domestic firms from even modest costs significantly, but less than outright prohibi- (5­10 percent) price preferences. In practice, the tions on foreign bidding. likely procurement cost reductions are very small One form of discrimination that is used by many and, given the realities of policymaking, will almost developing countries--price preferences--can certainly never be realized. This form of procure- actually lower procurement costs. McAfee and ment discrimination is bad policy, too. Still, when McMillan (1989) demonstrated this result in an the three forms of procurement discrimination are analysis of optimal procurement auctions. The sig- compared, price preferences tend to have the least nificance of this result is that it undermines the pre- adverse effect. Cost discrimination is more dis- sumption that a cost-minimizing procurement tortive than price preferences but less distortive official would always choose not to discriminate than outright bans on bids from foreign suppliers. against any bidder, domestic or foreign. To under- Unfortunately, the current plurilateral WTO agree- stand the intuition underlying McAfee and McMil- ment on procurement bans price preferences but lan's result, it is important to bear in mind two imposes much weaker disciplines on the other two points. First, in most analyses of auctions, if one forms of procurement discrimination--the more supplier raises its bid, then other suppliers have an distortive ones (Hoekman 1998b). incentive to raise their bids too. Second, price pref- erences only inflate the actual bids of foreign suppli- The WTO Agreement on Government ers when domestic and foreign bids are compared. If Procurement a foreign supplier's inflated bid is still the lowest, the government awards the contract to that foreign sup- The WTO Government Procurement Agreement plier but only pays the amount of the actual bid. (GPA), which entered into force on January 1, 1996, McAfee and McMillan's insight was to show that is a plurilateral agreement that binds only signato- foreign firms make their bids knowing that a price ries (Hoekman and Kostecki 2001). At the time of 422 Multilateral Disciplines and Government Procurement writing, the GPA had only 26 members: Austria, mitments on market access to that offered by other Belgium, Canada, Denmark, the European Union members of the GPA (Hoekman and Mavroidis (as an entity), Finland, France, Germany, Greece, 1997). This insistence on reciprocal access to pro- Hong Kong (China), Ireland, Israel, Italy, Japan, the curement markets reduces the benefits of enhanced Republic of Korea, Liechtenstein, Luxembourg, the competition in domestic procurement contracts. Netherlands, Norway, Portugal, Singapore, Spain, In addition to nondiscrimination, the GPA has Sweden, Switzerland, the United Kingdom, and the two other important features: measures to improve United States. Although China, Iceland, the Kyrgyz transparency in the procurement process, and Republic, Latvia, Panama, and Taiwan (China) are enforcement provisions. The agreement contains negotiating accession (in some cases as part of gen- numerous provisions on the procurement process, eral WTO accession), the current membership of covering the use of selected and invited tendering, the GPA is limited to high-income countries. the nature of technical specifications used in ten- ders, and the criteria for awarding contracts. For example, Article VI enjoins nations from creating GPA Principles and Disciplines "unnecessary" obstacles to trade when setting speci- The GPA establishes a framework of rights and obli- fications and calls for the use, "where appropriate," gations regarding parties' national procurement of performance rather than design standards. Con- laws, regulations, and procedures. Governments are tracts must be awarded to the supplier that is required to apply the principle of national treat- deemed capable of undertaking the contract and ment to the products, services, and suppliers of that has submitted the lowest bid or that best meets other parties to the GPA and to abide by the most- the evaluation criteria laid out in the original ten- favored-nation (MFN) rule, which prohibits dis- der. Essentially, the second requirement prohibits a crimination among goods, services, and suppliers of procuring entity from refusing to award a contract other parties. The agreement emphasizes trans- to a supplier on grounds that were not specified in parency of laws, regulations, procedures, and prac- the tender. The effectiveness of this requirement tices regarding government procurement. Five depends critically on the breadth of the evaluation annexes to the GPA specify, for each member, which criteria in the tender. government authorities (central government, sub- Transparency-related provisions emphasize pub- central governments, and other state bodies such as lication of tenders and notification of regulations to utilities) are covered and which purchases of goods the WTO. Ex post transparency norms are relatively and services are subject to multilateral disciplines. weak. The GPA allows a government to refuse to Only those transactions that exceed certain thresh- award a contract on "public interest" grounds, old levels are affected. Although a common thresh- which are not defined. Once a contract award has old of SDR 130,000 was established for central been made, the procuring entity does not have to government entities, thresholds for purchases of provide an explanation to the losing bidders; an construction services and procurement by noncen- explanation is required only when a member gov- tral government entities are often higher (see Hoek- ernment intervenes on behalf of a bidding firm man and Kostecki 2001). under the GPA's enforcement procedures. All goods purchases are subject to GPA disciplines The enforcement provisions of the GPA are par- unless a nation secured exemptions during the ticularly noteworthy. Recognizing that aggrieved negotiation of the agreement (or during subsequent bidders desire rapid and effective redress, the GPA accession to the agreement). By contrast, only those mandates that members establish impartial domes- services listed in the annexes are covered by the tic arbitration procedures capable of making swift GPA. Thus, as in the General Agreement on Trade in decisions. Complainants are allowed to invoke the Services (GATS), a "positive list" approach to cover- provisions of this WTO agreement in making their age was taken. Indeed, there is a very close corre- case in a domestic arbitration procedure. In addi- spondence between GATS specific commitments tion to these domestic procedures, members are and the coverage of the GPA. Although nondiscrim- allowed to bring cases to the WTO dispute settle- ination (MFN and national treatment) lies at the ment body. The effectiveness of this domestic core of the GPA, effectively banning price prefer- dimension to enforcing an international trade ences, many members have conditioned their com- agreement depends crucially on the manner in 423 " B E H I N D - T H E - B O R D E R " A N D R E G U L AT O R Y I S S U E S which national adjudication is carried out. There system. Srivastava (2000) explored the possible will inevitably be differences in such procedures effect of membership for India and concluded that among nations, as well as differences in the ability of membership could generate a welfare gain of adjudicators to resist pressures to favor domestic between 0.3 and 1.7 percent of GDP (see Box 40.1). suppliers. Options for Reform Effects of Membership All governments--whether industrial or develop- Few economic studies have been made of the actual ing--have an interest in establishing efficient, trans- or expected impact for developing countries of parent, and accountable procurement practices for signing on to the GPA. Most of the literature has the wide range of goods and services that are pur- focused on the European Union and has investigat- chased by the state. This observation alone suggests ed the impact of the EU's procurement directives. that governments should be unilaterally imple- Two recent studies do shed some light on the likely menting measures that encourage greater competi- effects of GPA membership for developing coun- tion between bidders, among other measures. tries. Choi (2001), who assessed Korea's experience Unilateral reforms, however, have foundered on the with accession to the GPA, notes that the share of rocks of strong vested interests in the status quo. A limited tendering in procurement by the Supply well-known finding in political economy is that the Administration of the Republic of Korea (SAROK), impetus for reform is undermined when the losers which accounts for about 60 percent of public pro- from reform (in this case, domestic suppliers to the curement, fell from 27 percent during 1993­-95 to government and, possibly, corrupt officials) are 22.5 percent in 1996­98 (Table 40.1). He also concentrated and the benefits of reform (in this observes that cost savings can be used as a proxy case, those to taxpayers) are diffuse. Multilateral measure for the impact of the new procedures on trade reform, by offering the prospect of greater efficiency. Using SAROK data on the price differen- market access to foreign procurement markets, tial between the market price and the actual con- alters the calculus of reform by encouraging lower- tract price of goods as the cost-saving ratio, Choi cost domestic firms to weigh the potential loss of found that the average ratio in 1993­95 was 5.8 per- domestic market share against prospective increases cent for domestic goods purchased and 18.5 percent in overseas sales. In principle, multilateral initiatives for imports.2 The ratio increased to 8.5 percent for can divide the domestic vested interests that oppose domestic goods and 23.1 percent for imports dur- unilateral reforms. ing 1996­98, which Choi interprets as reflecting an Although, as noted in Box 40.1, GPA membership improvement in the efficiency of the procurement is likely to translate into lower procurement costs, Table 40.1 Performance Indicators for Government Procurement in Korea, 1993­98 (percent) Share of Cost-saving Cost-saving Share of limited margin on margin on imports in goods Year tendering domestic goods imported goods purchased 1993 26.5 6.4 21.3 15.2 1994 27.0 5.8 19.1 16.2 1995 28.1 5.3 15.1 15.9 1996 22.6 6.4 29.4 16.6 1997 21.9 7.8 25.1 12.1 1998 23.1 11.2 14.7 9.4 Average, 1993­95 27.2 5.8 18.5 15.9 Average, 1996­98 22.5 8.5 23.1 12.7 Source: Choi (2001), based on data from the Supply Administration of the Republic of Korea (SAROK). 424 Multilateral Disciplines and Government Procurement B O X 4 0 . 1 A C C E S S I O N T O T H E G PA : I D E N T I F Y I N G C O S T S A N D B E N E F I T S There are two potential sources of benefit for sig- ences imply that India's policies are at odds with natories to the GPA. First, as a result of the trans- the GPA's national treatment requirement. parency requirements and the competition Although there is likely to be a large potential for resulting from the agreement, the tax-paying pub- the export of software services and possible gains lic could receive better value for money spent by from South-South trade, the analysis that follows the government on its purchases. Second, export ignores any such potential gains and focuses only markets could expand as a result of purchases by on the domestic impact of GPA membership. Esti- governments of other member countries. The gen- mates of the total value of government expendi- eral perception is that the second source of gain is ture, including that by state governments, above likely to be small for developing countries. It is GPA thresholds range from 0.8 to 1.5 percent of therefore important to estimate the potential gains GDP, and estimates for public sector enterprises from achieving transparency in procurement pro- range from 3.4 to 5.7 percent of GDP. The estimat- cedures and from nondiscrimination in the pur- ed social gains from GPA membership depend on chase of goods and services. As noted above, assumptions regarding the size of savings of public economic theory does not provide unambiguous funds and the relative weight attached to gainers. answers; there are cases in which discrimination is These estimates range between 0.1 and 0.6 per- welfare enhancing. What follows assesses the likely cent of GDP. Drawing on work by Domberger, costs and benefits of GPA membership. Hall, and Lee (1995), Hoekman (1998b) states that India has a central government, state govern- competitive tendering and outsourcing could pro- ments, and three tiers of local government, at the duce savings of about 20 percent without compro- village, intermediate, and district levels. In addi- mising quality. Transparency International (1997) tion, there are state-owned enterprises. No single notes that noncompetitive procedures may uniform law governs procurement by all these increase costs by about 30 percent. In a recent entities. Before 1991, public sector enterprises study reported in World Bank (2000e), 51 percent (PSEs) were protected from competition through of respondents indicated that in securing govern- reservation policies and requirements mandating ment contracts, bribes were paid always, usually, that central government departments and other or frequently, with payments amounting to 2 to 9 PSEs apply price and purchase preferences in percent of the contract value. favor of these firms. Since then, the list of indus- Results assuming a 30 percent saving are present- tries reserved for the public sector has been sig- ed in the table. To lend further perspective, the find- nificantly reduced, and the system of price ings are also expressed as a percentage of the fiscal preferences, including preferences for indigenous deficit. These calculations show that the social gain firms, has been discontinued. from GPA accession range from 0.34 to 1.7 percent The small-scale sector continues to be protect- of GDP. If these numbers could be translated into ed through compulsory purchase and price pref- savings for the central government, the fiscal deficit erences of 15 percent on procurement by central could be reduced by 8.4 to 42.5 percent, depend- government ministries, departments, and PSEs. If ing on what entities are included. These benefits the competition is with the duty-inclusive price of need to be compared with the costs of signing the imports, the implied preferential margin is higher GPA. The procedural requirements resulting from with respect to the duty-free price of imports. GPA membership are unlikely to lead at the margin Between 1994­95 and 1998­99, purchases by to any significant additional expenditure, since gov- the Directorate General of Supplies and Disposals ernment entities in India are already subject to com- (DGS&D) from the small-scale sector made up plex rules and procedures. Although there will be a about 8 to 10 percent of total purchases. Thus, one-time switchover cost, any net additional recur- despite important reforms of procurement proce- ring cost is likely to be small. In view of this, imple- dures, the remaining price and purchase prefer- mentation costs are not an important concern. (continued) 425 " B E H I N D - T H E - B O R D E R " A N D R E G U L AT O R Y I S S U E S B O X 4 0 . 1 ( C O N T I N U E D ) Estimates of the Social Gain from GPA Membership Entities Billions As percentage As percentage of involved of rupees of GDP fiscal deficita Central government only 67.5 0.34 8.4 Including state governments 90.3 0.45 11.3 Including public sector enterprises 340.6 1.70 42.5 a. Using the budgeted fiscal deficit of the central government for 1999­2000. The government is currently using preferential purchases involved a price preference of 15 per- treatment in government purchases as a tool of cent (which, in fact, they do not), the total "sub- industrial policy. In keeping with the spirit of sidy" to the small-scale sector would be about reform, there is no need to continue with the 1.2 to 1.5 percent of total expenditure. This preferential treatment of PSEs. Discontinuing could easily be paid out as an explicit subsidy preferential treatment for the small-scale sector, through a suitable mechanism. The other sectors however, may not be feasible politically. The indi- receiving preferential treatment are very small, rect subsidy to this sector is not very large; in and exceptions for them could be sought. 1999 the share of purchases by the procurement agency from the small-scale sector was no more Source: Prepared by the volume editors, based on Sri- than 8 to 10 percent of total purchases. If these vastava (2000). the agreement falls short of its potential for two rea- tougher penalties for graft by public officials and sons. First, the GPA bans price preferences, the least restrictions on postretirement employment oppor- (economically) costly and most transparent form of tunities for officials. procurement discrimination. Provisions on cost or As mentioned, the GPA has few developing coun- entry discrimination are far more opaque and allow try members. In part, this reflects a desire to be able for the use of local content criteria by developing to apply policies of discrimination and a concern countries in awarding contracts. Tightening up on that the gains from membership will be limited cost and entry discrimination and channeling pro- (Hoekman 1998b). Since the late 1990s, discussions tectionist use of policies toward observable price in the WTO have begun to focus on transparency, as preferences would improve the rules of the game opposed to market access. At the 2001 ministerial from an efficiency perspective and would provide a meeting of the WTO in Doha, agreement was focal point for negotiators (Hoekman and reached to launch negotiations on transparency in Mavroidis 1997). government procurement in 2003. Although not Second, the provisions for disciplining bribery explicitly linked to efforts to reduce corruption and and corruption are too weak. A reformed GPA bribery, transparency in procurement practice is might require signatories to delegate contract award intended to have this effect. decisions to independent public servants or agen- In setting priorities for reform efforts, a good case cies (where possible); to take steps to protect exists for tackling bribery and corruption first and whistleblowers; and to create an independent audit- discrimination second (Evenett and Hoekman ing or investigative office, which could be located in 2000). Graft probably always reduces welfare, under the legislative rather than the executive branch of any circumstances. By contrast, when the quantity government. These steps would complement demanded by the government is initially below the 426 Multilateral Disciplines and Government Procurement domestic industry's output, procurement discrimi- nation merely reshuffles sales from foreign to domestic firms in markets and has no consequences for national welfare. If political pressures are such that the elimination of procurement discrimination is not feasible, the WTO should be geared toward making discrimination as transparent as possible. This may involve encouraging price preferences on foreign bids instead of adopting measures that increase the costs, or reduce the number, of foreign supplies. More generally, the focus should be on increasing the transparency of procurement prac- tices of WTO members. Notes I thank Bernard Hoekman for helpful comments and suggestions. Any remaining errors are mine. 1 If the country imposing a ban on foreign procurement is large, its action may affect the terms of trade. If the ban raises the relative price of, say, steel and the nation was exporting steel, the ban will raise the relative price of steel for all purchasers-- including foreigners--who will then buy less. Thus, the pro- curement ban would be equivalent to an export tax. If the nation was importing steel before the ban was imposed, the increase in the relative price of steel reduces total (government plus private sector) domestic demand and thus reduces imports of steel. Note, however, that most developing nations are unlikely to be able to affect their terms of trade in this manner. 2 SAROK uses the wholesale price as the market price. Imports refers to goods procured directly from overseas with foreign currency, not imported goods procured in the domestic mar- ket with local currency. 427 41 J O H N S . W I L S O N Standards, Defining Terms Product and process standards Regulation, Standards are most commonly associated with the specifica- and Trade tions and characteristics of par- ticular products (auto parts, toasters, computer monitors, and so forth), services (hotel WTO Rules and Developing registration systems, accounting practices, medical qualification Country Concerns requirements, and so on), and materials (chemical polymers, A wood content in furniture, and lthough traditional trade barriers so forth) Process standards, in contrast, specify such as tariffs continue to decline, manufacturing or quality control measures to be technical and regulatory barriers are increasingly taken to ensure that product quality is maintained. employed to block trade. There has been a rise in the This might include, for example, the specifications use of technical regulations such as standards as on how an automated assembly line in a factory instruments of commercial policy in multilateral, producing car bodies is constructed. Standards may regional, and global trade. These nontariff barriers are be codified in written specifications or followed by of particular concern to developing countries, which custom in manufacturing processes. They are may incur additional costs in meeting such mandato- developed in various ways, with the most effective ry standards that apply to production for export mar- being created through private sector­driven sys- kets. In regard to voluntary standards, such as those in tems. In many industries, including information the International Organization for Standardization technology, product standards are developed (ISO) 9000 series on quality, developing nations face through a voluntary consensus of companies constraints in absorbing best-practice information on engaged in producing competing products. In norms and mobilizing the resources necessary to addition to voluntary or industry standards, regu- adopt appropriate process and production methods. latory standards are mandated by governments. Domestic regulations that affect imports through These regulatory standards are developed to meet technical requirements, testing, certification, and health, safety, or environmental objectives. In many labeling represent an important new area of empha- cases, these standards involve testing and certifica- sis in continuing liberalization efforts. For example, tion requirements. domestic regulatory systems may restrain trade and limit market entry through environmental, health, or Standards-Developing Bodies safety requirements that are not based on interna- Around the world, numerous standards-developing tional norms. bodies operate at the national level. They include 428 Standards, Regulation, and Trade: WTO Rules and Developing Country Concerns industry associations, professional societies, special- such regulations could produce social losses that ized standards-developing bodies, and third-party outweigh any economic efficiency gains. certifiers such as Underwriters Laboratories. Inter- Standards are designed to facilitate information national organizations are also involved in the exchange, ensure quality, and achieve the provision development of product standards; the most famil- of public goods. For example, emissions standards iar of these are the International Organization for and fuel economy requirements may contribute to Standardization (ISO) and the International Elec- cleaner air. Whether they are the most efficient trotechnical Commission (IEC), which develop vol- instruments for meeting this objective is a second- untary consensus standards, such as the ISO 9000 order question. Sanitary and phytosanitary require- series on quality and the ISO 14000 series on the ments can improve health and quality of life, with environment. These management systems stan- spillover benefits in higher productivity. In a similar dards are not mandatory, and there are no formal manner, interoperability standards leverage the mechanisms for requiring their adoption, but they gains from telecommunication and information can be used by governments to block imports. technology networks. As noted in Kindleberger (1983) and Casella (1996), it is possible to think of standards as public goods themselves, for a standard Conformity Assessment and Product Certification is available to multiple users or competitors. The existence of a standard does not guarantee Standards can improve information flows final producers or consumers that a product func- between suppliers and consumers regarding the tions as indicated in the technical specifications in characteristics and quality of products, thereby a standard. Product testing, plant inspections, and facilitating market transactions. The process of other procedures are conducted to determine standardization may reduce the costs of uncertainty whether a product conforms to those specifica- (as measured by time and effort devoted to search) tions. The conformity assessment usually involves that consumers face in assessing product quality several steps and is conducted by an authorized (Jones and Hudson 1996). Standards facilitate com- third party able to certify that a product meets parison by consumers across products with com- detailed technical specifications. Governments and mon essential characteristics. They also increase the consumers are increasingly demanding such certifi- elasticity of substitution among similar products cations of goods in international commerce. Certi- (Harrison, Rutherford, and Tarr 1996). By permit- fication involves testing a product against a ting producers to settle on a limited range of prod- voluntary, de facto, or regulatory standard and is uct characteristics or processes, standards and often carried out by organizations that have no link regulations can promote economies of scale, partic- to the manufacturer or purchaser. After testing, a ularly by forcing inefficient firms to close down or certificate is issued confirming that the product merge with stronger enterprises. They also provide meets a set standard. guidelines or focal points around which firms can organize their own quality or performance stan- dards (Maskus and Wilson 2001). The Role of Standards The implementation of standards involves costs. Standards and regulations differ fundamentally Some of these costs are inevitable; they arise from from taxes and quotas on trade. Those classic trade the testing and certification (conformity assess- barriers are inefficient and discriminatory taxes on ment) procedures necessary to determine whether a foreign sources of economic activity that raise costs product, such as fresh fish or produce, meets stan- to consumers and input users, inefficiently allocate dardized requirements (Box 41.1). Inefficient and resources, and protect entrenched domestic market duplicative testing and certification requirements, power (Maskus and Wilson 2001). Economic analy- however, represent unnecessary and significant costs sis demonstrates that countries benefit mutually to manufacturers, consumers, and society. In partic- from negotiating their removal or reduction. Regu- ular, the cost and complexity of determining con- latory policies, by contrast, exist, in principle, to formity with varying national technical regulations achieve important objectives that would go under- is high and rising (Wilson 2000b). This is especially served in the private market, such as protection of true with regard to the growth of mandatory third- public health or the environment. Elimination of party certification of conformity with technical 429 " B E H I N D - T H E - B O R D E R " A N D R E G U L AT O R Y I S S U E S requirements. Among the most important costs are multilateral negotiations. Total removal of technical those associated with mandates that industry retest regulations to trade would not necessarily achieve and recertify products that have already been tested efficiency gains sufficient to overcome losses from for conformity with similar standards in multiple reduced social protection. markets. This results in higher costs to consumers The historical record does show, however, that in and marks a key point at which trade policy inter- practice countries may regulate for purposes other sects with standards systems and regulation. than social protection. Technical regulations may In principle, standards and regulations are directly discriminate against foreign suppliers, both in their aimed at overcoming market failures. Indeed, they design and in their outcomes, and may be used to may be the least trade-restricting policies available gain strategic trade advantages for domestic firms for regulatory purposes. Under some circumstances, over foreign competitors. Standards are often non- standards may expand trade. In consequence, it is transparent and in some cases needlessly force firms not clear that the trade impacts of technical barriers to duplicate testing and certification costs. Regula- are inefficient or that they should be the subject of tions may be drafted to exclude both domestic and B O X 4 1 . 1 S P S M E A S U R E S : T H E K E N YA F I S H E X P O R T S C A S E In January 1998 the European Union (EU) banned lion, representing about 2 percent of the coun- the importation of fresh fish and fish products from try's total commodity exports. There were, more- Kenya, Mozambique, Tanzania, and Uganda, to over, fears that the ban could be extended to safeguard EU consumers from the risk of cholera. fresh fruits and vegetables, another of Kenya's The ban was still in effect in July 2000. This action leading nontraditional exports. The EU imposes was taken without regard to the disciplines of the controls that subject imported fruits and vegeta- WTO Sanitary and Phytosanitary Measures (SPS) bles to 10 percent sampling for microbial control, agreement, which provides that if a member is to and this requirement is already affecting Kenya's apply SPS measures, it has to prove scientifically exports of the products. This development is dis- that the product in question poses a real threat to turbing considering that in these areas Kenya has the health of consumers. The agreement requires a comparative advantage. (The EU has agreed to that assessment of the risk be carried out on the extend by one year the July 2000 deadline for basis of techniques developed by relevant interna- imposing new maximum pesticide residue levels tional organizations (if these exist). This is to on horticultural exports from the African, ensure that such action is not based merely on Caribbean, and Pacific countries covered by the fears or conjecture but that there is sufficient scien- Cotonou Convention.) tific evidence. Even after the risk assessment has The EU action caused considerable losses in the been conducted and sufficient evidence has been fish industry. The Kenyan Ministry of Health, gathered, an opportunity must be given to the which is the competent authority, needed to exporter to put in place measures that eliminate embark immediately on an action plan to address the health risk, and a timeframe for compliance the concerns raised by the EU. The SPS measures must be set. In the case of the African fish, the EU and quality assurance procedures outlined by the made it clear that the ban was not based on scien- EU have to be strictly adhered to in order to tific evidence but was, rather, motivated by the restore confidence. This requires capacity build- lack of a credible system in Kenya to safeguard the ing both by the Kenya Bureau of Standards, products from possible contamination. The EU where the sampling of food exports for contami- clearly stated that if this were not changed, the nation takes place, and by the Ministry of Health. products would be shut out of the EU market. Fish is a leading nontraditional export, and in Source: Prepared by the volume editors, based on 1994 Kenya exported fresh fish worth US$50 mil- Mwega and Muga (2000). 430 Standards, Regulation, and Trade: WTO Rules and Developing Country Concerns foreign entrants into a particular market, so as to smaller effect of international standards might support entrenched monopolies. Finally, standards reflect a tradeoff between more trade, based on may be stronger than necessary for achieving opti- higher economies of scale, and less trade, associated mal levels of social protection, thus imposing exces- with reduced product variety. sive costs on consumers and reducing net welfare. Otsuki, Sewadeh, and Wilson (2000) estimated A process for rationalizing costly technical regula- the impact of changes in the EU standard on afla- tions can therefore be beneficial. Such rationaliza- toxin levels in food using trade and regulatory sur- tion involves ending discriminatory treatment, vey data for 15 European countries and 9 African removing duplicative testing requirements, recog- countries between 1989 and 1998. The results sug- nizing that foreign standards could achieve the gest that the implementation of new aflatoxin stan- same level of social or consumer protection as dards in the EU would have a negative impact on domestic standards, making regulation more trans- African exports of cereals, dried fruits, and nuts to parent, and scaling regulations at levels that do not Europe. The EU standard, which would reduce impose excessive costs on consumers and firms. health risk by approximately 1.4 deaths per billion a year, could decrease African exports by more than 60 percent, or US$ 670 million, as compared with Quantifying the Trade Effects of Technical regulation based on an international standard. Barriers and Standards There is suggestive evidence that standards often There is considerable interest among policymakers act to raise costs and thereby serve to restrain trade. in empirical work on the impacts of technical regu- The OECD (1999b) found that the costs of meeting lations on trade. This interest reflects the belief that differing standards and technical regulations in its regulations often constitute important nontariff member nations, along with the costs of testing and barriers to trade and that their use is proliferating. certification, can amount to between 2 and 10 per- Many observers claim that the trade-restricting cent of overall product costs. Research stimulated effects of technical regulations in industrial coun- by the EU Single Market program in the mid-1980s tries are particularly costly for exporters in develop- illustrated how significant such standards-induced ing nations. These arguments arise from anecdotal trade barriers can be. A typical example was build- examples and case studies. ing tiles, where voluntary industry standards dif- Several recent studies have related trade flows to fered by EU country. Spain was found to be the measures of a country's standards. Swann, Temple, lowest-cost producer of such tiles, average prices and Shurmer (1996) and Moenius (1999) discussed being between 40 and more than 100 percent lower the multiplicity of economic hypotheses about than prices charged by producers in countries such trade and standards and note that standards can as Germany, France, and the Netherlands (Hoek- increase or reduce trade. Swann, Temple, and man and Kostecki 2001). Such price differences Shurmer regressed U.K. net exports, exports, and were the result of a combination of differing stan- imports over the period 1985­91 on counts of vol- dards and government procurement regulations. In untary national ("idiosyncratic") and international France nonstandard tiles could not be used in pub- standards recognized by the United Kingdom and lic works (about 40 percent of the market), and pri- Germany. Standards counts were taken from the vate firms were hesitant to use nonstandard tiles PERINORM database and were matched with the because insurance companies tended to require that three-digit Standard Industrial Classification (SIC) buildings meet industry standards. Pasta is another of the industry they applied to. The authors con- example; Italy's pasta purity laws required that pasta cluded that U.K. standards had a positive effect on be made of durum wheat, a high-quality type pro- both exports and imports; the standards served duced in the south of the country. This increased both to signal quality abroad and to raise import the cost of pasta in comparison with that in other demand in the United Kingdom. Swann, Temple, EU countries, where pasta tended to consist of a mix and Shurmer also concluded that German stan- of wheat qualities. dards reduce U.K. exports, perhaps suggesting a It has been estimated that over 60 percent of U.S. protective impact. Finally, they found that idiosyn- exports are subject to health, safety, and related cratic standards increased trade more than did standards in their destination markets. Govern- international ones. They hypothesized that the ment-issued certificates were required for 45 per- 431 " B E H I N D - T H E - B O R D E R " A N D R E G U L AT O R Y I S S U E S cent of exports to the EU; private, third-party certi- countries' SPS requirements, claimed that develop- fication was accepted for 15 percent; and manufac- ing countries are strongly constrained in their abili- turers' self-certification sufficed for the rest (Wilson ty to export food products by SPS mandates abroad. 1998). Certification in regulated sectors may involve Such rules were ranked as the most significant con- frequent and redundant sampling of products and straint on exporting agricultural and food products testing for conformity to standards. Some products to the EU, ranking ahead of transport costs, tariffs, may be subjected to 100 percent testing, which can and quotas. effectively block imports if applied only to foreign Maskus and Wilson (2001) discuss the results of firms. Duplicative testing and certification require- several partial equilibrium studies of the effects on ments have rapidly become more important as a prices and trade flows of particular food safety stan- barrier to international trade. Unter (1998) esti- dards. The studies they consider generally find that mates that redundant testing and conformity product standards amount to an equivalent tariff assessment procedures faced by the Hewlett- that exceeds legislated tariff rates. Finally, Messerlin Packard Company increased sixfold between 1990 (2001) provides extensive discussion of the EU's and 1997. sanitary and phytosanitary (SPS) standards and The United States International Trade Commis- technical regulations in manufactures. He argues sion (USITC 1998) interviewed corporate execu- that the EU's ban on imports of beef treated with tives, officers of trade associations, and government growth hormones has no basis in science as a safety officials regarding the importance of standards as measure but that it is a convenient way to block trade impediments in the information technology competition to domestic cattle farmers. (IT) industry in the United States, the EU, and vari- Box 41.2 provides examples of difficulties that ous countries in Asia and Latin America. Many IT Indian producers have faced in complying with firms considered duplicative and discriminatory standards in various export markets. testing and certification requirements to be a sub- Adoption and rationalization of international stantial barrier to trade, and some representatives of regulations can produce significant net welfare and the U.S. industry claimed that standards-related trade benefits. This is particularly true for develop- costs were the most significant trade restriction in ing countries that confront the challenge of meeting the industry. It is costly to meet multiple conformi- differing, duplicative, or discriminatory standards ty assessment procedures and labeling require- with their scarce resources. An international focus ments, and the delays involved are costly as well. For on rationalization of technical regulations world- example, meeting the EU's tests for telecommunica- wide would involve a commitment to the following tions equipment was estimated to take six to eight (Maskus and Wilson 2001): weeks, reducing product value by 5 to 10 percent. In many instances the requirement of a standard · Ending discriminatory treatment quality prohibits trade altogether. For example, an · Eliminating duplicative testing requirements EU regulation requires that dairy products be man- · Recognizing that foreign standards may achieve ufactured from milk produced by cows kept on the same level of social or consumer protection as farms and milked mechanically. This rule precludes domestic standards imports from many developing countries, especially · Making regulation more transparent those with numerous small producers for whom · Scaling regulation at levels that do not impose mechanization is too costly (World Bank 2000a). excessive costs on consumers and firms. The EU invoked this regulation recently to block imports of Mauritanian camel cheese, a new prod- WTO Disciplines uct developed at considerable cost by a small enter- prise in that country. The EU also raised the issue The WTO does not require that members have that Mauritania is not free of foot-and-mouth dis- product standards, nor does it develop or write ease, although there is little scientific evidence to standards. Instead, the WTO rules in this area aim at suggest that camels (or, in particular, camel milk) ensuring that technical regulations, voluntary stan- can transmit the associated virus. dards, and testing and certification of products do Henson and others (2000), in a study of problems not constitute unnecessary barriers to trade. Two faced by developing countries in meeting industrial WTO agreements deal with product standards: the 432 Standards, Regulation, and Trade: WTO Rules and Developing Country Concerns B O X 4 1 . 2 S TA N D A R D S - R E L AT E D P R O B L E M S E N C O U N T E R E D B Y I N D I A N E X P O R T E R S The following are examples of standards-related should be appropriately treated to ensure measures perceived as nontariff barriers by Indian destruction of any pathogens. What should mat- exporters. ter is the quality of the final product, which may Aflatoxin in Peanuts. In 1999 the European be attained through a flexible systems approach; Commission imposed new tolerance limits for it is neither feasible nor desirable to standardize a aflatoxin contamination in peanuts, as well as a specific systems approach. Similarly, the EU new testing procedure based on three tests of requires that raw milk must originate from cows material drawn randomly from a 30-kilogram or buffalo which have a minimum daily yield of 2 sample. If any of the three tests is found to be liters of milk per day. In the Indian context, this is over the limit, the lot is rejected. The revised simply impractical. The animal health require- standard is stricter than that specified by the ments stipulated by the EU are also far in excess Codex Alimentarius. Experts have concluded that of the requirements laid down under the Interna- 75 percent of the lots rejected under the pro- tional Animal Health Code (IAHC) of the Office posed procedure would be below the established International des Epizootic (OIE). The IAHC does tolerance level (i.e., uncontaminated material); not include any conditions specifically related to that the average EU citizen is at risk of eating a milk and milk products for diseases such as nut contaminated with aflatoxin every 27 years; rinderpest, pestedes petits ruminants, bluetongue, and that the health risk of consuming such a nut sheep pox, and goat pox, but the animal health is negligible. attestation conditions laid down by the EU do. Mango Pulp. Increasingly, what is under surveil- Egg Products. In 1997 Company O, a Banga- lance is not just the end product but the process lore-based egg product­exporting company, sent of production.* In India, where most primary pro- egg powder to Japan, which reported finding duction takes place in small-scale units, providing excessive BHC beta isomer levels in the product. primary-level quality assurance is difficult. Thus, A composite sample was analyzed in laboratories for example, an EU requirement that records be in Bangalore and Belgium. Both labs concluded kept for each delivery by farmers to mango that the BHC level was below the detectable pulp­processing units is very cumbersome. The limit. Thus, there is a great deal of uncertainty EU justification for this requirement is that in case associated with laboratory test results, and there a consignment of mango pulp is found to be is no assurance that products which have tested harmful, the farmer whose mangoes were bad as fit for export in Indian laboratories will be able can be traced. It would be more efficient to to enter export markets. ensure that pulp processors observe strict quality Tires. Brazil requires that all imported tires be checks on suppliers, rather than burden them marked "En-Metro," Brazil's national standard for with the cumbersome task of maintaining tires. The process of obtaining En-Metro certifica- records on farmers. tion is costly; a Brazilian team of experts has to Milk Products. EU standards for milk and milk inspect prospective Indian exporters, and all the products require that inspection be done at the associated costs--averaging about $20,000-- level of primary production and lay down norms must be borne by the exporter. Certification is for animal care, types of feed, and the like, as well valid for one year only. This process effectively as monitoring standards. In India, with its large bars small companies from exporting to Brazil. population, a dairy holding may have just one or Similarly, in Mexico imported tires must bear two draft animals, and milk from a number of "Norm" certification, which is awarded not to such holdings is pooled before it is processed. It is the tire company but to each type of tire, which not possible to monitor each animal. Under this has to be separately tested. The Norm certificate situation, the quality is again determined at the costs between $40,000 and $50,000 per tire entry point of the processing unit, where milk type. (continued) 433 " B E H I N D - T H E - B O R D E R " A N D R E G U L AT O R Y I S S U E S B O X 4 1 . 2 ( C O N T I N U E D ) Steel. Because of objections by Australia and pointed out, however, that in the context of food safety New Zealand to the use of untreated wooden regulations, process standards can sometimes be the dunnage, an Indian steel company was forced to optimal regulatory option; that is, a hazard analysis and use treated wood or wood substitutes. These critical control point (HACCP) system, which includes were not only more expensive but also in short flexible process standards designed to reduce microbial supply. Australia and New Zealand also require contamination in food, might be superior to specific fumigation of containers, involving an additional product standards, given the expense of microbiologi- cost of US$400 per container. cal tests and the recurrent nature of the pathogen haz- Packaging, Marking, and Language Barriers. In ard. The costs of enforcement and the degree of some European markets texturized yarn must be administrative discretion in enforcement are also impor- supplied in equal-length packages, creating addi- tant considerations in any evaluation of the relative effi- tional costs for exporters. German norms for engi- ciency of process or product standards. neering products are problematic in that technical According to Article 2.1.1.20 of the OIE International regulations are rarely made available in English. Indi- Animal Health Code, which deals with the import of milk an exporters must employ translators, and the trans- and milk products from countries considered infected lations may not be accepted by German authorities, with foot-and-mouth disease (FMD) or rinderpest, the generating additional uncertainty for exporters. veterinary authorities of exporting countries are required to produce international sanitary certificates attesting * A process standard might alternatively stipulate that that these products originate from herds or flocks that all mango pulp must be processed using a specific were not subject to any restrictions due to FMD at the method. Economists usually argue that product stan- time of milk collection and that the products have been dards are more efficient regulatory tools than process processed to ensure the destruction of the FMD virus standards, since product standards allow heteroge- according to the procedures laid down in the article. neous firms to choose the technology that minimizes the resource costs of achieving a specific regulatory tar- Source: Prepared by the volume editors, based on Car- get, while process standards do not. It has also been dona and others (2000). Agreement on Technical Barriers to Trade (TBT), importing countries to implement SPS measures and the Agreement on Sanitary and Phytosanitary that diverge from international norms (if these Measures (SPS). The TBT agreement addresses exist) but requires that they provide scientific justi- "product characteristics or the related processes and fication for such measures and apply risk assess- production methods" reflected in technical regula- ment mechanisms. International SPS standards are tions and requires that these regulations conform to developed by several organizations, including the basic principles of transparency and nondiscrimi- Office Internationale des Epizootic (OIE), interna- nation. Relevant international standards developed tional and regional organizations operating within by bodies such as the ISO--if they exist--must be the framework of the International Plant Protection used as the basis for technical regulations unless this Convention (IPPC), and the Codex Alimentarius would be inappropriate because of climatic, geo- Commission.1 graphic, or technological factors. A principal objective of the WTO SPS agreement WTO Enquiry Points and Notifications is to minimize the negative effect on trade from the adoption and enforcement of SPS measures. WTO To increase the transparency of technical regula- members are encouraged to adopt internationally tions and SPS measures, WTO members are recognized standards but are free to apply stricter required to establish enquiry points that are standards. The agreement recognizes the rights of responsible for providing answers to all reasonable 434 Standards, Regulation, and Trade: WTO Rules and Developing Country Concerns questions from interested members and for mak- In the first five years of WTO operations, notifica- ing available relevant documents on the standards tions based on the SPS agreement rose substantially, that apply. They must also notify new and pro- reflecting increased concerns about food, animal, posed standards to the WTO SPS and TBT com- and plant safety (Figure 41.1). Since SPS measures mittees. As of 1999, as shown in Table 41.1, normally remain valid once enacted, the cumulative three-quarters of developing country WTO mem- number of active measures has grown rapidly. Note bers had established enquiry points under the SPS that the number of notifications based on the SPS agreement, whereas 92 percent of high-income agreement is, in general, increasing both in high- members had done so. income and in middle- and low-income countries. Table 41.1 Total Number of Countries That Have Established SPS Enquiry Points Country group 1995 1999 Middle and low income Number of countries 78 98 Number of countries with SPS enquiry points 49 74 Percentage of total 63 76 High income Number of countries 34 36 Number of countries with SPS enquiry points 28 33 Percentage of total 82 92 Source: World Bank calculations, based on WTO data. 41.1 Number of Notifications under the SPS Agreement by Region, 1995­99 500 450 400 350 300 250 200 150 100 50 0 1995 1996 1997 1998 1999 Africa and Asia CEECs West Europe Oceania North Latin Middle-East America America Source: Wilson (2000a): 7. 435 " B E H I N D - T H E - B O R D E R " A N D R E G U L AT O R Y I S S U E S Disputes under the TBT and SPS mechanisms can be used to contest the trade- Agreements restricting effect of standards that do not comply with WTO rules. The ability to do so, however, Over two dozen claims that countries violated pro- depends on countries' capacity to bring cases. visions of the TBT and SPS agreements arose between 1995 and 2000, including disputes regard- Conclusions: Dealing with Developing ing domestic regulations that affect the process and Country Concerns on Standards production methods for manufacturing goods. The United States and the EU, for example, have been This chapter has identified several problems facing debating draft EU regulations on waste from elec- developing economies in managing problems asso- tronic and electrical equipment that would affect ciated with standards. It is useful to restate those the use of chemicals and other inputs used to pro- concerns and recommend approaches for dealing duce a wide range of consumer electrical products. with them. In procedural terms, WTO cases ranged from One important issue is that relatively little is requests for consultations through panel and appel- known about the cost impacts of differing product late body rulings. Although most of the complaints standards and how they affect exporters in develop- brought to the WTO to date were initiated by indus- ing countries. In addition to the direct costs of meet- trial countries, developing countries exercised their ing prescribed process standards and certification rights in a number of instances. An illustrative and testing requirements, there are numerous costs example of a WTO case is reviewed briefly here. associated with variability in standards across export On April 7, 1997, the United States requested con- markets and over time. At the same time, adopting sultations with Japan concerning Japan's require- stronger regulations can enhance technology trans- ment on testing and confirming the efficacy of fer and raise confidence in the products of exporting quarantine treatment for each variety of eight agri- firms. Thus, a concerted effort to study the cost and cultural products: apples, cherries, peaches (includ- productivity impacts of meeting international man- ing nectarines), walnuts, apricots, pears, plums, and dates is important. The results of such studies could quince. Japanese regulations prohibited the impor- feed directly into international efforts to limit the tation of these products from the United States on trade-restraining impacts of proliferating and non- the grounds that they were potential hosts of the transparent technical barriers to trade. Of particular coddling moth, an agricultural pest. The import interest would be the extent to which mutual recog- ban could be lifted, however, if the exporting coun- nition agreements may divert exports from try proved that an alternative quarantine treatment nonsignatories out of traditional export markets or achieved the required level of protection. even preclude entry of firms in the first place. In 1987 Japan's Ministry of Agriculture, Forestry, Second, developing countries should recognize and Fisheries imposed the testing requirement that that the imposition of product standards and tech- was the subject of the dispute. The law prohibits the nical regulations is not solely a trade issue. Effective importation of products requiring quarantine treat- regulation within their own markets is important ment until the treatment has been tested on that for ensuring consumer safety and promoting tech- particular variety, even if the treatment has proved nical change. To establish and enforce appropriate effective for other varieties of the same product. The standards requires building expertise and devoting United States maintained that these measures vio- additional resources to applied science and public lated the SPS agreement. In October 1998 a report management. To a great extent this effort can be left by a WTO panel found that Japan had acted incon- to private firms wishing to expand domestic and sistently with the SPS agreement. Japan appealed international sales, but there remains a role for gov- the panel's ruling. In February 1999 the WTO ernment in light of the public-good nature of effec- Appellate Body upheld the panel's conclusion that tive standards. In defining and implementing more Japan's varietal testing requirement, as it applied to effective standards, however, many poor countries apples, cherries, nectarines, and walnuts, was main- will need technical assistance from international tained without sufficient scientific evidence (a organizations and specialists with expertise. requirement of the SPS agreement). This case and Third, the SPS and TBT agreements within the others like it illustrate that WTO dispute settlement WTO have set a bar that must be met by exporting 436 Standards, Regulation, and Trade: WTO Rules and Developing Country Concerns firms in developing countries. These agreements environmental standards is viewed with serious strongly encourage importing nations to adopt alarm by many countries with regard to both manu- product standards that are at least as rigorous as factures and agricultural products. Among other those developed by international standards-setting issues, the lack of clear rules on the appropriate use of bodies. Over time, all WTO members can be expect- labels to indicate environmental impact and the rise ed to adopt such regulations, with the richer mem- in the use of standards for process and production bers choosing even stronger rules. Thus, developing measures in industrial countries have been noted in economies have no choice but to meet recognized developing country submissions to the WTO. international standards, at least for exports. It is Questions of how and under what circumstances likely, however, that such standards would have to mutual recognition agreements (MRAs) are best be applied to all production within each country implemented to facilitate trade have also been simply to inspire confidence in importing markets raised. Such agreements are used to reduce the that goods are produced safely by all potential sup- trade-impeding effect of technical barriers through ply sources. mutual recognition of national product testing and In this context, problems relating to the imple- certification procedures. To date, they have only mentation of obligations under the TBT and SPS been negotiated between industrial countries, agreements rank high among developing country although both the TBT and SPS agreements concerns. Lack of modern technical infrastructure encourage all WTO members to enter into MRAs. and capacity to engage in international standards- Developing countries may use the WTO dispute development activities and to provide internation- resolution mechanism to raise concerns about ally recognized testing and certification procedures whether particular standards in import partners meet for products is a common constraint. Without the SPS and WTO rules. Inevitably, resort to dispute set- resources necessary for building and maintaining tlement will increase, given the proliferation and modern standards and conformity assessment sys- complexity of modern product mandates. This situa- tems, it is difficult either to ensure rights or to exer- tion likely means that WTO panels must give greater cise responsibilities under existing WTO rules. If voice to scientific evidence and representations by developing countries lack resources to access infor- members of civil society. Developing countries need mation on international standards or to participate to monitor the development of dispute settlement in in their development, a key link between the rule of this regard and assert their own interests. law as specified in the WTO system and developing It must be recognized, however, that the WTO countries' ability to fulfill their obligations and itself is not a standards-setting body; it has neither defend their rights is called into question. Many the expertise nor the resources for this purpose. developing countries support a targeted review of Ultimately, the real concern of developing countries the TBT and SPS agreements in light of develop- must be to influence the development of global ment needs. standards in ways that at least pay attention to their Many countries are also concerned to clarify pro- concerns. One way forward might be the creation of visions regarding special and differential treatment a global standards forum, as outlined in Box 41.3. in the TBT and SPS agreements. India, for example, has recommended extending the timeframe for Notes compliance by developing country members with This chapter draws on joint work with Keith Maskus and Tsunehiro the existing provisions of WTO agreements refer- Otsuki. encing standards. In a related vein, a number of 1 The joint Codex Alimentarius Commission (CODEX) of the developing countries have cited problems with their Food and Agriculture Organization (FAO) and the World ability to react to notifications of new TBT and SPS Health Organization (WHO) is one of the more important stan- measures. A notification of intent to promulgate a dards bodies, from the viewpoint of trade and development new regulation, with a 60-day open comment rule, (see ). The CODEX, which involves 165 mem- have no capacity to respond. ber governments, develops standards on food safety, pesticide residues on foods, food additives, veterinary drug residues, Concern over the use of environmental standards food contaminants, and labeling. The group has also devel- to restrict imports is also prevalent among develop- oped principles for food import and export inspection and cer- ing countries. The use of trade measures to enforce tification. 437 " B E H I N D - T H E - B O R D E R " A N D R E G U L AT O R Y I S S U E S B O X 4 1 . 3 M O V I N G F O R WA R D : A P R O P O S A L What is most needed in the standards area from a mandatory government testing and certification development perspective is a framework to sup- that can be moved to "declaration of conformity port national capacity building and improve the status" should be undertaken. A multilateral design of international standards. An action plan global conformity agreement could then be to bridge the standards divide and address the developed on the basis of this list for negotiation problems confronting developing countries in vol- and agreement at the WTO or in another, more untary standards requires support for infrastruc- appropriate forum. It is critical that developing ture modernization and enhanced access for countries benefit from and participate in such an those countries to standards-development activi- agreement. A plan to provide technical assistance ties. There is currently no coordinated internation- and funds to support mechanisms such as post- al framework for addressing critical development market surveillance systems in developing coun- needs in this area. A global effort is required to tries must be part of the effort. develop a framework for a targeted financial assis- In agricultural trade (food products) the lack of tance plan for modernizing the standards infra- progress toward harmonized, internationally structures of low-income countries and enhancing accepted standards has the potential for seriously developing countries' capacity to participate in eroding the gains made through removal of tra- international standards-development activities. ditional trade barriers. The wide range of differ- Innovative ways of achieving this goal, including ing sanitary and phytosanitary standards the use of global information technology net- imposed by importers--standards that lack a works, should be explored. foundation in sound science and are not based In addition efforts should focus on promoting on risk assessments--is particularly costly to trade expansion through regulatory reform and developing countries dependent on agricultural removal of mandatory technical barriers that are exports. A concerted effort to accelerate the cre- discriminatory. Such efforts are in the long-term ation of appropriate international standards economic interest of both industrial and develop- would help support continued expansion of agri- ing countries. One way forward is for developing cultural and horticultural trade and reduce the countries to endorse the wider use of "supplier's number of SPS-related disputes and problems declaration of conformity" to regulatory require- that exporters from developing countries con- ments. A systematic review of products subject to front. 438 42 B E R N A R D H O E K M A N K A M A L S A G G I Multilateral choice of firms (Feketekuty 2000; Low and Mattoo 2000). For developing countries, a policy of Disciplines and great interest in this regard is locational subsidies (tax conces- National sions) offered by high-income country governments to attract Investment or retain FDI. Bora, in Chapter 19 in this Policies volume, discusses the imple- mentation of the Agreement on Trade-Related Investment Mea- sures (TRIMs) negotiated in the I Uruguay Round. This agree- n the past 50 years, little has been ment basically prohibits measures that are inconsis- done to negotiate multilateral disci- tent with the GATT's national treatment principle plines on policies that affect factor movement. With (Art. III GATT) and its ban on the use of quantita- the exception of the General Agreement on Trade in tive restrictions (Art. XI GATT). The GATS goes Services (GATS), there are no disciplines in the WTO further by including establishment and national on policies pertaining to labor and capital movement. treatment for investors as commitments that signa- Some WTO members have argued that there is a need tories may decide to make for specific services to negotiate multilateral rules for investment policies, industries. An important issue for WTO members is such as the right of establishment and national treat- whether to extend the trading system through gen- ment for foreign investors. In part, these arguments eral rules regarding investment policies, and, if so, have to do with market access objectives. In many sec- what form such rules might take. The 1996 WTO tors the preferred mode of supplying a market may be ministerial meeting in Singapore led to the creation through foreign direct investment (FDI), not exports. of a working group on trade and investment with If FDI is restricted by the host country, foreign firms the mandate of examining the relationship between have an interest in rules that enhance or guarantee trade and investment policies. At the 2001 ministe- their market access.Another line of argument empha- rial meeting, in Doha, agreement was reached to sizes the potential payoffs to developing countries of initiate negotiations on investment policies at the signing on to multilateral rules as a commitment 2003 WTO ministerial meeting, if consensus exists device--as a mechanism to implement rules that gov- on the modalities of such negotiations. ernments want to adopt but are constrained from This chapter surveys the main arguments that have adopting because of political-economy factors been suggested for why developing countries should (Markusen 2001; Moran 1998). Yet another rationale support the creation of a multilateral agreement on for considering rules in this area is to ensure that investment. As in other areas, the answer may vary investment policies do not distort the mode of supply depending on country circumstances. A key chal- 439 " B E H I N D - T H E - B O R D E R " A N D R E G U L AT O R Y I S S U E S lenge from a developing country viewpoint is to tional agreement may help overcome resistance to ensure that any agreement on domestic disciplines FDI liberalization by protected industries. makes sense from the perspective of the national In addition to trade-related investment measures, development strategy and that disciplines extend to many countries apply licensing and approval policies that are harmful to developing countries. regimes and impose related red tape costs on for- A basic question is to identify the problem that eign investors. They may also prohibit entry international cooperation is supposed to resolve. It is through FDI altogether or may impose equity own- important to bear in mind in this connection that the ership restrictions. Such policies may reflect wel- value of sales by foreign affiliates of multinationals fare-enhancing attempts to shift foreign profits to has been growing rapidly, driven in part by declines in the domestic economy, or welfare-reducing rent- communication and transport costs and by unilateral seeking activities by domestic industries or govern- actions in many countries to privatize state-owned ment bodies. (See the section and readings on FDI enterprises and liberalize FDI regimes. Eagerness to in the CD-ROM that accompanies this Handbook.) attract FDI is reflected in the use of fiscal and financial The TRIMs agreement does not apply to such non- incentives to investors and in the proliferation of trade-related policies, nor does it affect service bilateral investment treaties (BITs). UNCTAD reports industries. The latter, however, are covered by the that as of 2000, over 1,600 BITs had been negotiated, GATS. As mentioned, the GATS extends to FDI as against 400 at the beginning of 1990. policies, in that countries can make specific market access and national treatment commitments for this mode of supply for any or all services. Conceptual Issues The current situation suggests a number of ques- Economic theory dictates that in the absence of tions. What is the payoff to seeking general invest- domestic market failures and externalities, the opti- ment rules, as opposed to expanding the coverage of mal FDI policy ought to be no policy at all--that is, the GATS? Much can already be achieved via the governments should allow for unfettered market GATS, as the agreement includes FDI as a mode of transactions. A rationale for restricting FDI supply, and services tend to be subject to higher FDI depends on the presence of domestic policy distor- restrictions than do manufacturing sectors. How sig- tions or market failures. Since multinational firms nificant are existing barriers to entry through estab- typically arise in oligopolistic industries, the exis- lishment (FDI) in nonservices sectors? What is the tence of imperfect competition is a possible motiva- effect of these barriers? Taking into account that tion for intervention by host country governments. restrictive FDI policies can be eroded in tradable Multinational firms wield considerable market industries by contesting the market through exports, power and will typically use it to extract rents from what is the relative payoff for trade liberalization the host economy. Theoretical analyses of content compared with investment liberalization? If the for- protection and export performance requirements mer is higher, this would suggest that priority should under conditions of imperfect competition (Rodrik be given to trade liberalization and related trade facil- 1987; Richardson 1991, 1993) illustrate that the wel- itation efforts. If trade barriers are low, domestic fare effects of such policies can be positive under industry will not have as large an incentive to support certain circumstances. In most situations, however, restrictive FDI regulations. (Restrictions on inward more efficient instruments than investment mea- FDI may be motivated in part by the existence of sures can be identified; for example, vigorous com- high trade barriers, as this provides an incentive for petition policies are better suited for encouraging tariff wall­hopping FDI.) Perhaps most important is competition (Bora, Lloyd, and Pangestu 2000).1 the question of why governments do not reform FDI Whatever the rationale of restrictive policies, the policies unilaterally. As explained in Box 42.1, reduc- available empirical evidence suggests that local con- ing restrictions on foreign entry through FDI can tent and related policies (on transfer of technology have very beneficial effects for developing countries. and joint ventures) are ineffective and costly to the A key issue is how a multilateral agreement could economy (Moran 1998).2 Furthermore, protected help achieve this if domestic forces are blocking FDI industries may create problems for future liberaliza- liberalization and a reduction in red tape. Another tion because they have an incentive to lobby against key question concerns the magnitude and effects of a change in regime. In such a scenario, an interna- policies that seek to attract or retain FDI. 440 Multilateral Disciplines and National Investment Policies B O X 4 2 . 1 W T O R U L E S O N F D I : A P O S I T I V E V I E W FDI has been subject to various types of policies on equity share, domestic content requirements, pro- the part of both host and parent countries, from duction or export requirements, and restrictions extremely negative ones such as nationalization or on remittances of profits. Many countries use expropriation to positive incentives such as tax screening, negative lists (sectors in which FDI is not holidays, as shown in the table. Relatively common permitted or is restricted), foreign equity caps, and negative incentives include restrictions on foreign limitations on landownership. Examples of FDI Policies Positive incentives Negative incentives 1. Tax holidays 1. Nationalization or appropriation 2. Tax treaties to avoid double taxation 2. Double taxation 3. Exemptions on import duties on capital 3. Domestic content requirement for goods and raw materials intermediate inputs 4. Other exemptions or relaxations of rules 4. Domestic employment restrictions in priority sectors 5. Export requirements 6. Screening 7. General foreign equity limits 8. Sectoral foreign equity limits 9. Landownership restrictions 10. Joint-venture requirement 11. Restrictions on remittance of profits 12. Limitations on transfer of shares or liquidation of the company There is, in principle, a fundamental similarity power will be self-constraining. If there is evi- between the case for freer trade in goods and the dence of predation, the remedy is competition or case for freer FDI. From a rule-making perspec- antitrust policy. Lack of an effective institutional tive, what is needed is to apply two sensible prin- framework for competition policy should not be a ciples that have been used in the trade policy justification for imposing restrictions on FDI. A setting. First, distortions should be handled by more open FDI policy may itself act as a catalyst the appropriate policy instruments that most for the development of these institutions. In any directly deal with the respective distortion. Sec- event, the first policy principle applies. ond, if there is no forceful theoretical support for Crowding out or scaling down of domestic the welfare-enhancing effects of a policy inter- entrepreneurship is another concern. Although vention, the benefit of the doubt is given to the such effects do arise, it is equally true that down- market, not to policy activism. That is, the pre- sizing of inefficient domestic firms is welfare ferred choice is no intervention. improving because of the associated rationaliza- These principles should apply to FDI policy, as tion and the increase in the choice of quality that well. Perhaps the single most important reason would be available to consumers. Those domestic for resistance to a more open FDI policy is the firms that are unable to undertake technological presumed market power effect of multinational innovation would be relegated to serving the corporations. A more open and transparent FDI lower end of the market, whereas firms that are policy, however, would invite not just one multi- willing and able to innovate would serve the high national but many and would thereby foster end, along with the foreign firms. Moreover, competition among the multinationals them- through mergers and acquisitions, and by infus- selves as well as between domestic and foreign ing new technology, FDI can prevent some firms. Hence, the scope for exercising market domestic industries from being wiped out. (continued) 441 " B E H I N D - T H E - B O R D E R " A N D R E G U L AT O R Y I S S U E S B O X 4 2 . 1 ( C O N T I N U E D ) FDI is a direct instrument of development and timetable must be set for removal of the restric- growth. Since growth strategy should vary from tions. This is where multilateral rules can help. country to country depending on factor endow- There is a yawning gap between FDI approvals ment, technology, and so on, FDI policy ought to and actual inflows in India. Since liberalization in be country-specific to some extent. In the Indian 1991, the ratio of actual to approved FDI has context, for example, this translates mostly into been no more than 25 percent An alarming sectoral prioritization. Infrastructural problems absolute decline in FDI in India since 1998 sug- continue to be India's biggest bottleneck, fol- gests that the existing incentive packages and lowed by the poor quality of the services sector vows of commitment are not enough to attract and shortcomings in the agriculture sector--lag- foreign investors. Further assurance and security ging modernization and availability of critical for foreign investors are needed and might be inputs. The priorities for FDI in India are, accord- obtained from WTO rules. ingly, relatively straightforward: the infrastructure Such rules should revolve around a most- sector (energy, transport and communications, favored-nation (MFN) code of conduct aiming cement, and so on) comes first, followed by the at gradual, time-bound removal of restrictions services sector (including the financial and insur- on FDI, with defined prioritization deadlines for ance industries) and agricultural machinery, different developing countries and with safe- chemicals, and fertilizers. guard provisions that allow for well-defined An effective prioritization scheme does not temporary deviations from free foreign entry, require a complex system of regulations and but on grounds of industry-specific ills only, not incentives. Reform should aim for simplicity--for on account of balance of payments or other example, the removal of many of the arbitrarily problems. set caps on foreign equity in different sectors. Any prioritization scheme carries the danger of Source: Prepared by the volume editors, based on Das allowing costly discretion for too long, and a (2000). International Spillovers positive externalities. An example is when FDI gen- erates technological spillovers for local firms, there- Investment-related policies may rationally attempt by making more efficient use of national resources.3 to shift rents (profits) from source to host countries There exists a large literature that tries to determine through measures that effectively tax investors. The whether host countries enjoy such spillovers. opportunity for this arises because FDI usually Spillovers may arise when local firms adopt tech- occurs in imperfectly competitive markets, and such nologies introduced by multinational enterprises policies can therefore give rise to spillovers. The through imitation or reverse engineering; when same is true for policies that encourage FDI. Clearly, workers trained by a multinational transfer informa- both types of policies can provide a basis for interna- tion to local firms or start their own firms; and when tional cooperation. What follows focuses primarily derived demand (both upstream and downstream) on incentive policies, as these are most obviously from multinationals leads to local provision of ser- potentially detrimental to developing countries. Dis- vices or inputs that are also used by local firms. tortions created by imperfect competition call for The empirical support for positive spillover competition policies (see Chapter 43, by Holmes, effects is ambiguous (see Chapter 34, by Saggi, in and Chapter 44, by Evenett, in this volume). this volume). Nevertheless, if governments believe From an individual country's perspective, incen- that there exists a solid economic case for promot- tives to attract FDI may be justified if FDI generates ing inward FDI via incentives because of positive 442 Multilateral Disciplines and National Investment Policies externalities, countries may find themselves in a between competition for FDI between developing bidding war for attracting FDI.4 This can be to the countries, which may be efficient, and locational detriment of the parties involved if it leads to exces- incentives used by industrial nations. The latter are sive payment to the investor--that is, transfers that much more likely to be inefficient because they exceed the social value of the expected spillovers. attract or retain industries that otherwise would The proliferation in the use of incentives for FDI locate in developing countries. Such incentive poli- suggests that this is an important possibility and cies, as well as complementary policies that protect that there may be a case for international coopera- industries which cannot compete (examples are tion to ban or discipline the use of fiscal incentives. restrictive rules of origin in regional agreements and Clearly, a key issue here is whether fiscal incen- antidumping), are prime candidates for discipline tives are effective. The empirical evidence on this through international negotiations (Moran 1998). issue, too, is far from clear. Many studies have con- The foregoing suggests there are valid reasons to cluded that incentives for inward FDI do not play an question the rationale for a multilateral agreement important role in altering the global distribution of that seeks to discipline all incentives. If incentives FDI (Wheeler and Mody 1992; Caves 1996). Others fail to alter the global allocation of FDI, restricting conclude that incentives do have an effect on loca- their usage has mainly distributional consequences. tion decisions, especially for export-oriented FDI In this case unilateral action to cease granting (see Guisinger and associates 1985; Hines 1993; incentives is the optimal policy. If incentives are Devereux and Griffiths 1998).5 When incentives do effective in altering location decisions, a case may not distort the global allocation of FDI, they basi- exist for subsidy freedom, since countries may be cally end up as transfers to multinationals. It is pre- able to signal important information to potential cisely when such incentives fail to attract FDI that investors. Developing countries, however, have an developing countries have the most to gain from unambiguous incentive to push for multilateral dis- committing to not using them. The case for cooper- ciplines on industrial country policies that have the ation under these circumstances is based mainly on effect of keeping firms from relocating to develop- distributional grounds. ing countries. A key need is to increase information If incentives do affect FDI, there may be an effi- on the use of incentives and analysis of their effects. ciency case against competition for FDI. It must be recognized, however, that competition for FDI via Spillovers due to Regional Integration incentives may actually help ensure that FDI goes to those locations where it is most highly valued. Some regional integration agreements (RIAs) extend Incentive competition may act as a signaling device the reach of national treatment to investors from that improves the allocation of investment across partner countries. Examples include the European jurisdictions by ensuring that FDI moves to where it Union (EU), where freedom of investment is a basic has the highest social return. It can do so in situa- principle; the North American Free Trade Agree- tions where investors locate in countries or regions ment (NAFTA); and various association agreements in which the social return to FDI is lower. In such that the EU has concluded with neighboring coun- situations governments should apply incentive poli- tries. Insofar as RIAs lead to discrimination between cies on a nondiscriminatory basis. insiders and outsiders in FDI policies, they impose In practice, locational competition is generally not negative externalities over and above whatever driven by information asymmetries that can lead to investment "diversion" occurs because of the prefer- FDI not flowing to countries where social returns are ential liberalization of trade barriers (see Chapter highest. This is the case, in particular, for efforts by 55, by Hoekman and Schiff, in this volume). Elimi- high-income countries to retain or attract FDI that nating this discrimination can be a powerful argu- would be more efficiently employed in developing ment in favor of multilateral rules. An important countries. Labor unions and groups representing the empirical question is whether such discrimination interests of local communities may oppose plant clo- occurs and how large it is. This is difficult to deter- sures and efforts by firms to transplant facilities. mine, as doing so requires careful and detailed Similar motivations underlie the use of trade policy assessments of the applicable legislation on both a instruments such as antidumping by industrial horizontal and a sectoral basis. Some agreements-- countries. It is important, therefore, to distinguish for example, the EU and some of the agreements the 443 " B E H I N D - T H E - B O R D E R " A N D R E G U L AT O R Y I S S U E S EU has in turn negotiated with neighboring coun- sals and policy uncertainty.6 To assess the relevance tries--embody a right of establishment for nationals of the credibility argument for an investment agree- of parties. Most RIAs only have disciplines of the ment, it is necessary to identify how much of what type found in bilateral investment treaties, which might be embodied in such an agreement can be require national treatment (often subject to excep- pursued and implemented unilaterally. tions in the form of negative lists) and which disci- Countries that are seeking to attract FDI can pline the use of performance requirements. already use a variety of existing credibility-enhanc- Given the role of regulation and the political sen- ing institutions. One is to commit to accept arbitra- sitivity associated with foreign ownership of many tion of disputes under the Convention on the services industries, one way of assessing whether Settlement of Investment Disputes between States RIAs have a discriminatory effect is to determine to and Nationals of Other States; by the International what extent they go beyond the GATS in eliminat- Chamber of Commerce (ICC); or by the UN Com- ing discrimination in services markets. Since FDI mittee on International Trade Law (UNCITRAL), will be a major mode of supply, the more RIAs go depending on the preferences of the investor. Some- beyond the GATS, the greater the potential negative times such commitments are embedded in RIAs spillovers. such as NAFTA. Another is to conclude bilateral Hoekman (1998a) argues that with the exception investment treaties with the major source countries of the EU, most RIAs do not go much beyond the for FDI. Countries that are in the market for credi- GATS. Most RIAs also do little toward effectively bility can also use existing WTO disciplines to constraining the ability of governments to provide schedule market access­opening policies for ser- incentives for FDI. The most far-reaching RIAs are vices (including the right of establishment) and to those involving the EU, which seek to apply com- bind low tariffs under GATT rules. There is great mon disciplines in areas such as antitrust, state aid, scope for developing countries to use the WTO as a and state monopolies. Periodic disputes regarding credibility-enhancing instrument; the coverage of the use of incentives by local governments to attract services commitments is often very limited, and tar- FDI and recurring claims of "social dumping" illus- iff bindings for merchandise imports are frequently trate that even the far-reaching EU disciplines are significantly higher than applied rates. Although insufficient to constrain the ability of governments credibility with respect to investment-related poli- to adopt the tax and factor market policies they cies can certainly be pursued via a multilateral believe will be most conducive to stimulating investment agreement, governments that are con- investment, be it foreign or domestic. vinced they have a need to use external instruments Insofar as RIAs cause negative investment to achieve such objectives could start by exploiting spillovers, these effects will be attenuated if the existing instruments much more fully. trade discrimination associated with RIAs is reduced by negotiating lower external tariffs and Issue Linkage and the Grand Bargain other trade barriers. The WTO process allows countries to define a nego- tiating set that allows a variety of potential tradeoffs Reputation and Policy Credibility and deals to be crafted. Because countries are As noted above, an important question is what a restricted to the equivalent of barter trade in multi- multilateral agreement can do that a government lateral trade negotiations, achieving a superior cannot do on its own. One possible answer is that a cooperative outcome may require that issues be multilateral agreement may help countries that seek linked (Hoekman and Kostecki 2001). Determining to attract FDI by acting as a signaling device or when such linkage is necessary and successfully instrument through which the perceived credibility designing globally beneficial packages is a difficult of a set of policies intended to foster FDI can be task, given that this occurs in the context of rent- enhanced. It is sometimes argued, for example, that seeking lobbying and often involves issues that are the countries of Central and Eastern Europe sought difficult to analyze. Insofar as developing country to conclude association agreements with the EU in governments are confronted with domestic con- part to overcome perceptions by foreign investors straints that inhibit the abolition of welfare-reduc- that these countries had a high risk of policy rever- ing restrictive FDI policies, engaging in investment 444 Multilateral Disciplines and National Investment Policies policy talks may come at zero cost and allow addi- Account must be taken, however, of the potential tional gains to be obtained through quid pro quo downside--issue linkage can be a two-edged sword. negotiations (see Box 42.1). Given that for most Efforts to expand the agenda to investment may developing countries capital exports through out- allow groups in society to seek cross-issue linkages ward FDI flows is largely a nonissue, a good case can in areas that could be detrimental to developing be made that the quid pro quo for accepting invest- countries. Bhagwati (1998) has argued that this ment-related disciplines should be sought outside Pandora's box possibility provides a powerful justi- the investment area. Within the investment policy fication for leaving general investment rules off the area, however, there are also important potential WTO agenda. The failure of the OECD to reach gains for developing countries, most importantly in consensus on a multilateral agreement on invest- the area of disciplines on the use of incentives by ment (MAI) illustrates the practical difficulties that high-income countries. will arise.7 If OECD countries, with their much While there is certainly scope for gains to be more uniform policy environment and similar obtained from an agreement on investment, the size goals, could not reach an accord, agreeing to a com- of the negotiating chips developing countries can mon set of multilateral principles on investment bring to the table will determine what is attainable. can be expected to be difficult. It should also be Developing country investment policies may not be noted that in the MAI negotiations no progress was regarded as a particularly valuable negotiating chip made in disciplining the use of investment incen- by other WTO members, especially nations that tives--one of the primary issues where developing already have liberal regimes. If so, other policies are countries stand to gain substantially. likely to be more powerful in inducing concessions The fact that the GATS includes establishment as from trading partners. Among these, further liberal- a mode of supply on which commitments can be ization of trade under existing agreements (GATT made should also be considered, given that FDI in and GATS) figures prominently. Investment policies services is more important in contesting markets may prove useful, but more may have to be brought than is FDI in goods (since goods can be traded). to the negotiating table by developing countries. Much remains to be done in liberalizing access to What determines the net payoff from agreement to services markets through establishment. Govern- negotiate on investment rules depends importantly ments may also be able to achieve much of what is on the constraints that developing country govern- beneficial unilaterally, by applying national treat- ments face in pursuing domestic reforms, whether ment and MFN principles and by adopting the right these reforms are consistent with what major WTO of establishment in national law. members desire in terms of multilateral rules, and That said, it is the case that the current architec- the extent to which industrial countries are willing ture of the WTO is asymmetric; there is no a priori to impose restrictions on locational incentives used rationale for incorporating FDI as a mode of supply by their local and provincial governments to attract in the GATS while excluding FDI in manufacturing and retain firms. and primary sectors. It is often emphasized that trade and investment have increasingly become complementary. Conclusion This suggests that there is a case for negotiating Negotiating a WTO agreement on investment poli- on investment. As is always the case in trade agree- cies may prove useful in arriving at a grand bargain ments, the devil is in the details. Much depends on that extends to issues of particular interest to devel- country-specific circumstances and on the objec- oping countries. This possibility must be considered tives of the government. These will differ signifi- carefully. A broader agenda will be necessary in any cantly across countries. event, both for countries that confront domestic In some cases the primary issue may be to political-economy constraints on the adoption of improve FDI regulatory policies and to reduce red better FDI policies and for those that seek to use tape. In some countries incentives may be needed to FDI policies strategically. In both cases, addressing attract FDI, and these may be beneficial because of investment issues in a broader context can help FDI-induced positive externalities. In others the key mobilize interest groups that have an incentive to issue may be to eliminate inefficient tax incentives. engage groups which benefit from the status quo. Determining the impact of prevailing policies on 445 " B E H I N D - T H E - B O R D E R " A N D R E G U L AT O R Y I S S U E S FDI is therefore critical in identifying reform prior- industry. In chemicals and petrochemicals, local content ities and determining how multilateral rules and requirements and export requirements are employed exten- sively. agreements may assist in meeting the objectives that are defined. Undertaking such analysis will require 3 The use of the word "spillovers" is somewhat unfortunate, information on existing policies--including poli- since productivity improvements are unlikely to be costless cies in partner countries--and assessments of their and automatic. effects. 4 Government officials are often not convinced of the inefficacy of incentives, as illustrated by the use of such instruments by many countries. Notes 5 Fiscal incentives are found to be unimportant for FDI geared 1 In the case of other domestic policy distortions, the optimal toward the domestic market. This type of FDI is more sensitive policy is well known: remove the distortions at the source, if to the extent to which it will benefit from import protection. necessary through appropriately designed regulatory interven- tion that is applied on a nondiscriminatory basis (i.e., that 6 See Markusen (2001) for a discussion of the credibility case for applies equally to foreign and domestic firms). Thus, the adop- an investment agreement and Fernandez and Portes (1998) for tion of low and uniform tariffs is preferable to the use of invest- an analysis of how international agreements may support cred- ment policies to offset the effects of high protection. This ibility. point of view is implicit in the WTO, which not only aims at The International Centre for the Settlement of Investment progressive liberalization of trade but also prohibits the use of Disputes (ICSID), which operates under the aegis of the World most TRIMs. Bank, is responsible for applying the convention. The ICC has a Court of Arbitration. UNCITRAL has adopted a set of arbitra- 2 Investment measures have tended to be concentrated in spe- tion and conciliation rules that can be used in the settlement cific industries, with automotive, chemical and petrochemical, of commercial disputes. and computer industries leading the list (Moran 1998). Local content requirements are most important in the auto industry; 7 See Henderson (1999) for a comprehensive analysis of the export requirements are more important in the computer OECD-based MAI negotiations. 446 43 P E T E R H O L M E S Trade, power of the East India Compa- ny, which he argued hurt both India and the United Kingdom. Competition, Smith presciently drew attention to the significance of interna- and the WTO tional anticompetitive behavior in the services sector and to the symbiotic role of private and public actors.2 Before World War II, international cartels had a high profile, and after the war the 1947 Havana Charter of the abortive International Trade T Organization (ITO) included, in his chapter looks at the trade and Article 46.1, a requirement that members police competition debate, which has been international restrictive behavior: a preoccupation for more than 50 years. Studies and policy experience show that although the case for Each Member shall take appropriate measures some form of multilateral rules is in principle clear, and shall co-operate with the Organization to there is extensive debate about how their introduc- prevent, on the part of private or public com- tion might be effectively brought about, what the mercial enterprises, business practices affecting institutional locus of any rule-making effort should international trade which restrain competition, be, and how international norms might be limit access to markets, or foster monopolistic enforced. control, whenever such practices have harmful effects on the expansion of production or trade and interfere with the achievement of any of the Background other objectives set forth in Article 1.3 The topic "trade and competition policy" was put on the WTO agenda at the Singapore ministerial The Havana Charter contained no general obliga- meeting in 1996, and a decision was made to set up tion to adopt a competition law. The ITO as an a working group on this interface. The working organization would have been called on to investi- group is not a negotiating forum, but its work has gate any complaints not resolved by consultation influenced the attitudes of WTO members toward and to report its recommendations for action. The the possibility of negotiations, and its reports give a Havana Charter allowed for intergovernmental valuable account of the evolution of the debate.1 cooperation and also provided a means for dealing Concerns about competition and trade have a with disputes in the services sector. Although the long pedigree. Adam Smith, in the Wealth of debate has moved a long way from the Havana Nations, had already denounced the monopolistic Charter in many respects, it is striking that the aims 447 " B E H I N D - T H E - B O R D E R " A N D R E G U L AT O R Y I S S U E S of 1947 seem quite sensible today (Mathis 2000): European Common Market, only Germany had a that is, preventing private business practices from vigorous domestic competition policy. The Treaty nullifying the benefits of the removal of govern- of Rome that established the European Community mental barriers to trade. put in place a form of supranational competition Before discussing why anyone might object to an policy, giving the European Commission powers to agreement along the lines of the Havana Charter, it directly intervene and control anticompetitive is worth briefly reviewing some of the reasons why behavior by firms that could distort trade between anticompetitive practices in one jurisdiction, and member states. It thus addressed the problem iden- policy toward these practices, may have spillover tified in the Havana Charter but went further than effects in the global economy. These reasons are the charter in giving the executive the authority to admirably reviewed in WTO (1997c). The heart of act, not merely to report and recommend. The the problem is that anticompetitive behavior can jurisdiction of the European Commission, however, have cross-border effects but that national competi- was strictly limited to practices affecting cross-bor- tion authorities have as their objective only the der trade. The European Community did not create interests of their own jurisdictions. The European an internal competition regime for its members. Union (EU) and the United States take into account Market access was the driving force, and the avoid- in their decisions the impact of actions taken else- ance of distortions to trade was the prime goal. where that affect their economies but not the Market access considerations led to a focus on the actions that affect only foreign markets. A dominant interests of exporters rather than consumers firm or cartel may use market power in one market (although within the EU this is no longer so). to leverage its position in another--for example, by International antitrust concerns are somewhat dif- predatory behavior. It is not unusual for competi- ferent. Here, the main preoccupation is the effect on tion laws to specifically exclude sectors of activity in consumer welfare of agreements or acts by dominant which foreigners may be the main victims of anti- firms that may, for example, charge excessive prices competitive practices. Shipping conferences have for exports. The remedy is likely to be different, since traditionally been exempted, and many jurisdic- the cross-border effect is now that of failure of the tions exempt export cartels either explicitly or home jurisdiction to act against anticompetitive implicitly.4 International mergers pose problems behavior, including mergers and export cartels. not foreseen in 1947. Two firms may merge in a The third reason for advocating some form of home market where there is plenty of competition agreement is that with globalization, the transac- but may also have subsidiaries in markets where the tions costs of allowing each national jurisdiction to new combined entity may have a dominant position deal separately with international mergers, for (see Chapter 44, by Evenett, in this volume). example, is becoming intolerable, as in the Boeing- In all these cases it is unlikely that governments in McDonnell Douglas and General Electric­Honey- the affected country will be able to get the evidence well cases. In the first case the EU and the United needed to pursue infringers even if they have effec- States managed to avoid a conflict; in the second tive laws. The capacity of competition authorities to case they did not. And, as noted above, in many adequately assess all the factors in transnational cases national or regional authorities simply lack mergers may well be limited even if there is no access to vital information. problem of willingness. The increasing number of We thus have a powerful case for some form of international mergers and strategic alliances where international agreement on the regulation of com- the authorities are forced to rely on whatever infor- petition in the world economy, and the WTO seems mation the parties supply clearly raises the question a natural place for this. In contrast to the Havana of the need for enhanced cooperation. Charter language, there is, at the time of writing, no Overall, the reasons for wanting some form of suggestion of an active role for the WTO as an international competition agreement can be divid- organization. Rather, the debate is about whether ed into three themes: international antitrust, market the obligations of member states should include access concerns, and transactions costs. The distinc- having a competition policy that incorporates inter- tion between international antitrust and market national dimensions. access concerns is not always clear, but it is impor- Given the arguments in favor, it is sometimes sur- tant. In the early years of the establishment of the prising how much opposition there is to starting 448 Trade, Competition, and the WTO negotiations on trade and competition. On close request is made can only act on matters that would examination, the skeptics have a case that must be be illegal under its own laws. The United States has answered before any agreement is possible. sought to replace the cooperation agreements with To establish the case for international competi- multilateral legal assistance treaties that will contain tion rules under the WTO, it is helpful to pose a much stronger obligations than those in the EU- "subsidiarity test" (Rollo and Winters 2000): U.S. agreement. Whish and Wood (1994) argue that in the case of · Are there significant cross-border competition mergers, despite the existence of overlapping juris- issues, in practice? dictions, the costs are manageable, and, moreover, · Do current national or regional (e.g., EU) regimes the different objectives of different regions cannot deal adequately with these issues? easily be reconciled with compulsory common pro- · If not, could the existing or currently envisaged cedures. Janow (2000) reports that EU and U.S. non-WTO cooperation arrangements provide a authorities routinely manage to agree on common solution? definitions of the relevant market when both are · If they cannot, is the reason for this a revealed looking at the same case. But conflicts such as political preference not to address competition occurred in the GE-Honeywell case in 2001, where issues? the EU blocked a merger approved in the United · Or is there an assurance problem that can be States, cannot be ruled out. solved by a binding agreement? If existing regimes are inadequate, could the cur- · If an agreement is needed to deal with competi- rently envisaged non-WTO cooperation arrangements tion policy, would some form of WTO regime provide a solution? The United States has long have benefits greater than the costs? favored bilateral agreements as a solution. U.S. offi- · If so, what form? cials assert that the lack of common rules across jurisdictions would make any WTO rules unsustain- We address each question in turn. able. They acknowledge the need for more coopera- What is the evidence for the existence of significant tion but stress that it must be voluntary. But despite cross-border competition issues? In the 1990s the U.S. the existence of a multilateral legal assistance treaty Department of Justice became enormously active in between the United States and Canada, Canada has investigating and prosecuting cartels in industries yet to change its law to allow exchange of confiden- such as vitamins, steel, and animal feeds. The tial information. The U.S. International Antitrust department uncovered massive evidence of global Enforcement Act does provide a framework for violations. The citric acid and lysine cartels involved bilateral cooperation, but the only agreement so far world markets of about US$2 billion in the late is one with Australia. There are almost no bilateral 1990s. These developments prompted Justice to set agreements with developing countries, and setting up an International Competition Policy Advisory up a new web of bilateral arrangements could be Committee (ICPAC), which reported that cartels unmanageable. The evidence that existing coopera- existed on a large scale (ICPAC 2000; see also Chap- tive agreements are adequate is less than compelling. ter 44, by Evenett, in this volume). Thus, there is If cooperative agreements are not the answer, is the powerful evidence that cross-border cartels are not reason a revealed political preference not to address being adequately policed. competition issues? Perhaps the existing situation Do current regimes deal with the problem? Numer- reflects actual preferences. EU industry is strongly ous bilateral agreements between the major indus- opposed to any agreements that would oblige the trial countries, such as the EU-U.S. agreement EU to give confidential information to U.S. authori- providing for "positive comity," allow countries to ties. Certain business spokespersons regularly request other jurisdictions to take action under express fears that information given to the Euro- their own laws when the requesting party's interests pean Commission, if shared, might lead to criminal are affected. The ICPAC report and Janow (2000) imprisonment penalties in the United States. The show the weakness of this approach. There is no United States recently proposed an OECD agree- obligation to exchange confidential information-- ment that would enjoin members to act against car- in fact, there is no obligation to do anything--but tels which are already illegal under their existing only a right of request. The jurisdiction to which the laws, but this did not include a commitment to 449 " B E H I N D - T H E - B O R D E R " A N D R E G U L AT O R Y I S S U E S police all export cartels. Even the EU, desirous as it want their turf taken over by trade officials. The is to see enhanced cooperation, has some problems debate often focuses on vertical restraints; trade with a commitment to police the anticompetitive officials sometimes demand that exclusive distribu- behavior of its firms in export markets. There are tion systems be allowed and that parallel imports be also, of course, governments that in fact want their banned, and this is opposed by competition offi- firms to be able to make monopoly profits. cials, who do not have a market access perspective The competition policy community itself has (see Box 43.1). As discussed below, however, the shown strong political reticence. Discussions at the emphasis of the discussions and recent proposals OECD revealed that competition specialists do not has moved away from market access. B O X 4 3 . 1 W H Y C O M P E T I T I O N P O L I C Y S H O U L D N O T B E S E E N T H R O U G H A M A R K E T A C C E S S L E N S Bargaining over market access concessions has age one supplier to undertake relationship-specific been a time-honored and highly effective means of investments (such as the purchase of special lowering border barriers in multilateral trade nego- machines or the cost of tailoring parts to the man- tiations. Unsurprisingly, when the view emerged ufacturer's particular needs). Without a long-term that inadequate competition policy enforcement contract, the supplier would be unwilling to could undermine the expected market access gains undertake such commitments. The benefits of from trade reforms, some began to argue that such relationship-specific investments would be international disciplines on competition policies lost if long-term contracting with suppliers were were necessary to ensure market access. banned on the grounds of preserving foreign mar- This viewpoint--which for some time had con- ket access. siderable currency among leading U.S. policymak- The potentially divergent effects on market ers--arose out of frustration on the part of U.S. access and economic efficiency of certain competi- exporters over their attempts to gain access to the tion policy decisions suggests that the former may Japanese market. From the mid-1980s on, many not be a good proxy for the latter--unlike the case U.S. trade policy analysts argued that Japanese of most negotiations over border barriers. This authorities had tolerated anticompetitive business explains the view of most competition authorities practices by Japanese firms, in particular through that any multilateral initiatives on competition pol- vertical integration between producers and dis- icy ought to be evaluated solely in terms of their tributors. A vast legal and economic literature effects on economic efficiency and not on market grew up on whether Japan's markets were rela- access (Hoekman 1997). There may be other tively more closed than other nations' markets, explicit goals for competition policy (for example, with little agreement emerging. (Evenett and Sus- black empowerment in South African law), but low 2000 survey much of this debate.) many consumer groups argue that broad "public Unlike trade reform, where reduction in border interest" provisions in competition laws are easily barriers typically improves both access to domestic captured and distorted by producer interests. markets and the domestic allocation of resources, Summing up, national competition enforcement there are numerous instances in which prohibiting should not focus on market access. (This is not horizontal and vertical agreements between firms because other aims such as redistribution are might enhance market access at the expense of unimportant but because competition policy is economic efficiency. For example, a manufacturer rarely the best instrument for achieving them.) may sign a long-term exclusive contract with a During 1997­2000, a clear consensus emerged in supplier that effectively shuts out foreign suppliers the WTO Working Group that competition policies from competing for the manufacturer's business. and international competition policy initiatives The manufacturer may forgo the benefits from should not be evaluated in terms of their effects on competition among suppliers in order to encour- trade flows. 450 Trade, Competition, and the WTO Is there an assurance paradox that can be solved the EU could be said to have seriously sought to through a binding multilateral agreement? Even if achieve convergence on substantive as well as proce- there were maximum goodwill, international dural issues at the WTO, its aims are now much antitrust enforcement is in large part a public good. more modest: it is unimaginable that the EU would Any country that offers to take other countries' propose that everyone adopt a WTO competition interests into account without an assurance that code so specific as to rule out any of the existing others will do the same incurs a cost. But is it the provisions of U.S. law. case that everyone is likely to be willing to adopt effective antitrust enforcement if others agree to do Developing Countries the same? As noted above, certain interests oppose more effective international enforcement. Never- A WTO agreement that obliged members to have a theless almost all governments see the desirability of competition law of some kind would clearly prima- competitive markets. The fact that there are real dif- rily affect developing economies that do not have a ferences of opinion does not mean that agreement competition law, as well as those such as Hong Kong is impossible, but it helps in understanding why it (China), which argue that their own open trade has not yet been reached. regimes are sufficient to guarantee fully competitive Can a WTO regime be devised that offers greater markets.5 As discussed in Box 43.2, the latter argu- benefits than costs for all members? If so, what form ment is not compelling. would it take? What is clear from the discussions is Why would developing countries need a WTO that the idea of an international agency reporting competition code when they could adopt domestic on and eventually policing international competi- laws unilaterally? The tally of countries with com- tion issues is not even being discussed at the WTO. petition laws shows that the number has been The EU would like all (or nearly all) WTO members increasing every year. Competition policy is spread- to have their own competition law and calls for the ing in Sub-Saharan Africa, and there is a growing creation of stronger modalities for cooperation, not recognition of the need for vigorous competition a global equivalent of its Competition Directorate. policy to assist the transition from economies dom- Unlike the abortive ITO, the WTO has a judicial inated by parastatals. South Africa has been in the arm but no executive. Bringing in the WTO essen- vanguard of promoting the use of competition pol- tially means bringing in the Dispute Settlement icy as a way of achieving both economic and social Body (see Chapter 9, by Delich, in this volume). ends, to make industry more competitive in export Fox (1999) argues that the international competi- markets and to deconcentrate power. Developing tion issue should be split, with a general WTO obli- countries are also increasingly aware of internation- gation that members have a competition law al cartels and anticompetitive practices regarding ensuring that market access is not unreasonably distribution. The Consumer Unity & Trust Society impaired, and with other matters left to other fora. (CUTS) recently examined an interesting case in Many analysts have argued that there are such sub- which the Indian subsidiary of Unilever was found stantial differences in philosophy between jurisdic- to be abusing its control of the distribution system tions (and between trade and competition officials) in neighboring Bhutan.6 In this case the Bhutan that any substantive rules on competition law authorities were able to take effective action by would risk forcing a single option in a domain insisting that Unilever allow a second distributor to where opinions are divided and would even risk the carry its products. subordination of efficiency considerations in verti- Although there is little empirical evidence on the cal restraint to market access considerations (Hoek- link between competition and growth (Tybout man and Holmes 1999; Marsden 2000). But as 2000), technical assistance and voluntary adoption discussions have proceeded at the WTO and the of laws should not be controversial. There is an OECD, there has been a clarification of what is interesting precedent in the treatment of basic commonly agreed and what is not. There is a con- telecommunications in the General Agreement on sensus that nondiscrimination and transparency are Trade in Services (GATS), where negotiators pro- key disciplines that everyone should respect. What duced a reference paper laying down regulatory and remains to be agreed is not whether these principles competition principles that countries could sign on should be made to apply, but how. Whereas in 1995 to à la carte but had to stick to once agreed (see Box 451 " B E H I N D - T H E - B O R D E R " A N D R E G U L AT O R Y I S S U E S B O X 4 3 . 2 F R E E T R A D E A N D C O M P E T I T I O N P O L I C Y It is sometimes asserted that, in the absence of to do so would be frustrated by overseas rivals government-sponsored barriers to trade and willing to sell at the world price of the commodi- investment, national competition policy cannot ty in question. improve the allocation of resources in a small Empirical studies have certainly borne out the economy. Over 30 years ago, Bhagwati (1968) claim that trade reform helps reduce the exercise showed that if a monopoly producer in a small, of market power by domestic firms (Feenstra open economy was exposed to unimpeded com- 1996; Tybout 2000). It has not been shown, petition from abroad, the monopolist would be however, that eliminating tariffs eliminates unable to charge a price above the world price. domestic market power. Even if such findings did More important, the monopolist would expand emerge, they would only apply to markets in output until the marginal cost of production which the goods are tradable. Although it could equaled the world price--the outcome that be argued that open foreign direct investment would prevail in a competitive market. Thus, regimes could inject competitive pressures into competition from abroad could not only tame nontradable sectors, EU and U.S. antitrust experi- but actually eliminate domestic market power. If ence demonstrates that even in competitive this theoretical proposition were borne out in the economies, some markets remain vulnerable to empirical literature, it would call into question anticompetitive practices. Thus, the available the rationale for certain national competition empirical evidence suggests that trade liberaliza- policies. For example, there would be no need to tion is not a perfect substitute for an active com- worry that a merger of two domestic firms would petition policy, although it is an important and result in excessive market power, as any attempt powerful procompetitive policy. 29.2, by Carsten Fink, in this volume). A similar trade and competition should be on the post-Singa- quasi-voluntary approach could be pursued in the pore WTO agenda came essentially from the EU. competition area. Better international cooperation The EU has made a point of incorporating into in antitrust could also be a positive attraction."Pos- nearly all its recent preferential trade agreements itive comity" would not go very far, but a real com- with Central and Eastern European countries mitment to sharing information and expertise (CEECs) competition provisions that ban "(i) all where cartels are discovered in industrial countries agreements between undertakings, decisions by could be a major benefit, which could only be taken associations of undertakings and concerted practices up by a country that has a competition law. between undertakings which have as their object or One of the most sensitive issues in this area is the effect the prevention, restriction or distortion of potential abuse of patents. The WTP Agreement on competition; and (ii) abuse by one or more under- Trade-Related Aspects of Intellectual Property Rights takings of a dominant position in the territories of provides a right for host and importing countries to the Community or of [the partner country] as a control patent abuses through competition law--a whole or in a substantial part thereof, if they may little used provision that should be reaffirmed and affect trade between the two partners."7 The agree- strengthened in any competition agreement. ments declare that the meaning of these terms shall be that used in the EU under what are now Articles 81 and 82 (formerly Articles 85 and 86 of the Euro- The Positions of the Major Players pean Community treaty). The full wording of this In the 1990s the United States claimed that Japan's and equivalent agreements has been widely seen as competition laws were operating in a manner that was dictation by the EU of the form of domestic compe- biased against U.S. imports. By 1996, however, the tition policy its partners should have. In fact, the United States had lost interest, and the demand that wording of the agreement technically leaves the 452 Trade, Competition, and the WTO associates free to achieve this end by their own core principles of domestic competition law means, but the CEECs' relationship with the EU and policy; 2) co-operation between competi- means that the latter can make such demands, large- tion authorities, including both specific cases ly due to the carrot-and-stick of EU membership. and more general co-operation and exchange of The 1999 EU­South Africa Free Trade Agreement information; and 3) technical assistance and contains similar provisions but does not spell out capacity-building for the reinforcement of detailed competition policy criteria.8 The EU-Mexi- competition institutions in developing coun- co Free Trade Agreement (2000) contains equivalent tries. (Mogens 2001) provisions.9 The EU is thus creating a web of bilateral agree- However, in a communication of June 29, 2001, ments, including agreements with the United the EU suggested, "Individual developing countries States--which is doing the same. The EU, however, may . . . wish to reserve their judgement on the costs is arguing that having so many overlapping and and benefits of such an agreement. In this connec- possibly inconsistent agreements is inadequate and tion, a possible option to explore could be to provide inefficient. The United States maintains that given for the possibility for developing countries to decide, the different interests and capacities of different at the conclusion of negotiations, whether they wish parties, this approach is inevitable. to subscribe to the competition agreement."10 The European Commission would like to see a The European Commission argues strongly that generalization of the provisions it has been putting the adoption of common core principles, including into its trade agreements introduced into the WTO. nondiscrimination and transparency, should not The bilateral agreements of the form referred to conflict with other development objectives. But the above are feasible only if a state has a competition commission also suggests that if countries do want law. The EU would like a basic agreement on the to exclude sectors from competition rules or from need for domestic competition law in WTO mem- national treatment provisions, they should be free bers, although its proposals provide for opt-outs to do so, subject only to the proviso that exclusions (see also Garcia Bercero and Amarasinha 2001). must be transparent and hence predictable for eco- nomic actors--the model perhaps being similar to What we are proposing is to introduce, into the some of the elements in the GATS, referred to WTO, provisions that require its members to above. The EU places great stress on the benefits of a adopt certain minimum standards and core common framework for cooperation, which it principles as it relates to their domestic compe- argues would be helped by some agreement on tition laws and regulations and to respect cer- common principles. It stresses, however, that the tain requirements of international co-operation obligation to undertake cooperation will be "volun- between competition authorities. The establish- tary" in that the strict obligation will be to consider ment of a domestic competition policy and a requests, not to act on them, although there could competition authority with sufficient enforce- be an obligation to explain any refusal. This gets ment powers remain the basis, in the absence of around the problem of onerous procedural obliga- which a country would not be able to address tions, but it leaves open the complaint that the EU anti-competitive practices of a domestic or can also refuse assistance to developing countries, international character. What we are aiming at even though the EU's aims are said to include is therefore, somewhat similar to what we did in "responding to the longstanding developing coun- the TRIPs Agreement, namely the establish- try concerns about the importance of assistance by ment of a certain number of basic principles for `home' competition authorities in those cases in inclusion in domestic laws. (Mogens 2001) which foreign firms may be engaged in anticompet- itive practices with an impact on developing coun- The EU is thus asking in some ways for more in a try markets."11 WTO agreement than was provided in the Havana One solution is that the obligation to offer coop- Charter, but it insists that its aims are modest: eration could be asymmetrical. Here we can also see a possible parallel with GATS, where countries may More specifically, we suggest that WTO negoti- decline to open up sectors but are required to apply ations should focus on three key issues: policies on a most-favored-nation (MFN) basis. 453 " B E H I N D - T H E - B O R D E R " A N D R E G U L AT O R Y I S S U E S The Clinton administration in the United States, information with the United States, although however, argued strongly for more bilateral and vol- whether this is a priority for the latter will have to be untary cooperation outside the WTO. The ICPAC determined. report of April 2000 highlighted the global dimen- Developing countries are divided. South Africa and sion of antitrust and the weaknesses of existing some Latin Americans are favorably inclined, but arrangements but firmly argued against WTO Asian countries are more skeptical. Officially, India involvement: has argued that while it recognizes the multilateral dimension of competition policy and supports the At this juncture, the majority of the Advisory work of UNCTAD, it wants to see some of the basic Committee believes that the WTO as a forum systemic issues of the WTO sorted out before new for review of private restraints is not appropri- areas are included in negotiations. There are, howev- ate. Given the possible risks, and the lack of er, influential counterviews within India.15 international consensus on the content or Meanwhile, in October 2001, the OECD held its appropriateness of rules or dispute settlement first Global Competition Forum, along the lines in this area, this Advisory Committee believes suggested by the ICPAC. The forum was attended by that the WTO should not develop new compe- 30 OECD members and 26 nonmember govern- tition rules under its umbrella.12 ments (including India), as well as selected business and consumer groups. The mandate of this group is Most economists with a disinclination toward dialogue and exchange of experience regarding best harmonization will have an instinctive sympathy for practice and modes of cooperation; it is not expect- this position. There is, nevertheless, an inconsisten- ed to address trade issues per se.16 cy in it: the ICPAC report shows that the voluntary approach is not delivering. Conclusion The U.S. view has evolved under President Bush. In mid-2001 U.S. Trade Representative Robert Zoel- The Doha Ministerial Declaration calls for negotia- lick stated: tions on competition to be launched at the Fifth WTO Ministerial Conference, in 2003, if consensus In competition policy, U.S. trade and anti-trust can be reached on the modalities of negotiations. authorities recognize the significance of the The EU has an increasing number of countries on issue. Therefore, we are working to understand its side now that it has become more flexible and has more clearly what the EU seeks, and are dis- moved well away from its initial focus on market cussing with the EU how it can accommodate access. It is ready to discuss opt-outs or phase-ins the concerns of the United States and other for developing countries that do not have the incli- countries.13 nation or the capability to introduce a competition law. It will, however, have to seriously consider But doubts remain: going beyond its present offer on export cartels. And it will probably have to make concessions on What is not clear to us, however, is how compe- other topics--for example, on antidumping, which tition obligations based on the core principles has to be seen as an issue distinct from competition should be assessed; for example, the important in this context.17 question of how dispute settlement might oper- There is perhaps a deal that would interest devel- ate or whether other forms of oversight such as oping countries, under which those countries that peer review might be more satisfactory.14 agree to have a competition law would obtain signif- icant additional rights, above all in terms of cooper- The EU responds that only the consistency of laws ation. This is more likely to involve discussion and with agreed core principles, not cases and enforce- exchange of information than an attempt to address ment, should be subject to the Dispute Settlement really sensitive matters such as multijurisdictional Body, and it favors peer review. The EU does have review of mergers. No country is likely to give up the something to offer the United States: enhanced right to review cases, but there are things that can cooperation on antitrust enforcement. The EU usefully be done even here--for example, agreeing might have to be ready to exchange confidential on ways to define relevant markets. 454 Trade, Competition, and the WTO The key challenge for developing countries is to 6 See . The CUTS ensure that negotiations strengthen the competitive Website, , contains comprehensive material on competition policy in developing countries. disciplines in global markets and that progress is made toward addressing their needs and recogniz- 7 EU-Polish Association Agreement, . 8 Available at . I thank numerous friends and colleagues at the World Bank, the 9 Available at ; see Article 39 and Annex XV. and assistance for this chapter. The chapter draws heavily on an 10 "Draft Communication from the EC and its Member States: A earlier collaboration with Bernard Hoekman and benefited greatly WTO Competition Agreement and Development," June 29, from his editorial assistance. 2001. 1 On the pros and cons, see the 2001 report to the WTO Gener- 11 Ibid. al Council of the WTO Working Group on the Interaction between Trade and Competition Policy (WT/WGTCP/5, Octo- 12 Available at . 2 See Smith (1976 edition), vol. 2, ch. 7, pt. 2: 87­103. 13 In a statement of July 17, 2001, after meeting Pascal Lamy; see 3 Available at . CHAPTERV>. 14 Ibid. 4 Some U.S. experts believe that repealing the U.S. Webb- 15 See "Let's Be Proactive on Multilateral Competition Policy," Pomerene Act, which allows export cartels, would have no available at , "Viewpoint." impact, as any adverse effects affecting only foreigners are legal in any case in the United States--and, indeed, elsewhere. 16 See . 5 This is the strongly held position of the Hong Kong govern- 17 For a full discussion of why antidumping and competition poli- ment, but it is contested by the Hong Kong Consumer Council. cy are different, see Lawrence (1998). 455 44 S I M O N J . E V E N E T T Merger and an integrating world. The approach adopted is essentially microeconomic, with a focus on Anticartel national policies toward merger review and anticartel enforce- Policies in an Era ment. Commercial developments in the 1990s suggest that these of Integrating two policies are where most of the action is, in both developing Markets and industrial economies. The Changing International Commercial Landscape T he last decades of the 20th century Among the changes in corporate strategy during the saw extensive liberalization of trade 1990s, one of the most important was the surge of and investment by developing and industrial coun- mergers and acquisitions, which accelerated after tries. Most of that reform was unilateral, but some 1995. In contrast to the mergers of the 1980s, firms took place through regional integration initiatives in developing economies--notably in Latin Ameri- and multilateral trade rounds. These reforms ca and East Asia--participated in this new wave of occurred in the context of a changing international corporate consolidation. Even before the trend commercial landscape. Among the changes were the accelerated in 1999­2000, mergers and acquisitions first truly global wave of privatization, mergers, and of Latin American and Asian firms by companies acquisitions, affecting firms from developing as well outside the regions totaled just under US$400 bil- as industrial economies--in particular, in Latin lion during 1994­98. Over the same period, firms in America and East Asia--and a rise in international these two regions also acquired US$150 billion of anticartel enforcement. assets from outside their respective regions. In The connection between these changes and trade effect, the two regions experienced two-way merg- reform has not attracted much attention. There may ers and acquisitions. Much of the inward foreign be strong incentives for firms to circumvent the addi- direct investment (FDI) in Latin America came in tional competitive pressures created by trade and the form of mergers and acquisitions, not greenfield investment reform through mergers or acquisitions. investments, as is often supposed (JETRO 2001; Recognition of the incentives to engage in anticom- UNCTAD 2001b). petitive behavior has heightened interest in the role Many factors account for this upsurge of global of and effectiveness of national competitive policies mergers and acquisitions. They include the avail- in developing and industrial economies alike.1 ability of cheap financing from banks and capital This chapter describes the changing commercial markets in the late 1990s; the deregulation and pri- landscape and the role of national merger policy in vatization of many services industries; liberalization 456 Merger and Anticartel Policies in an Era of Integrating Markets of rules against foreign takeovers of and acquisition cies--and such policies are almost certainly weaker of stakes in domestic firms; and, in certain indus- in developing economies than in the United States tries--notably in the financial sector and telecom- and the EU. One can therefore be confident that munications--a belief that only large, consolidated developing economies have been detrimentally firms could develop the capital base required to affected by these cartels. In sum, the evidence paints compete against leading firms in the world market. a potentially disturbing picture of the effects of car- Not all mergers or acquisitions result in higher tels in the developing world. prices or distort resource allocation. Such transac- Another pertinent development is the widespread tions may result in efficiencies that enable firms to adoption of competition laws by developing lower production costs, and these cost reductions economies. Although estimates vary, as of 2000 at may then be passed on to consumers in the form of least 80 countries had antitrust laws or planned to lower prices. The efficiency motive for mergers-- have them (see Palim 1998; ICPAC 2000). What fol- the significance of which has been vigorously debat- lows discusses the efficacy of such national antitrust ed--need not call for public policy intervention. enforcement in a world of integrating markets and The same cannot be said for another motive for whether international initiatives in the competition consolidation, acquisition of market power. This policy area might help reinforce the independence, results in prices being raised above marginal costs integrity, and coherence of antitrust enforcement in and sales falling short of the competitive bench- developing economies.3 The emphasis is primarily on mark. During an era of trade reform--which puts the control of mergers and on anticartel enforcement. downward pressure on the prices and profit mar- gins of incumbent firms--mergers can attenuate National Merger Policies in an Integrating World and in principle offset the benefits of liberalization at the border. The recent explosion of merger and acquisition The second feature of the changing commercial activity is documented in Table 44.2. The effects of a landscape is the rise of cartel enforcement in indus- proposed merger need not be confined to the mar- trial economies (Table 44.1).2 Cartels are increasing- kets of those nations where the firms involved in the ly seen to have an international dimension, affecting merger are located. National antitrust authorities more than one national market. A recent study have therefore asserted a right to review proposed showed that in a sample of 40 international cartels mergers that may have potentially adverse effects in prosecuted during the 1990s by the European Com- their jurisdictions. This, in turn, has led to a prolif- mission and the U.S Department of Justice, 24 (40 eration of "merger notification" regimes and merger percent) lasted at least four years. It would seem that review procedures and introduces the possibility-- these cartels do not collapse rapidly under their own recently demonstrated in the General Electric­Hon- weight (Evenett, Levenstein, and Suslow 2001). eywell merger case--of disagreement between Levenstein and Suslow (2001) documented pur- reviewing bodies. In principle, each reviewing body chases by developing countries of 16 goods the sup- can veto a proposed merger. In cases where this is ply of which was internationally cartelized (by EU likely to happen, the firms that wish to merge typi- or U.S. firms) at some point during the 1990s. The cally offer to amend their proposed transaction so authors found that in 1997 developing countries as to meet the objections of a reviewing body. Often, imported US$81.1 billion in goods from industries this involves divesting certain subsidiaries or that had been found to engage in a price-fixing con- accepting restrictions on the operations of the spiracy during the 1990s. This represented 6.7 per- merged firm. cent of developing country imports and 1.2 percent Merger review is no longer the preserve of indus- of their GDP. The products concerned represented trial countries. Many developing countries--by an even larger share of the trade of the poorest some counts, about 40--have merger notification developing countries, 8.8 percent. Note that the requirements and review procedures. Cicerone study looked at international cartels that were (2001) provides a summary of these requirements found to operate within the EU or the United States. for 46 jurisdictions. Argentina, for example, It cannot be assumed that these cartels operated in demands that mergers involving worldwide sales of every developing country, but cartels are more like- 2.5 billion pesos be notified to its antitrust authori- ly to operate in nations with lax anticartel poli- ties, even if not a single peso is earned within its 457 " B E H I N D - T H E - B O R D E R " A N D R E G U L AT O R Y I S S U E S Table 44.1 Economies with Firms Convicted of Price Fixing by the United States or the European Commission during the 1990s Economy Cartel Angola Shipping Austria Carton board, citric acid, newsprint, steel heating pipes Belgium Ship construction, stainless steel, steel beams Brazil Aluminum phosphide Canada Carton board, pigments, plastic dinnerware, vitamins Denmark Shipping, steel heating pipes, sugar Finland Carton board, newsprint, steel heating pipes France Aircraft, cable-stayed bridges, carton board, citric acid, ferry operators, methionine, newsprint, plasterboard, seamless steel tubes, shipping, sodium gluconate, stain- less steel, steel beams Germany Aircraft, graphite electrodes onboard, citric acid, aluminum phosphide, lysine, methionine, newsprint, pigments, plasterboard, seamless steel tubes, steel heating pipes, vitamins Greece Ferry operators India Aluminum phosphide Ireland Shipping, sugar Israel Bromine Italy Carton board, ferry operators, newsprint, seamless steel tubes, stainless steel, steel heating pipes Japan Graphite electrodes, lysine, methionine, seamless steel tubes, shipping, sodium gluconate, sorbates, thermal fax paper, vitamins Korea, Rep. of Lysine, methionine, shipping Luxembourg Steel beams Malaysia Shipping Mexico Tampico fiber Netherlands Carton board, citric acid, ferry operators, ship construction, sodium gluconate, tampico fiber Norway Carton board, explosives, ferrosilicon Singapore Shipping South Africa Diamonds, newsprint Spain Aircraft, carton board, stainless steel, steel beams Sweden Carton board, ferry operators, newsprint, stainless steel Switzerland Citric acid, laminated plastic tubes, steel heating pipes, vitamins Taiwan (China) Shipping United Kingdom Aircraft, carton board, explosives, ferry operators, newsprint, pigments, plaster- board, seamless steel tubes, shipping, stainless steel, steel beams, sugar United States Aircraft, aluminum phosphide, bromine, cable-stayed bridges, carton board, citric acid, diamonds, ferrosilicon, graphite electrodes, isostatic graphite, laminated plastic tubes, lysine, maltol, methionine, pigments, plastic dinnerware, ship con- struction, shipping, sorbates, tampico fiber, thermal fax paper, vitamins Zaire Shipping Note: Italics denote that the product is currently under investigation. Source: Levenstein and Suslow (2001): table 1. 458 1994-8 46.1 30.7 251.8 161.5 432.4 213.6 234.1 115.4 659.3 413.2 231.4 153.3 1262.1 1193.8 1922.4 1453.8 Sum 9.4 7.0 9.6 1998 73.8 41.3 75.9 56.0 71.4 74.5 96.5 27.1 51.0 28.1 22.8 34.5 42.1 79.3 66.7 44.2 33.0 18.0 49.2 54.5 77.3 99.6 76.6 73.8 179.5 480.6 478.6 731.7 540.0 7.2 1997 69.2 43.8 58.4 63.3 15.1 59.5 47.7 48.4 27.6 47.6 47.4 33.5 42.1 70.7 41.8 72.6 61.1 41.7 42.0 68.3 82.0 99.9 72.6 83.3 101.6 148.9 108.1 233.1 232.9 359.2 299.3 5.8 5.2 7.1 45.9 22.3 50.2 48.6 80.7 89.7 92.4 55.5 60.1 51.9 26.9 19.9 51.8 88.3 23.1 68.0 49.2 35.1 29.4 71.5 76.4 95.3 63.6 84.4 1996 129.8 195.4 186.2 283.5 239.3 7.3 2.8 7.7 1995 32.8 11.4 53.1 34.8 75.6 38.4 73.3 38.6 52.7 42.7 21.6 29.4 50.6 70.3 18.4 66.6 45.6 25.1 34.1 55.0 76.6 79.0 62.4 69.2 105.5 211.5 167.0 306.5 212.2 1991­98 8.5 8.5 1994 30.1 14.8 21.1 49.2 29.2 68.6 44.0 12.9 64.0 69.3 23.8 27.8 67.2 95.6 67.2 18.9 70.4 42.5 33.4 27.4 78.5 74.9 91.3 61.5 67.5 100.0 141.5 129.1 241.5 163.0 Pacific, Asia 3.4 1993 13.7 27.9 65.7 33.5 15.3 23.4 21.8 72.5 64.5 18.4 88.9 34.9 27.4 27.4 78.6 97.8 56.2 70.4 44.0 65.5 138.9 205.8 134.9 and 5.1 8.8 America 1992 10.4 59.3 89.2 21.2 16.9 59.1 51.1 38.2 31.8 74.7 22.7 22.7 65.8 83.7 73.6 69.6 99.2 59.3 55.1 100.0 120.3 180.0 Latin in 3.9 0.7 6.4 5.0 4.3 8.3 5.4 1991 24.4 17.7 31.8 41.7 13.8 18.2 33.2 31.8 64.6 71.4 65.2 62.2 79.9 59.3 42.1 114.8 189.8 Acquisitions and FDI FDI FDI FDI FDI FDI FDI FDI FDI FDI FDI FDI FDI FDI FDI FDI M&A stakes M&A stakes Pacific M&A stakes M&A stakes M&A stakes M&A stakes M&A stakes M&A stakes (2000c). America Inward Inward and Inward Inward Economies Inward Inward Economies AD Inward Inward Latin Mergers of Outward Outward of Outward Outward of Outward Outward of Outward Outward majority majority UNCT % of Asia majority majority % of majority majority % of majority majority % of cross-border cross-border % cross-border cross-border % cross-border cross-border % cross-border cross-border % with with with with with with Industrial with with Developing (2000); % % % % % % % % acquisitions. Inward Inward Inward Inward and Cross-Border Outward Outward Outward Outward Johansson mergers and 44.2 Kang M&A, ce: ableT Note: Sour 459 " B E H I N D - T H E - B O R D E R " A N D R E G U L AT O R Y I S S U E S borders. Brazil, by contrast, exempts from merger sumer welfare. It would not be surprising if such a review any proposed commercial transaction merger were permitted by both nations--especially between firms with no corporate presence within its if the merged entity's additional market power borders. increased its profits by more than enough to com- The diversity of notification requirements, differ- pensate (in the eyes of the reviewing authorities) for ences in the deadlines for making merger review any consumer welfare losses. In this case a proposed decisions, and the presence or absence of judicial merger that--as measured by the traditional micro- review add considerably to the cost of undertaking economic standard--distorts the global allocation international mergers and also increase the proba- of resources is permitted. bility of different and inconsistent decisions being The absence of any compensation mechanism is made by national antitrust authorities. In the 1990s what differentiates multijurisdictional merger the magnitude of these transactions costs became a review from many other areas of international source of concern for the International Bar Associa- cooperation and negotiation, including trade tion, as well as for policymakers. Several sensible reform. Antitrust officials and practitioners are proposals for reducing these costs were made (see reluctant to make tradeoffs across cases in the same ICPAC 2000: ch. 4). In the remainder of this section way that tradeoffs across sectors occur during trade I focus on the potential for resource misallocation negotiations. Consequently, any merger that has sig- due to multijurisdictional merger reviews. (I nificant adverse effects in one jurisdiction is likely to abstract from the associated transactions costs.) be rejected, even though the benefits created by the There are two potential types of resource misallo- merger in other jurisdictions could more than offset cation that merger reviews can generate. First, a the harm done. Thus, a merger that improves the proposed merger that improves the global alloca- global allocation of resources might never come tion of resources may be vetoed by at least one about, having been vetoed by at least one jurisdic- antitrust authority. Second, a proposed merger that tion. Essentially, the effects of international mergers distorts the global allocation of resources may be spill over national borders, and, in the absence of allowed to proceed. In the first case the potential for any compensation scheme, mergers that could in improving resource allocation is forgone; in the sec- principle improve the global allocation of resources ond case actual resource allocation is adversely will not take place. affected. The following three features of national It is useful to keep this trinity in mind when enforcement--call them the "international antitrust assessing international initiatives on competition trinity"--can generate these inefficient outcomes. policy. One class of proposals advocates procedural cooperation between national antitrust authorities, · Multiple veto. Each nation has the right to veto a in part to encourage the convergence of substantive proposed transaction. standards for evaluating mergers. Another set of · National standards. Each nation evaluates a pro- proposals calls for internationally agreed minimum posed merger in terms of its effects on firms and standards. Yet another argues for explicit harmo- consumers within its borders only. nization of standards, which could be accomplished · No compensation scheme. Each nation makes a through the adoption of common regional, plurilat- decision on whether to allow a given merger with- eral, or multilateral norms. out reference to any other merger or policy matter Even if a common standard were adopted, it need that might be of importance to the nations not be based on, or be consistent with, the efficiency reviewing the merger. standard advocated by microeconomists. Further- more, even if efficiency were adopted as the com- The multiple veto feature of national merger mon standard, the resource misallocation created review has already been discussed. The standards by the two other elements of the trinity--multiple adopted by each nation can differ and need not veto and lack of a compensation mechanism-- focus on the effect of the merger on the sum of con- would remain. The sources of resource misalloca- sumer and producer surplus within a nation. Sup- tion go far deeper than the adoption of conflicting pose that a proposed merger affecting only two merger review standards by nations. nations is reviewed using standards that place far However politically infeasible it may seem at pres- more weight on producer interests than on con- ent, the logic of these arguments suggests that only a 460 Merger and Anticartel Policies in an Era of Integrating Markets supranational decisionmaker able to consider the · Difficulties in collecting evidence abroad, inter- aggregate efficiency effects of a merger in all affect- viewing witnesses overseas, and extraditing per- ed jurisdictions can avoid the resource misalloca- sons from other jurisdictions. tion created by national merger review. (A national · Lessened attractiveness of national corporate decisionmaker with a cosmopolitan viewpoint, to leniency or amnesty programs to firms that fear which other nations defer, could also fulfill this applications for leniency in one jurisdiction will role.) Such a decisionmaker would consider only leave them exposed to investigations and poten- the aggregate impact of a merger on resource alloca- tial punishment in other jurisdictions. This is par- tion, irrespective of the distribution of welfare gains ticularly worrisome because such programs have and losses in different national markets. Since, of been instrumental in encouraging firms to course, the distribution of gains and losses is a pri- "defect" from a cartel agreement and to supply mary reason why national competition authorities evidence of wrongdoing by other firms to state differ in their assessments of mergers, such a supra- authorities--and rarely are cartel members suc- national arrangement may well have to involve the cessfully prosecuted without a former member creation of structural adjustment funds as well. turning state's evidence. These could mitigate the effects of decisions under · "Multimarket effects" that increase the incentive which certain states lose while the aggregate welfare to collude. Fines for cartelization are a function of of the group of nations improves. the cartel's effects within a single jurisdiction, Although supranational decisionmaking is rather without taking into account the enhanced profits unlikely in the near or medium term, the point is derived from cartelizing other nations' markets, that a supranational solution might overcome the as well. inefficiencies created by the international antitrust trinity and do so far more effectively than the other Practitioners have recognized the difficulties cre- proposals outlined above. ated by the first two factors, but the third factor has received little attention in policy circles, although it has been extensively analyzed in the economics lit- National Anticartel Policies in an Integrating World erature. Only a few jurisdictions have taken steps to Private cartels attempt to raise profits through overcome the first set of difficulties, principally agreements to restrict the quantity supplied to a through signing bilateral cooperation agreements market, to fix prices, or to rig bids. Such agreements on antitrust matters (for example, an agreement in distort the allocation of resources away from the 2000 between Australia and the United States) or by competitive norm, and for this reason many nations invoking mutual legal assistance treaties (as in the have passed laws that restrict or ban agreements case of the United States and Canada).4 What more likely to cartelize markets within their jurisdictions. could be done to enhance the deterrent posed by By contrast, many nations often exempt firms from national cartel laws in a multijurisdictional world? anticartel laws with respect to attempts to cartelize A first option would be to extend the set of exist- overseas markets, presumably on the grounds that ing cooperation agreements on antitrust matters to no harm is done to domestic consumers. An inter- include more nations. A slightly different variant esting question is whether the effectiveness of might be to negotiate regional, plurilateral, or even national anticartel laws is compromised in a multi- multilateral agreements that permit signatories to country world. request other signatories to collect evidence, inter- National anticartel laws have a dual purpose: to view persons, and consider extraditing those punish cartels that are found to exist, and to deter charged with cartelization. Yet another possibility cartels from forming in the first place. The effective- would be to agree on minimum standards for coop- ness of the deterrent depends critically on the sever- eration between jurisdictions and then allow ity of the punishment and on the probability of nations to tailor cooperative agreements to their assembling sufficient credible evidence of carteliza- specific needs. tion. The following three factors undermine the A second option, which could easily build on the deterrent provided by national anticartel law in a first, would be to allow for the simultaneous appli- world of many jurisdictions (Evenett, Levenstein, cation of corporate leniency in multiple jurisdic- and Suslow 2001): tions. In addition, nations could agree on a 461 " B E H I N D - T H E - B O R D E R " A N D R E G U L AT O R Y I S S U E S minimum degree of leniency for successful peti- different reform options for the two types of tioners and minimal conditions for such leniency. antitrust policies considered here. While there is in Each nation would retain its own leniency program, principle a case for supranational approaches to and there would be no pooling of sovereignty. A merger review, nations need not pool sovereignty more ambitious variant would be for a group of for the sake of fighting international cartels. nations to agree to establish a common leniency Although the analysis in this chapter is economic program, whereby firms would apply to a single in nature, it must be borne in mind that the princi- body for leniency in all of the jurisdictions that are pal practitioners of antitrust law are lawyers, who parties to the agreement. This would avoid the diffi- are typically attuned to operating in a single juris- culties created by inconsistent leniency decisions diction. Given differences across nations in legal being made by multiple antitrust authorities. traditions and procedures, legal practitioners, when A third option would be to establish a panel that dealing with other jurisdictions, place much would estimate the pecuniary gains a cartel has emphasis on procedural cooperation. Such cooper- made from all the markets in which it has been ation may over time enhance trust in foreign juris- found to operate. Such a panel could be entirely dictional procedures and practitioners and is advisory, suggesting fines that each jurisdiction arguably the first step in forging new approaches to could impose and doing so in such a way that total international competition policy. This form of pro- fines equal or exceed the total gains from cartelizing cedural cooperation is to be encouraged and should many markets. Panels could be formed on a case- not be confused with the skepticism expressed by-case basis, and their membership would draw above about the efficiency-enhancing effects of pro- from the professional economic and antitrust com- cedural cooperation aimed at convergence of munities. national standards for merger review. In sum, much can be done to enhance the effec- tiveness of national anticartel laws against interna- Notes tional cartels without creating a new supranational agency or pooling national sovereignty. Much could 1 For the purpose of this chapter, I use the terms "competition also be accomplished regionally, plurilaterally, or policy" and "antitrust policy" interchangeably. In other words, the discussion of competition policies is confined to the ele- through organizations such as the OECD. Nations ments of antitrust policy--policies toward mergers, acquisi- with more aggressive anticartel regimes could move tions, takeovers, cartels, vertical restraints, monopolization, ahead more quickly, taking the steps outlined above and anticompetitive practices. in cooperation with like-minded nations. 2 This pickup in anticartel enforcement started in the United States with the reform of its corporate leniency program, Conclusion which grants amnesties to qualifying firms that come forward with evidence of cartelization. The sizable fines imposed by In determining a policy stance with respect to inter- U.S. authorities have not gone unnoticed by other nations and national competition policy matters, decisionmak- have provided a pecuniary incentive for enhanced anticartel enforcement elsewhere. For a discussion of the enforcement ers are well advised to give primary consideration to record against international cartels in the 1990s, see Evenett, efficiency, not market access. This chapter has dis- Levenstein, and Suslow (2001). cussed why national antitrust laws may perform 3 See Tavares (2001) for an account of how various Latin Ameri- suboptimally in an integrating world economy from can initiatives on competition policies are affecting the imple- an efficiency perspective. The reasons differ for car- mentation of antitrust enforcement in that region. tel and merger policies, illustrating that the general 4 Waller (2000) describes the extensive and fruitful cooperation debate over trade and competition policy needs to between the United States and Canada in these matters and be replaced by specific analyses of the problems contrasts it with the absence of any sustained cooperation on faced by national enforcement of different aspects cartel enforcement between the European Commission and of antitrust laws. These analyses point to markedly the U.S. Department of Justice. 462 45 S I M O N TA Y Trade and where sanctions have not tradi- tionally been acceptable, efforts have been made to curb child Labor labor and other abuses through trade measures. Unilateral efforts by cities and states to connect trade and economic Text, Institutions, and Context measures to labor and human rights have also been increasing.1 This has been most noticeable in the case of Myan- mar, where a number of human rights abuses, including forced I labor, have been reported.2 n the run-up to the 1996 WTO minis- In 1999 the ILO developed a convention (No. terial meeting in Singapore, the 182) against the worst forms of child labor that United States and the European Union (EU) sup- potentially has implications for trade. This high- ported a "social clause" that would amend the WTO lights the real possibility that treaties in this area to permit countries to impose trade measures to may come into conflict with the WTO, with no clear ensure that minimum labor standards were met by priority between conflicting international rules and their trading partners (Leary 1996). Much heated institutions. The Vienna Convention on the Inter- debate resulted, largely along North-South lines, pretation of Treaties provides that the most recent with many nongovernmental organizations and most specific treaty should prevail. In the area (NGOs) and federations of unions supporting the of trade and labor this could lead to rapid changes introduction of such a social clause. The final state- in international law and policy and to inconsistency. ment from the Singapore meeting recognized the In the run-up to the Seattle ministerial meeting in reality of lower wage costs in developing countries 1999, there were signs that it would no longer be as a legitimate advantage and designated the Inter- possible for the WTO to continue to exclude labor national Labour Organization (ILO) "the compe- standards. Labor issues have featured prominently tent body to set and deal with [labor] standards" on the U.S. agenda, as a result of lobbying by unions (WTO 1996: 4). This was reaffirmed at the Doha and other interest groups.3 In Europe labor protec- ministerial meeting in 2001. Box 45.1 summarizes tion has been a concern for many countries, espe- some of the key labor standards at issue. cially regarding child labor. This chapter considers To date, efforts to link labor rights to trade within the connections between trade and labor. The cur- the WTO have been stymied. Nevertheless, the idea rent debate in the WTO is a beggared choice of using trade measures and sanctions to promote between, on the one hand, ignoring the interde- and protect labor rights resurfaces regularly, both in pendence between trade and economic issues and the WTO and outside it (Rodrik 1997). In the ILO, the protection and promotion of human dignity 463 " B E H I N D - T H E - B O R D E R " A N D R E G U L AT O R Y I S S U E S B O X 4 5 . 1 W H AT A R E C O R E L A B O R S TA N D A R D S ? The term "labor standards" covers a wide range set too high, this will tend to hurt the poorest, of rights and requirements. Maskus (1997) pro- who generally fall outside the ambit of the stan- vides a useful fourfold classification: dards. Moreover, to be legitimate, and hence effective, the standards need to command wide · Basic rights, including prohibition of slavery, support within the society. This suggests that the physical coercion, discrimination, and exploita- debate is not about labor standards as such but tive child labor about whether internationally determined and · Civic rights, including freedom of association, enforced standards are appropriate for develop- collective bargaining, and expression of griev- ing countries. ances Labor standards have been the responsibility of · Survival rights, including the rights to a living the ILO since 1919. The organization has a tripar- wage, limited hours of work, information about tite structure, with representation from labor, hazards of the job, and compensation for acci- business, and government in each member coun- dents try. The ILO's first major function is to promote · Security rights, including protection against higher international labor standards through the arbitrary dismissal and rights to retirement and preparation of a series of conventions, of which survivors' compensation. there are more than 180. Ratification of these conventions is sporadic, partly because some Some of these rights may clearly be termed countries simply disagree with certain conven- basic or core labor standards in the sense that tions but also because some conventions are the principles they embody command universal regarded as too inflexible and as not allowing for respect as a matter of humane treatment of even minor national variations in labor practices. workers. Adherence to others, by contrast, The table lists the conventions that cover core depends on the country's level of development labor standards, along with the number of coun- and national preferences. There is not much dis- tries that had ratified them by 2001. In many agreement that labor standards are a legitimate instances the failure to ratify a convention does and useful tool for governments. Societies have not necessarily imply that equivalent national leg- the right to regulate themselves in such areas, islation is lacking. Conversely, several countries but the standards need to be appropriate to a that have ratified conventions have failed to country's level of development. If standards are implement their provisions in practice. Core ILO Labor Standards Conventions Convention Date Number of number Title adopted adherents 29 Forced Labour Convention 1930 158 87 Freedom of Association and Protection of the Right to Organise Convention 1948 137 98 Right to Organise and Collective Bargaining Convention 1949 149 100 Equal Remuneration Convention 1951 153 105 Abolition of Forced Labour Convention 1957 156 111 Discrimination (Employment and Occupation) Convention 1958 151 138 Minimum Age Convention 1973 111 182 Worst Forms of Child Labour Convention 1999 89 The second major function of the ILO is to serve complaints depends on persuasion and peer pres- as a clearinghouse and publicity mechanism for sure. A Committee of Experts issues interpreta- complaints about government and private actions tions on the operation of various conventions. that contravene national obligations regarding Subject to these findings, the ILO compiles docu- labor standards. The process for dealing with ments on each country's compliance with conven- 464 Trade and Labor: Text, Institutions, and Context B O X 4 5 . 1 ( C O N T I N U E D ) tions it has ratified (see Maskus 1997). Complaints forced labor. This increased activism, coupled with about country practices are considered, and the the fact that recent improvements in information ILO's findings are publicized. Until very recently, and communications technologies and the bur- no sanctions beyond public opinion existed, but geoning activities of NGOs have made "naming in 1999 the ILO excluded Myanmar from its pro- and shaming" more effective than ever, means ceedings for its persistent use of forced labor. In that the ILO is becoming a more potent force in December 2000 the ILO approved the imposition improving labor standards. of sanctions for the same reason, requiring other members to review their dealings with Myanmar Source: Prepared by the volume editors, based on to ensure that they were not abetting the use of McCulloch, Winters, and Cirera (2001): ch. 15. and, on the other hand, seeking to join these issues conditions. This has led to a number of voluntary in a negative way, centered on sanctions. In order to labeling schemes, such as the "Rugmark" warranting move beyond the impasse, it is important to find the that a carpet or rug was made without child labor. happy middle ground. Similar efforts have led to labels on soccer balls. Com- panies such as Nike, Levi Strauss, and Reebok have declared their own codes of conduct, which include Controversies and Underlying Debates labor issues, with varying success (Cassel 1996; Gib- To those who favor links between labor and trade, ney and Emerick 1996; Steiner and Alston 1996). basic labor rights are indisputable public goods that In general, there are three intertwining but dis- the international community should promote and tinct themes in the arguments for and against links protect. These proponents argue that international between trade and social issues (Ward 1996): eco- principles have evolved in this area, with dynamic nomic spillovers and competitiveness; "psychic" or developments in labor and the allied human rights, "moral" spillovers; and unilateralism, compulsion, and that the WTO should promote human dignity, and protectionism. not simply freer trade and economic growth. The fear is that freer trade will spark more human mis- Economic Spillovers and Competitiveness ery, in a "race to the bottom" in standards of worker protection (Rodrik 1997). Those concerned about competitiveness see labor Those who oppose such links emphasize that rights as a cost factor that affects the overall ability freer trade and economic growth are complex and of a country to produce goods at attractive prices, difficult issues in themselves. They view resolution compared with other countries. Companies operat- of trade issues in such areas as agriculture as more ing in industrial and other countries that seek to pressing than taking on new issues such as labor enjoy higher labor standards are put at a cost disad- rights. Moreover, those who focus on trade see in vantage compared with countries with lower stan- the proposed labor rights provisions a potential dards (see Box 45.2). Industrial and developing pretext for protectionism. They emphasize that the countries differ on these issues, partly because of success of the WTO, and its predecessor, the GATT, lack of adequate legislation or weak implementa- is bound up with these organizations' focus on tion capacity in the latter and partly because of dif- trade and trade-linked economic issues. fering social priorities. Although it is recognized The proposals for a social clause and for sanctions that one country's standards may be inappropriate against violation of labor rights have attracted consid- for another, some still see lower labor standards as erable media and NGO attention. There is special an "unfair" advantage and worry that countries will concern about products that are alleged to have been lower their standards of labor protection to attract produced by child labor, by forced labor, and in sweat- investors. Although these fears have not been borne shops that fail to meet basic standards for working out in empirical studies on industrial migration 465 " B E H I N D - T H E - B O R D E R " A N D R E G U L AT O R Y I S S U E S B O X 4 5 . 2 T R A D E A N D L A B O R S TA N D A R D S : T H R E E D E B AT E S Commonly expressed complaints about limited EU and the United States, but although this adherence to core labor standards (CLSs) and the increase has coincided with the opening of mar- alleged effects on foreign economies focus on the kets to international trade, the correlation does operations of export-processing zones (EPZs), not necessarily imply causation. Changes in tech- international wage spillovers, and the putative nology, demographics, regulation, and unioniza- "race to the bottom" in labor standards. tion rates could also be responsible for these labor market trends (Richardson 1995). It is mis- Export-Processing Zones leading to ascribe differences in wages across Core labor standards may be introduced across countries (or changes in these differences) to dif- most of the economy but waived or weakened in ferential labor standards. Policy changes that the export sector. Questions then arise as to the weaken labor standards are not necessarily corre- effects of EPZs on wages and whether export- lated with lower wages in poor countries, larger specific exemptions from CLSs constitute an volumes of low-priced exports from those coun- export subsidy. The available evidence (Maskus tries, and, ultimately, downward wage pressure 1997) indicates that firms in EPZs tend to pay in the importing countries. higher wages than firms outside the zones. There are a number of reasons for this. Take the case of Competitive Impacts on Standards an EPZ in an economy that has a large informal Another common complaint is that competition sector with low wages, a small formal sector with will reduce standards in the higher-standard higher wages, and substantial unemployment. (OECD) countries--that there will be a "race to Exempting from tariffs imports that are interme- the bottom" in labor standards. This claim has diate goods in EPZ production in order to attract two variants: (a) there is a danger that standards labor-intensive assembly operations will raise the will decline to low levels worldwide, or (b) the demand for low-skilled labor. Favorable tax treat- competition in standards may be concentrated ment and subsidies to fixed costs within EPZs are among low-wage nations, preventing countries likely to attract foreign capital, further expanding that would otherwise move toward higher stan- labor demand and raising wages in and outside dards from doing so. The simple argument that the area. To the extent that workers are then competition can push standards toward their low- trained and acquire skills, employers are likely to est levels is wrong on its face; it presumes that the seek to retain them, putting further upward pres- lowest standards will prevail as market outcomes. sure on wages in the EPZ. It is unlikely that, accounting for global income levels and technologies, African or South Asian Comparative Advantage and International Wage labor standards would prevail in international Spillovers competition. Moreover, it is questionable that Organized labor interests in high-wage countries integrated markets must see convergent labor are often concerned about the effects on their standards. Open economies can sustain high stan- own labor markets of limited CLSs in low-wage dards through some combination of higher taxes, countries. Differences in labor standards, howev- lower wages, and exchange rate devaluation. er, will have negligible implications for employ- ment, wages, and wage inequality in high-wage The Empirical Evidence on Labor Standards and countries. For example, less than 5 percent of Trade children working in developing countries are The empirical evidence suggests that differential engaged in export sectors, and their contribution CLS levels have little effect on exports and export to output is small (although in particular sectors, prices. OECD (1996), a study relating measures such as carpets, footwear, and apparel, child of export performance to indicators of labor stan- labor is more prevalent). Wage and income dards, found no relationship between the two. inequality have increased since the 1980s in the The study detected no effects of differences in 466 Trade and Labor: Text, Institutions, and Context B O X 4 5 . 2 ( C O N T I N U E D ) core labor standards on U.S. import prices in sec- firms. Aggarwal found no association between tors such as textiles and apparel, nor was there U.S. FDI and poor labor standards in developing any indication that export prices for handmade countries. In fact, she concludes that U.S. FDI is carpets are lower in countries in which use is not concentrated in countries or sectors with low made of child labor. Rodrik (1996) econometri- labor standards and that countries with weaker cally related basic measures of labor standards labor standards do not have higher import pene- across countries to international trade flows, tration rates in the United States than countries using data such as country ratification of ILO con- with stronger labor standards. ventions covering core labor standards and an Such evidence will not satisfy those who are indicator of enforcement problems in child labor concerned about the impacts of differences in standards. He, too, was unable to determine any labor standards (or of EPZs) on competitiveness. relationship in the data or to find any suggestion The studies can be criticized for their inability to of a positive statistical relationship between low measure CLS effectively, given the inherent diffi- labor standards and flows of FDI from the United culties with data in this area. Further criticisms are States to particular countries. Indeed, there was that the studies did not adequately control for some evidence that FDI is lower than expected in other significant impacts on trade and FDI and countries with low labor standards. that they are static (cross-sectional). Many Aggarwal (1995) notes that it is common in observers, for example, point to the rapid developing countries for labor standards to be increases in manufactured exports from EPZs in lower in less export-oriented sectors and in non- China, where CLSs are not fully respected traded goods than in export-oriented industries, (although it is not known whether this fact including even textiles and carpeting. Within all explains export growth). manufacturing, workers in firms with high export-output ratios tend to receive higher wages Source: Prepared by the volume editors, based on and benefits than those in less export oriented Maskus (1997). (see Revesz 1992; Stewart 1992; Esty 1996), percep- and other forms of forced labor, but they question tions of a "race to the bottom" persist. the effectiveness of trade sanctions in resolving such evils. As an alternative, they propose assistance and the transfer of resources to promote and enable "Psychic" or "Moral" Spillovers solutions. For example, a common rebuttal regard- Some argue that as economic interdependence ing child labor is that banning the goods produced, increases, so does concern about moral issues such as without further action, may lead to factory closures the treatment of our fellow human beings. They see and to the child workers being abandoned to street the growth of the world economy and world trade as life as prostitutes or to criminal behavior. The com- being in tandem with the promotion of human mon defense against the "moral spillover" argument rights, both being post­World War II phenomena. In is thus most often a debate not about morality itself human and labor rights the international communi- but about the effectiveness of using trade measures ty has been increasingly willing to take action against to address moral issues. violators. This is especially true where the abuses are gross and large scale (Steiner and Alston 1996). Unilateralism, Compulsion, and Protectionism By and large, developing countries that have been accused of violations do not seek to justify them in Although unilateral actions per se are not illegal the name of sovereignty. They agree that there is a under international law, such measures can be a core of basic labor and human rights that have to be form of compulsion, practiced by larger and richer observed, such as the prohibition against slavery states against less powerful countries. Moreover, 467 " B E H I N D - T H E - B O R D E R " A N D R E G U L AT O R Y I S S U E S measures ostensibly taken for labor protection may The objection to this approach is that the implied be forms of disguised protectionism. The interna- wide scope for unilaterally decided standards would tional labor regime cannot itself enforce labor stan- seriously undermine the coherence, stability, and dards. Rather, the ILO seeks to promote compliance strength of the international trade regime. through monitoring and reporting systems and by Two other general points should be noted about providing assistance to help states meet their obliga- attempts to find exceptions within the existing tions. Such efforts and assistance are justified by the GATT language. First, the chapeau of Article XX is principle of common but differentiated responsibil- open to different interpretations. It states that meas- ity: countries differ in their abilities to discharge ures should not be applied arbitrarily or in an their responsibilities. unjustifiably discriminatory manner, leading some to suggest that the WTO must accordingly take pains to try to discern genuine labor concerns from Environment and Labor in the GATT/WTO disguised protectionism. Others argue that the and Other International Institutions WTO can and should give more leeway to sincere Many suggestions have been made for amending the attempts by nations to legislate their exceptions in WTO to respond to the controversies over linking the areas of concern identified by Article XX, there- trade and labor. Before considering amendments, it by providing greater scope for restricting trade in an is important to first study existing disciplines and effort to attain the objectives listed. the extent to which these already allow for an inte- Second, concerns about labor protection most gration of social concerns. often deal with production and process methods The core principles of the GATT/WTO are most- (PPMs), rather than with products. What is at issue favored-nation (MFN) and national treatment. is how a product has been produced. Those seeking General exceptions to these obligations are found in to use trade sanctions to combat perceived viola- GATT Article XX, which contains no specific men- tions of labor standards question the traditional tion of "labor rights."At present, those who propose trade view of treating "like" goods alike. For exam- that labor rights be accepted as exceptions to trade ple, is a football made by child labor under poor rules have to seek grounds in more general or more workplace conditions different from one made by limited provisions. Article XX(e) specifically recog- reasonably paid adults in a proper factory? Those nizes a very narrow exception for products made by who care about labor rights would say that it defi- prison labor. (The ILO recently invoked this provi- nitely is. From the perspective of trade rules, howev- sion against Myanmar.) This clause is too narrow to er, the two items are not different. encompass "basic" labor rights (Leary 1996) Those arguing for a less narrow interpretation of Beyond this exception for forced labor, the argu- Article XX have not met with success to date in dis- ment that labor rights should be recognized as an putes brought before the WTO regarding environ- exception to GATT/WTO trade rules is highly spec- mental measures. This may reassure those who ulative. Some suggest that these rights should be an oppose using trade as an instrument for attaining allowed exception under Article XX(a), which per- noneconomic objectives, but it does not please mits an exception to trade rules where "necessary to those who marched in the streets of Seattle. Both protect public morals." This clause is presently con- groups can share three concerns over the present strued narrowly as permitting countries to ban situation. The first is that several WTO dispute set- imports of immoral or pornographic materials. tlement decisions seem to suggest that even multi- Some suggest, however, that it can be read more lateral treaties on labor must give way to trade rules. widely to include moral objections when goods are The second is that the interpretations given by dif- made under conditions that defy labor rights and ferent Dispute Settlement Understanding (DSU) fail to observe human dignity (Charnovitz 1998). panels might displace the need and ability of WTO These arguments imply that under Article XX(a) members to negotiate and reach agreement. The each country would be allowed to restrict market third, and related, concern is that a number of the access on the basis of "moral" objections to goods cases on trade and the environment have demon- made in violation of "basic" labor rights. It would strated shortcomings in the DSU process. be up to each country's moral judgment to decide Three basic deficiencies have shown up in the unilaterally which labor rights are or are not basic. DSU process in trade and environment cases. First, 468 Trade and Labor: Text, Institutions, and Context the panels often take a narrow and pro-trade inter- These concepts have developed relatively recent- pretation of the GATT, especially of Article XX. Sec- ly, and they have developed outside the WTO, ond, the premises for making the decisions largely in the United Nations and in UN-related sometimes appear weak, some decisions reveal a organs and conferences. Although the WTO charter lack of expertise and interest in issues outside of makes concessions to these concerns, the issues trade, and others have shown an ambition to reach have not been embraced by the WTO, which beyond trade law into the realm of more general remains focused on trade, with few or no connec- public international law, which would include envi- tions between the promotion of freer trade and the ronmental and labor concerns. Third, the process larger concepts. There is therefore a perceptible gap and the access of NGOs and other interest groups to between the breadth of concepts of development the DSU proceedings have come into question. The and the narrowness of institutions such as the WTO charter allows labor or environmental experts WTO that promote economic policy. In this regard, to be appointed to panels and provides for presenta- the debate over the proposed links between trade tions to the panel by interested third parties, but few and labor rights is not only about issues but also of these possibilities have been explored and prac- about institutions. ticed (Lang 1996). The narrow trade view of the WTO's mandate has Institutions been challenged on several grounds. One is that, historically, the original and aborted attempt to The international community is anarchic, without launch an International Trade Organization (ITO) democracy, equity, or world government. Effective envisaged that the organization should take labor cooperation between sovereign nations is impossi- issues into account (Botsch 1995; Charnovitz 1996). ble without their agreement--and difficult even Another, perhaps more important, argument seeks when such agreements are reached. The interna- to situate the GATT/WTO in the context of new and tional cooperation and institutions that do exist are dynamically changed concepts of development that fragmented among many subjects and responsibili- include human development and security. ties, functioning with little coordination, even under the umbrella of the United Nations. These characteristics of the international system are Sustainable Development and Human among the fundamental reasons for the controversy Development in International Institutions about the links between trade and the social factors The UN Development Programme (UNDP) has of labor and environmental protection. pioneered the concept and measurement of human Since its early days, the UN has set up specialized development, which embraces civil and political agencies for different subjects. While this has led to freedoms as well as social and economic human focus and specialization, it has also meant a strong rights.4 The concept seeks to focus on human sense of separateness between international eco- beings in the development process, rather than see nomic policy--the realm of the World Bank, the development as being concerned only with macro- IMF, and the GATT/WTO--and other aspects of economic indicators and megaprojects. Other allied international policy that have developed, such as concepts that are still evolving include "human peace and security concerns, labor rights, and security," which examines the bases of development human rights. The WTO thus faces a basic dilemma for individual human beings and seeks to ensure in approaching issues beyond its original mandate that access to such bases is widened. of trade and allowing links between trade and such In these clusters of concepts, attention centers on issues as labor rights. the needs and priorities of developing countries. There is merit in distinguishing between sugges- This emphasis is based on principles such as "intra- tions for changes in the WTO that require formal generational equity," "common but differentiated amendment and those that relate only to present responsibility," "the right to development," and practice, culture, or habits (Sampson 1999). In the "comprehensive" security. Such ideas seek to recog- second group an ample framework is already at nize the interdependence of the world's people, as hand for a number of helpful steps that would assist well as the interrelationships between economic and the integration of trade with social concerns. Such other social and political spheres of life. steps include: 469 " B E H I N D - T H E - B O R D E R " A N D R E G U L AT O R Y I S S U E S · Greater transparency and timely public access to · Accept the equality of multilateral agreements on WTO documents labor rights with those on trade rules, and work · Policy discussions with civil society, especially toward their congruence. Unilateral measures NGOs and other groups working on social con- should be avoided. cerns5 · Recognize that efforts at the state level must be · Arrangements for access and participation of emphasized in order to achieve a "first-best" rec- NGOs and experts in the dispute settlement onciliation of trade and social concerns in the process production and manufacture of goods. Such a · Increasing dialogue and cooperation between the reconciliation should be sought both in policies WTO and international environmental bodies and in institutions such as trade ministries and such as the secretariats of multilateral environ- labor unions. mental agreements, the United Nations Environ- · Emphasize cooperation and assistance to facilitate ment Programme (UNEP), and the ILO compliance with labor obligations, especially in · Increasing coordination between economic and developing countries and in vulnerable sectors trade agencies and their environmental and labor such as small and medium-size enterprises. Trade counterparts at the national level. sanctions and measures to restrict or prohibit market access should be avoided. By such means, coordination between trade and · Recognize the appropriate role of specialized fora social policies could be increased.6 This would bet- on labor issues and increase their dialogue and ter achieve the first-best solution of reconciling the coordination with the WTO. Proposals for differing concerns at the level of the production of amendment of the WTO or for the creation of goods rather than in their trade. The increased new international institutions should be put cooperation between different institutions would aside, at least for the short to middle term. allow for better understanding. This path could be · Change habits, culture, and outlook within the pursued both at the international level and within WTO in areas that would benefit labor concerns national governments. as well as improve governance in general. There is also a need for support and mediation on social concerns outside the WTO, especially with WTO members have taken steps to address some regard to extending assistance and resources to help of these issues. For example, WTO documents have countries strengthen labor rights protection. The been made much more accessible to academics and WTO as an institution faces several limitations in the public in the past few years. This improvement dealing directly with these issues. It lacks expertise, has taken advantage of Internet technology, as well as will, and culture, and some developing countries the interest of NGOs that follow the WTO process, view it with skepticism as representing certain such as the International Center for Trade and Sus- Northern countries' interests in promoting freer tainable Development. More can and should be done trade. The prospect of discussions in the WTO to create a more positive context in which the WTO appears to be threatening because it could lead to and the international community can deal with the new, binding rules. Finally, and perhaps most intersections between trade and social concerns and important, the WTO has no strong mechanism or foster better coordination and cooperation. principles regarding assistance.7 Efforts to deal with contentious issues of social concern may be better Conclusion received and more effectively implemented if they are carried out through other institutions. What the Controversies among nations over labor rights pro- WTO must then do is to be open to greater tection are likely to continue. Recent incidents and exchange, improved access, and better coordination cases, such as the treatment of Myanmar by the ILO, with these other entities and processes. have further polarized discussion. In approaching The above recommendations should be seen as new issues, the initial reaction of many people and achievable and effective in the short to medium institutions is, understandably, to avoid changes in term. A more positive agenda for coordination and paradigms. ("If it ain't broke, don't fix it.") This is cooperation in the longer run may be tentatively especially so for the WTO, which many have sketched as follows (Runnals and Tay): adjudged a success in its field. But there are also those 470 Trade and Labor: Text, Institutions, and Context who believe that something is broken, who embrace paper, and Will Martin, Patrick Low, Kym Anderson, Bijit Bora, and change, and who see a need for vast and dramatic Mari Pangestu for feedback from the World Bank­Pacific Economic Cooperation Council (PECC) seminar. changes in the WTO and other international institu- tions. These groups share a common viewpoint, even 1 Some such efforts by cities in the United States have been suc- if analyses of what, more precisely, is broken and cessfully challenged and set aside by federal courts as an intru- sion on national foreign policy. what the remedy is continue to be debated. Countless agendas and counteragendas have aris- 2 The action taken against Myanmar was decided by the ILO in en, and more will come up. There are no easy solu- the Resolution on the Widespread Use of Forced Labour in tions. There may, however, be points on which Myanmar, 87th Session of the ILO International Labour Con- ference, June 1999, available at . For a report not stand in isolation; it stands within a wider inter- on the measure, see "Burma Barred from ILO Meetings and national community that has dynamic concerns Technical Aid," ILO News, ILO U.S. Branch Office, at other than trade, including labor and human rights. . The trade organization was transformed in the 3 In January 1999 and again in July 1999 the U.S. government Uruguay Round to shoulder new issues outside the proposed that the WTO establish a working group on trade traditional purview of trade in goods. In many and labor. See the U.S. proposal on labor rights, instances, it did not take in these new areas in WT/GC/W/139, 1999, International Centre for Trade and Sus- tainable Development (ICTSD). entirety but, as with investment policy and intellec- tual property rights, strove to find and deal with the 4 See the UNDP Human Development Report series. nexus of these issues with trade. 5 Pursuant to Article V.2 of the Marrakesh Agreement that estab- A similar approach may be used for labor rights. lished the WTO, the General Council in mid-1986 adopted It is wrong to call for the WTO to displace the ILO guidelines for arrangements on relations with NGOs. For a in protecting workers. But it would be equally general discussion of NGOs in the WTO, see Esty (1999). wrong to ignore the legitimate economic, social, 6 The WTO has of late come to recognize the need for coordina- and political imperatives that call for improved dia- tion and to take steps accordingly. For instance, the Commit- logue, better coordination, and mutually supportive tee on Trade and the Environment has granted observer status policies and institutions in the international com- to 20 intergovernmental organizations concerned with the environment, including the secretariat of the Convention on munity. A new context for improved coordination International Trade in Endangered Species of Wild Fauna and and exchange is needed before and for deciding on Flora (CITES). A number of symposia with NGOs have also an agenda. Such a context can and should be the been held since 1994. foundation for progress on these issues. 7 The general preference for developing countries is at best a vestigial principle in the WTO. There have been some calls for Notes assistance to least-developed countries, but they have focused on technical assistance and training for trade matters and I thank Gary Sampson, Daniel Esty, Veena Jha, Martin Khor, Bea- trade facilitation rather than broader developmental assistance trice Chaytor, Halina Ward, and James Cameron for inputs to the and resource transfers. 471 46 V E E N A J H A Environmental Trade and Environment in the WTO Regulation and The Uruguay Round introduced into the multilateral trading sys- the WTO tem some recognition of the concepts of sustainable develop- ment and environmental pro- tection. Although environment was not a separate negotiating topic, the preamble to the agree- ment establishing the WTO includes, in contrast to the E GATT, references to the objec- nhancing understanding and tive of sustainable development and the need to strengthening policy coordination protect and preserve the environment. A number of in the areas of trade, environment, and develop- Uruguay Round agreements explicitly or implicitly ment are of key interest to developing countries. take into account environmental considerations. The context for interactions between countries on Under the GATT, a Group on Environmental these subjects changed significantly in the 1990s, Measures and International Trade had already been particularly after the Seattle ministerial meeting in established in 1971, but it remained inactive until 1999. Two distinct streams of influence can be iden- 1991. Then, impelled by a dispute between Mexico tified. The first is related to discussions (or lack and the United States involving restrictions that the thereof) in the WTO Committee on Trade and latter placed on imports of tuna to protect dolphins, Environment (CTE). The second revolves around it began to examine trade-related aspects of envi- disputes related to trade and environment that have ronmental policies that might have significant trade been brought to the WTO. effects for GATT contracting parties. It considered This chapter briefly outlines the negotiating his- trade provisions contained in existing multilateral tory of trade and environment in the WTO and environmental agreements, the transparency of describes the positions of the major players as of national environmental regulations likely to have 2001. This is followed by a review of major develop- trade effects, and the trade effects of new packaging ments on the dispute settlement front and their and labeling requirements aimed at protecting the implications for developing countries. The chapter environment. With the establishment of the WTO, ends with a discussion of the options confronting the group was replaced by the CTE, which was given developing countries. the mandates of identifying the relationship 472 Environmental Regulation and the WTO between trade measures and environmental meas- by "developing the full use of the resources of the ures in order to promote sustainable development world" through reduction of trade barriers and and of making appropriate recommendations on elimination of discriminatory practices. Efficient whether modifications of WTO provisions are environmental regulation can help achieve this required. The CTE was the first, and remains the overall objective--indeed, it is a necessary condi- only, international organization charged with tion. It is not clear that trade barriers have a role to examining these issues. play in attaining environmental objectives (see Box The CTE's work program covers nearly all issues 46.1). Nonetheless, in the Tuna case the United that arise under the heading of trade and environ- States sought to use trade barriers in this way ment. Primarily, however, it looks at the effects of (although it has been argued that protection of the environmental measures on trade, rather than the domestic industry was a factor). Whatever the case reverse, as the WTO itself has no competence in may be, the dispute illustrated that the boundary environmental policy. The CTE's work program is between trade and environment policy is blurred. driven by proposals from WTO members concern- The subsequent Shrimp dispute (discussed below) is ing issues on which they hope the committee can another illustration. It is difficult to draw the line produce concrete results to submit to ministerial between free traders and environmentalists; most conferences. Environmental experts have partici- disputes and trade restrictions appear to be champi- pated as members of national delegations or have oned by the very countries that have sought and been invited by the CTE to make presentations. achieved prosperity through free trade. CTE discussions are clustered around market access An added complication in the trade and environ- and linkages between the multilateral trade and ment debate is the precautionary principle. Usually environment agendas. For a good part of the 1990s, invoked as a justification for protective measures the CTE constituted the cornerstone for most taken in the absence of full scientific certainty--for debates on trade and environment. example, about the potential health hazards of a pes- The Seattle ministerial meeting in 1999 marked a ticide--the principle has generated considerable watershed in the trade and environment debate. It controversy, both between industrial countries and became clear that environment is a factor in the between industrial and developing countries. The mandated negotiations on agriculture and services, European Union (EU) defends the principle, while as well as in the review processes for the Agreement the United States maintains that all measures should on Trade-Related Intellectual Property Rights be based on sound science.1 The issue is further (TRIPS) and the Agreement on Sanitary and Phy- complicated in cases where there is no consensus tosanitary Measures (SPS). In Doha in 2001 it was view. Thus, in the Beef Hormone case the EU Scien- agreed to launch negotiations on certain aspects of tific Committee on Veterinary Measures concluded environmental policy, in particular the relationship that there was a "substantial body of recent scientific between environmental agreements and WTO rules. evidence that hormones exerted both tumor-initiat- The environment is emerging as a useful bargaining ing and tumor-producing effects." Although the risk chip for both industrial and developing countries in was not quantifiable, the fact that it existed was defining the contours of trade agreements. For one deemed sufficient to justify imposition of trade thing, it can be used to temper otherwise radical free restrictions. The United States countered by citing a traders. Thus, for example, environmental consider- joint WHO/FAO scientific expert panel which ations in a possible multilateral agreement on invest- reconfirmed that residues of the hormones in ques- ment could be used to temper aggressive supporters tion would not have a harmful effect if administered of these agreements. Similarly hard-line postures on in accordance with good veterinary practices.2 production and process methods (PPMs) can be It is ironic that ecolabeling, initially strongly used both for protecting the environment and for championed by Canada and the United States, is protectionist purposes. now being questioned by these countries in the con- text of genetically modified organisms (GMOs). The United States opposes labeling for GMOs on The Blurring of Trade and Environment Policy the grounds that the products are safe and that test- The preamble of the GATT states that members ing for traces of GMOs would be prohibitively seek to increase the efficiency of global production expensive and complicated--it would require segre- 473 " B E H I N D - T H E - B O R D E R " A N D R E G U L AT O R Y I S S U E S B O X 4 6 . 1 B A S I C P R I N C I P L E S O F E N V I R O N M E N TA L R E G U L AT I O N Environmental problems can be usefully divided rain) or intangible (mistreatment of animals, con- into two types: domestic, in which the damage is sumption of ivory). In such cases there is a basis contained within the borders of a country, and for cooperation and negotiation. Unilateral trade transboundary, in which the damage affects policy, however, is not the appropriate instru- more than one country. Dealing with the latter ment for dealing with the externality; standard generally requires international cooperation. The economic theory requires that externalities be emergence of global environmental problems addressed at the source. This means either that such as global warming and ozone depletion has the producer or consumer must bear the real led to a growing recognition of the interdepen- costs of the activity, or that property rights must dence of the global environment and to the con- be assigned that give owners an incentive to clusion of around 200 multilateral environmental manage and price resources appropriately. Trade agreements. sanctions cannot offset an environmental exter- At both the domestic and international levels, a nality efficiently because they affect both con- widely accepted principle of environmental poli- sumers and producers of a good and usually have cy is that "the polluter pays": public policy an impact on only a part of total production or should seek to ensure that firms pay not only the consumption. direct cost of production but also the social cost Only if international trade itself were the prob- of the pollution they generate. The marginal lem would intervention in trade be the appropri- social cost associated with a unit of pollution will ate policy response, but that is a rare occurrence. vary across countries. Countries may have very For example, there might be a ban on internation- different preferences regarding environmental al trade in harmful products as a supplement to protection, reflecting differences in the absorp- domestic bans if moving the products were dan- tive capacity of their ecosystems, in income levels gerous. But if the concern is about international (wealth), and in culture. Thus, the "polluter pays" transport, it is the transport that should be taxed principle is entirely consistent with the existence (made to pay its full social cost), not the trade; of widely different environmental standards that is, the aim is to encourage cleaner forms of among countries. transport, not to prevent goods from moving. The manner in which environmental policy is Although environmental policies may reduce implemented varies greatly from country to the ability of enterprises located in countries with country. Some countries make efforts to ensure high standards to compete with those operating that polluters do pay according to an estimate of in nations with low standards, this is exactly what the costs that they impose on society, while oth- the policy aims at. If high standards are what a ers fail to enact such policies, thereby providing society wants, the result should be the contrac- their producers with an implicit subsidy. This sub- tion of the affected activities. Restricting imports sidy does not arise from countries' adoption of makes no sense, as it promotes the domestic lower standards but from their failure to ensure activities that the environmental policy is that polluters bear the cost of the chosen stan- attempting to constrain. This, of course, is one dard. Like all subsidies, the welfare cost of such a reason why domestic industries may seek to policy is borne principally by the country provid- "level the playing field" through trade policy--it ing the subsidy (it is the country's own environ- is one way of avoiding part of the impact of envi- ment that is polluted), and the welfare benefit ronmental regulation. More generally, if con- accrues primarily to consumers in other countries sumers prefer environmentally friendly goods, who receive goods or services at lower prices. they should be willing to pay for them. Production and consumption activities in one country may have detrimental impacts on other Source: Prepared by the volume editors, based on countries. Such negative spillovers or externalities McCulloch, Winters, and Cirera (2001) and Hoekman may be physical (air and water pollution, acid and Kostecki (2001). 474 Environmental Regulation and the WTO gating GMO from non-GMO crops during cultiva- tiations. This reflects the fact that any negotiating tion, transport, and any subsequent industrial pro- agenda that does not pay adequate attention to the duction, entailing unnecessary costs. The irony is environment will fail to generate political support that when developing countries raised similar ques- in industrial countries. Developing countries are tions regarding ecolabeling of textiles and footwear, less than enthusiastic, to say the least. They worry their arguments were dismissed on the grounds that about a resurgence of protectionism and the use of consumer preferences should be catered to. Interest- unilateral measures under the guise of environmen- ingly the issue of consumer preference has not been tal concerns and are fearful that accommodating the given much emphasis in discussions on GMOs.3 use of trade-restrictive measures for nontrade pur- poses may spill over into other areas such as labor standards. They also perceive the current trade and Trade Law versus Environmental Law: environment debate as lacking balance. Proposals A Crisis of Legitimacy by industrial countries focus on accommodating Trade negotiators seek to extract better terms of measures that restrict trade, as opposed to measures trade and to open other countries' markets while that promote trade (Jha and Vossenaar 2001). A keeping their own as closed as possible. The focus is recent assessment by the UN General Assembly of on national or regional, not global, welfare. In the progress in implementing Agenda 21, the environ- case of the environment, the concept of interna- mental action plan adopted at Rio de Janeiro in tional society is not metaphorical but real. Attempts 1992, showed that little headway has been made on to introduce nonconsensual politics through instru- so-called supportive measures, such as finance, ments such as trade barriers goes against the grain access to environmentally sound technologies, and of environmental politics. To accommodate differ- capacity building. This engenders skepticism con- ences in development and in domestic environmen- cerning the sincerity of industrial countries about tal regimes, the Rio Declaration, adopted at the UN advancing sustainable development objectives on Conference on Environment and Development (the the basis of the principle of common but differenti- "Earth Summit") in 1992, embraced the principle of ated responsibilities. Doubts often expressed by common but differentiated state responsibility.4 industrial countries about the extent to which such This creates an incongruity: although the focus in supportive measures can be specified in WTO the trade policy­WTO context has been on disci- agreements aggravate the problem. plining sovereign rights to impose tariff barriers, At the risk of oversimplification, the developing members now seek to resurrect sovereign rights to country position on trade and environment could erect these barriers for environmental purposes. be summarized as follows. Environment and sus- A crisis of legitimacy arises because the consensu- tainable development are important policy objec- al character of environmental law is being over- tives, but there is no convincing evidence that turned by recourse to nonconsensual trade disputes, existing trade rules stand in the way of legitimate often in an incoherent way, depending on the eco- environmental policies and the promotion of sus- nomic stakes. When substantial commercial inter- tainable development. Compatibility between trade ests are involved, as in the case of GMOs, the effects and the environment is the rule; conflict is the on the environment may be examined only in a cur- exception. Where conflicts have arisen, GATT Arti- sory manner. Elected developing country govern- cle XX has proved sufficient for dealing with them ments have voiced concern that nonelected NGOs while at the same time providing the checks and representing limited constituencies may have a balances to avoid abuse. The CTE has made an greater influence on trade disputes than govern- important contribution to enhancing understand- ments do.5 The trade and environment debate is ing of trade and environment linkages, and this increasingly straitjacketed into North-South lines. process continues. Finally, environmental and development objectives should be achieved through supportive measures and international cooperation, Developing Country Positions as well as through better coordination at the nation- In general, industrial countries place high political al and multilateral levels. All this leads to the posi- priority on the inclusion of environmental and sus- tion that it is premature to engage in negotiations tainable development considerations in trade nego- on trade and environment. 475 " B E H I N D - T H E - B O R D E R " A N D R E G U L AT O R Y I S S U E S In the course of the 1990s the major protagonists national cooperation and supportive measures changed their strategies on trade and environment. involving a number of stakeholders remain the pre- Just before the Singapore ministerial meeting in ferred way for moving toward the use of more envi- 1996, the proponents of non-product-related PPMs ronmentally friendly PPMs (Vossenaar 1999). The and the advocates of reform of GATT Article XX overwhelming emphasis of the views expressed by made several proposals for amending the rules to developing countries with respect to this issue is accommodate environmental concerns. At Singa- that current WTO rules and practices are adequate pore it became clear that developing countries could for addressing any inconsistencies that may arise not be coerced into accepting an agenda on environ- between trade measures used in MEAs and WTO ment that did not include changes in TRIPS, domes- rules. They also point out that since so far there has tically prohibited goods (DPGs), and safeguards on never been a dispute between WTO rules and the use of environmental standards for protectionist MEAs, the conflict is more hypothetical than real. purposes.6 After 1996, the United States began to Opening the door for such measures may also open rely more on the WTO dispute settlement process, the way for using trade measures for protectionist whereas the EU shifted its emphasis to civil society purposes and for pressuring countries to join MEAs and political declarations on trade and environ- that do not serve their economic interests. ment. According to some NGOs from the North, developing countries need to demonstrate that they Ecolabeling. Ecolabeling programs are a fact of the are not antienvironment but that they simply have international marketplace; the challenge is how to different priorities, focusing more on issues of envi- accommodate such programs in the Technical Barri- ronmental justice, sustainable livelihoods, and sus- ers to Trade (TBT) agreement without compromis- tainable development (see Najam and Robins 2001). ing basic WTO rules. The situation is complicated by Surprisingly, NGOs that brand developing countries the lack of an agreed interpretation on whether pri- as antienvironment are silent on the environmental vate, voluntary ecolabeling schemes are within the agenda of the developing countries regarding TRIPS scope of the TBT agreement. Discussions in the CTE and DPGs and, more important, the need to safe- have focused on multicriteria ecolabeling schemes, guard market access from unnecessary environment especially those that are based on non-product- and health standards. related PPMs. The effects of "type-1" ecolabeling on In approaching the environment agenda in the the marketplace and international trade have been WTO, developing countries have taken a two- limited. (Type 1 labels compare products with others pronged strategy, resisting changes in WTO rules in within the same category, awarding labels to those a number of areas (the defensive agenda) and seek- that are environmentally preferable through their ing changes in others (the proactive agenda). whole life cycle.) It would appear that interest in eco- labeling in the context of international trade is at least in part attributable to the fact that it involves The Defensive Agenda many complex issues, such as PPMs, the definition Among the defensive measures in developing coun- of international standards, and equivalency. So far, tries' strategy are resistance to changes in trade mea- little has been done toward dealing with the PPM sures to conform with environmental agreements, a issue in the context of ecolabeling. In particular, in skeptical attitude toward ecolabeling, and opposition the debates in the WTO and the International Orga- to environmental reviews of trade policies. nization for Standardization (ISO), not much progress has been made on developing the concept Trade Measures Pursuant to Multilateral Environ- of "equivalency."8 mental Agreements. Proposals have been made to Clarification of the status of ecolabeling with press for adaptation of GATT Article XX to accom- respect to WTO rules may result in greater WTO dis- modate the use of trade measures specifically man- cipline in certain sectors (forest products, textiles, dated by multilateral environmental agreements and cut flowers) where trade of the exporting coun- (MEAs). Many countries have indicated that allow- tries has been adversely affected by such schemes. ing unilateral trade measures based on non-prod- Clarification may provide an opportunity to force uct-related PPMs could mean imposing domestic greater WTO discipline on purely private programs environmental policies on trading partners.7 Inter- and NGO campaigns in areas where trade has been 476 Environmental Regulation and the WTO adversely affected, and it may reduce pressures for In view of these issues, consideration could be unilateral measures. (Information would be provid- given to: ed as an alternative to an outright ban: let the con- sumer decide.) But there is also a danger of · Shortening the lifetime of some patents for envi- establishing precedents, particularly if such prece- ronmentally sound technologies dents apply to labor and human rights issues. There · Making the procedures for compulsory licensing is also a risk that clarifying the status of ecolabeling of environmental technologies less cumbersome with respect to WTO rules would encourage its · Excluding life forms from patentability and wider use in international trade and might serve ensuring compatibility between the Convention protectionist intent. In addition, it may become on Biological Diversity and the TRIPS agreement more difficult to challenge an ecolabeling measure. · Granting protection for traditional knowledge Even if the disadvantages of clarification were not through systems other than patents significant, there may be little direct benefit, as · Extending protection under geographical indica- developing countries do not use ecolabeling to any tions to other products besides wines and spirits. great extent in their domestic markets and have little to gain from ecolabeling programs in OECD coun- Market Access. Safeguarding market access for tries. Few, if any, examples can be found in which products exported by developing countries has been type I ecolabels have allowed developing countries to discussed extensively at the WTO. Developing obtain price premiums, expand market share, or countries may be more vulnerable to environmental improve their environmental performance. measures because of the composition of their exports. They may also find such standards difficult Environmental reviews. Since 1997, industrial coun- to meet on account of several constraints--having tries have suggested that an environmental impact to do, for example, with the nature of the operations assessment of trade policies be included in the WTO of small and medium-size enterprises, which Trade Policy Review Mechanism. Many developing account for a large share of their exports. Much countries argue that while environmental impact emphasis has been placed in this context on identi- assessments are useful domestic policy instruments, fying"win-win"opportunities in trade and environ- there is no need to multilateralize them. ment. Win-win situations arise when the removal or reduction of trade restrictions (high tariffs, tariff escalation, and remaining nontariff obstacles to The Proactive Agenda trade) and distortions has the potential to yield The second prong of developing countries' strategy both direct economic benefits for developing coun- is to seek changes in the TRIPS agreement, market tries and positive environmental results access provisions, and controls on trade in domesti- (WTO/CTE/W/67). Much of the discussion so far cally prohibited goods. has concentrated on removing trade distortions in sectors such as fisheries, agriculture, and energy. The TRIPS Agreement. Of special concern are pro- visions in the TRIPS agreement dealing with trans- Domestically Prohibited Goods (DPGs). Many fer of technology and the protection of biodiversity. developing countries are concerned about the (See also the chapters in Part V of this volume.) The health and environmental effects of exports to their system of intellectual property protection should markets of goods that are prohibited or severely find ways of safeguarding indigenous knowledge restricted in the exporting country. Their importers and species preservation, which may be of consider- need adequate information about the risks that able value in protecting biodiversity.9 In the manu- such products could pose to public health and the facturing sector the TRIPS agreement may: environment. Apart from information problems, countries may also lack the infrastructure (includ- · Affect technology transfer by restricting the use of ing testing facilities) and other capabilities to moni- compulsory licensing mechanisms tor and control imports of DPGs. Industrial nations · Increase the price of goods and technologies argue that a number of multilateral agreements and · Adversely affect innovation in the area of envi- instruments already address this issue. Although ronmentally sound technologies. duplication must be avoided, there is a need to 477 " B E H I N D - T H E - B O R D E R " A N D R E G U L AT O R Y I S S U E S examine whether existing instruments, such as the The AB made several determinations in this case prior informed consent (PIC) procedure, are suffi- on the interpretation and application of Article cient from the perspective of developing countries XX(g). The means-to-an-end relationship between to prevent environmental damage from such the trade measure and the policy of conserving an imports, in particular with regard to product cover- exhaustible, endangered species, sea turtles, was age and procedures. observably close and real, and therefore the measure was one "relating to" the conservation of an exhaustible natural resource. The AB also deter- Dispute Settlement mined that the imposition of the embargo was an The contours of the trade and environment debate even-handed measure, in that it was implemented have been affected by panel and Appellate Body in conjunction with restrictions on domestic har- (AB) decisions, some of which were quite contro- vesting of shrimp, as required by Article XX(g). versial. (See the submission by Pakistan, In justification of the wider meaning to be given WT/GC/W/162.) Several suggestions for reform of to "natural resource" in Art XX(g), the AB cited the the dispute settlement mechanism have arisen as a UN Convention on the Law of the Sea (UNCLOS), response to the perceived political nature of dispute the Convention on International Trade in Endan- resolution on trade and environment matters.10 gered Species of Wild Fauna and Flora (CITES), and This section reviews aspects of the Shrimp case and other bilateral and multilateral actions to protect a number of proposals for rectifying some of the living natural resources. It also cited GATT 1947 imbalances created by the panel in this case.11 panel rulings that fish was an "exhaustible natural In 1997 India, Malaysia, Pakistan, and Thailand resource." In taking this view, the AB held that requested the establishment of a WTO panel to con- XX(g) covers both living and nonliving natural sider U.S. trade restrictions on shrimp imports. resources. According to the defendants in the case, Under the authority of the Endangered Species Act the AB extended further the ruling in the Gasoline (ESA), the United States had imposed embargoes case (relating to U.S. imports of gasoline from on the import of shrimp from a number of its trad- Venezuela and other countries), opening the way for ing partners for the purpose of protecting the sea a country to adopt national measures operating turtle population. The ESA makes access to U.S. beyond its legal jurisdiction to protect the environ- shrimp markets conditional on certification that a ment and conserve "natural resources" and then to country has adopted conservation policies that the enforce these measures through trade barriers, pro- United States considers comparable to its own with vided this is preceded by a process of getting target- respect to regulatory programs and incidental tak- ed countries to negotiate with it on bilateral, ing of turtles. The United States unsuccessfully regional, or multilateral environmental agreements. argued that this trade measure satisfied Article In the Shrimp case, however, the AB went on to XX(g), which allows trade restrictions if needed to rule that the U.S. embargo was applied in a manner conserve an exhaustible natural resources. The that would constitute a means of both unjustifiable panel rejected the U.S. argument on the basis of its and arbitrary discrimination between countries interpretation not of Article XX(g) but of the cha- where the same conditions prevail, contrary to the peau to Article XX. The panel found that the U.S. requirements of the chapeau of Article XX. The measure constituted unjustifiable discrimination body reasoned that unjustifiable discrimination between countries where the same conditions pre- includes the application of a trade measure, such as vail. The United States appealed the panel's reason- the U.S. embargo, that does not allow for any ing. The AB found fault with the panel's inquiry into the appropriateness of the regulatory interpretation and took Article XX(g) into consid- program for the conditions prevailing in the export- eration. It decided, however, that although the ing countries. The failure to engage in serious nego- embargo served an environmental objective that is tiations to conclude bilateral or multilateral recognized as legitimate under the article, the mea- agreements or to undertake cooperative efforts for sure was applied in a manner that constituted arbi- the conservation of sea turtles before enforcing the trary and unjustifiable discrimination between embargo, and the unilateral application of that WTO members, contrary to the requirements of the embargo, further underscored its unjustifiability. chapeau of Article XX. Furthermore, that same rigidity and inflexibility 478 Environmental Regulation and the WTO applied to the certification procedures adopted by attention of the dispute panels by the WTO secre- the United States, and this amounted to arbitrary tariat or by any of the parties to the dispute, and for discrimination. the panels to take notice of such briefs. It is feared The AB stated that "perhaps the most conspicu- that while this ruling would help NGOs get "envi- ous flaw in this measure's application relates to its ronment and sustainable development" reflected in intended and actual coercive effect on the special the WTO process, it would in the course of time policy decisions made by foreign governments, enable other interests and pressure groups, whether Members of the WTO." An interesting question of industry or of labor, to use this route as well. Fur- would be to examine whether most, if not all, trade thermore, many international environmental NGOs measures would fall into this category of measures are funded by corporate donors, and this method of with "intended and actual coercive effect." indirect influence may now increase. NGOs from The AB also ruled,"It is not acceptable in interna- developing countries appear to be more interested tional trade relations for one WTO member to use in transparency in the negotiating processes and in an economic embargo to require other WTO mem- the proposals submitted for negotiation. bers to adopt essentially the same comprehensive With the Shrimp dispute, a question has arisen as regulatory program, to achieve a certain policy goal, to whether panels should take account of amicus as that within that member's territory, without tak- briefs submitted to them by public interest groups ing into consideration different conditions which or NGOs. The relevant applicable provisions of the may occur in the territories of those other mem- DSU (Article 13.2) suggest that panels may seek bers." As there is little environmental justification information from any relevant source and may con- for harmonization of regulatory regimes, the "same sult experts to obtain their opinions on certain conditions prevail" test may be difficult to meet. aspects of the matter at hand. Pakistan has noted Added to this is the issue of the "appropriateness" of that this does not appear to suggest that the AB can the regulatory programs of the importing countries accept unsolicited briefs; Article 13.2 should there- when applied to the exporting countries. fore be clarified to state that panels and ABs should Following this interpretation, there were several not take account of unsolicited information. Many suggestions by developing countries on the scope of developing countries take the view that amicus the functions. Many developing countries expressed curiae briefs and greater NGO participation more the view that as long as no consensus has been generally would unnecessarily distort the balance in achieved in the CTE on the interpretation of Article favor of industrial countries that can more easily XX, the AB should not interpret rules in such an fund the participation of NGOs. They have also evolutionary manner as to overturn the existing pointed out that inclusion of NGO opinions and consensus. According to Pakistan's proposal to the inputs should be done through national coordina- General Council (WT/GC/W/162), it is necessary to tion; there is no need to internationalize essentially clarify the relevant provisions of the Dispute Settle- national conflicts. ment Understanding (DSU) to make clear that the responsibility for clarifying or modifying the provi- The Role of the Appellate Body. According to the sions of the WTO agreements clearly rests with DSU, the AB is only expected to examine issues of WTO members and that it would not be appropri- law covered by the panel report. In the Shrimp case ate for the AB to usurp these functions under the the AB examined de novo the facts of the case and guise of interpreting law on the basis of contempo- made a finding on legal issues that had not been rary developments. Pakistan added that the AB addressed by the panel. Pakistan has proposed that should defer to the General Council for making in all such cases the AB should be required to modifications in the relevant rules, as the member remand the case to the panel for reexamination. It countries consider appropriate. has also suggested that to avoid unnecessary delays in the settlement of the dispute as a result of such Amicus Curiae Briefs. The defendants in the remands, the panel should complete its examina- Shrimp case also claim that the AB ruling regarding tion within a period of one month. the admission of amicus curiae briefs opens the way Although the shrimp decision appears to have for NGOs to file briefs before the WTO Dispute Set- broadened the scope of measures that would be tlement Body, for these briefs to be brought to the clearly considered acceptable under the chapeau of 479 " B E H I N D - T H E - B O R D E R " A N D R E G U L AT O R Y I S S U E S Article XX and to have opened the way for seeking law over more systemic approaches, are, however, inputs from NGOs and perhaps also other lobbyists, being examined elsewhere in the framework of the it also brought to the fore the numerous grounds on WTO. Case law on the environment has shown how which such measures would be considered discrim- evolutionary dispute settlement panels may over- inatory. In fact, there may be a plethora of reasons turn a consensus. for finding trade measures for meeting nontrade As mentioned, there is an important link between objectives de facto "arbitrary and unjustifiably dis- environmental issues and concerns and the TRIPS criminatory." It is not the panel or the AB decision agreement. An example is the Convention on Bio- that is of greatest concern but, rather, the trend logical Diversity (CBD) and diverging views on the toward judicial activism that panels may inadver- patentability of plants and animals under TRIPS tently encourage. What WTO committees or negoti- Article 27.3(b). TRIPS Article 29, covering patent ating groups may fail to achieve because of lack of a disclosures, prior informed consent procedures, and consensus, panels and ABs may impose de facto. technology transfer, can also be mentioned. More Although the DSU makes clear that the dispute should also be done under the TRIPS agreement settlement process cannot add to or subtract from toward dissemination and generation of environ- the rights and obligations of the members under the mentally sound technologies. Useful partnerships WTO and its annexed agreements, through a grad- could be forged in this context with Southern ual process of citing earlier rulings and adopting NGOs such as the Southern and Eastern Africa them as their own panels and ABs may be spear- Trade, Information and Negotiation Initiative heading a process of judicial activism that may (SEANTINI), the Third World Network (TWN), eventually change the balance of rights and obliga- and Focus on the Global South, and in organiza- tions. The solution lies in review of the DSU process tions such as UNCTAD. and perhaps in creation of guidelines for the DSB. Mention should also be made of the biosafety protocol, where much more needs to be done on risk assessment techniques. A mechanism for deal- Conclusion ing with possible conflict with science should also Reconciliation or convergence can only take place be developed. Although these are complicated areas, within the existing paradigm of globalization. What work by the development community in partner- has to be understood is that the real debate on trade ship with other organizations would be beneficial. and environment, like that on globalization, is not Environmental issues also arise in the context of just about the efficiency of markets, or the impor- the General Agreement on Trade in Services tance of modern technology, or values such as envi- (GATS), in the area of environmental services; most ronment and labor. The debate, rather, is about the of the proposals made to date would benefit coun- inequality of power, for which there is much less tol- tries of the North with well-developed environmen- erance now than in the world that emerged at the end tal expertise. Developing countries, however, of World War II, when the GATT was formed.12 perform useful environmental services such as the Thus, attention should be focused on those aspects of provision of carbon sinks through their forests and the trade and environment debate that would restore are important repositories of biodiversity. These balance, such as identifying win-win scenarios in the services should be studied and a way found to make areas of market access and subsidies, improving the them tradable through the GATS. TRIPS agreement, and controlling trade in DPGs. Nobody questions the validity or urgency of Focusing attention on trade measures for environ- improving environmental policies; it is the means to mental purposes, by putting a disproportionate bur- the end that is at issue. It is of primary importance den on developing countries--not least because of that other concerns of developing countries also be the inequities of the multilateral trading system and addressed. These include serious concern about the DSU--would only exacerbate existing inequities. excessive recourse to dispute settlement by industri- Resorting to case law to arm-twist countries into al countries and the limited and back-loaded imple- accepting insupportable positions on trade and mentation of liberalization commitments for environment may prove disastrous to the dispute textiles, clothing, and agriculture. If the trade and settlement system itself. Reforms in dispute settle- environment issue is to be addressed by the WTO, it ment, and the primacy that should be given to case has to be taken up as part of the wider trade agenda, 480 Environmental Regulation and the WTO in which there are glaring inequities. Addressing ities. Thus, industrial countries took on a commitment to pro- environmental concerns as part of the implementa- vide improved access to their markets for products from devel- oping countries and to promote developing countries' access tion agenda of developing countries could generate to environmentally sound technologies and finance. more support but would necessitate more research and creative thinking to identify win-win solutions. 5 "Dispute Settlement Corner," Bridges Monthly Review 4 (3, April 2000): 5. No matter how legitimate environmental concerns are, if their trade impacts are inequitable between 6 Recently, several cases have been brought to light in which the North and South, an agenda on trade and environ- trade concessions won during the Uruguay Round are being ment will be very difficult to generate. Thus, when systematically eroded by very high and sometimes unneces- sary environmental and health standards. See the papers sub- examining issues such as environmental impact mitted to a meeting sponsored by the U.K. Department for assessments of trade measures, due consideration International Development, held in Cuba in 2000. should be given to distributional effects. Poor coun- 7 Obviously, this would be a recipe for protectionism and restric- tries should be given the economic and environmen- tions on market access for developing countries' exports, with tal space to fulfill their development aspirations.13 In inevitable negative consequences for sustainable development. such an analysis, it is important to examine not only See "Trade and Environment: A Developing Countries' Per- the environmental effects but also the income gener- spective," statement by the Dominican Republic, Egypt, Hon- ation effects of trade liberalization. The distribution- duras, and Pakistan at the WTO High-Level Symposium on Trade and Environment, Geneva, March 15­16, 1999. al effects of trade liberalization will be crucial for understanding their potential environmental effects; 8 Progress has been made in the ISO on developing guidelines it is important to have a good understanding of the on transparency, conformity assessment, and mutual recogni- effects on environmental degradation of both trade tion. liberalization and trade distortions. Solutions to the 9 In accordance with the TRIPS agreement, in order to be trade and environment debate must also be sought patentable an invention must be new, involve an inventive outside the WTO. Cooperation with international step, and be capable of industrial application. TRIPS seems to contemplate only the Northern industrialization model of organizations such as UNCTAD, the United Nations innovation. It fails to address the more informal, communal Environment Programme, and development agen- system of innovation through which farmers in the South pro- cies is an important ingredient in any recipe for duce, select, improve, and breed a diversity of crop and live- effecting a reconciliation. stock varieties. Thus, Southern germ plasm is assigned a status inferior to that of contemporary biotechnologists' varieties. The intellectual property of Southern farmers is apparently Notes denied recognition, and hence protection. See WWF, "The UN Biodiversity Convention and the WTO TRIPS Agreement," Dis- 1 See the U.S. submission to the High-Level Symposium on cussion Paper. Trade and Environment held in Geneva, March 15­16, 1999. 10 See Chakravarthi Raghavan, "NGOs Have More Rights Now 2 See "Dispute Settlement Corner," Bridges Monthly Review 3 (4, than WTO Members," South-North Development Monitor May 1999): 5 (published by the International Centre for Trade (SUNS) 4719; see also "DSU Architect Critical of Judicial and Sustainable Development; available at Activism at WTO," SUNS 4685. . 11 Appellate Body report on "United States: Import Prohibition of 3 See "Committee on Technical Barriers to Trade," Bridges Certain Shrimp and Shrimp Products," October 12, 1998, Monthly Review 2 (6, September 1998): 2. WT/DS58/AB/R. U.S.C.A. 1531­44. 4 At the "Earth Summit," it was agreed that all countries have a 12 See A. Sen, Bridges Monthly Review, September 2000. responsibility for achieving and must cooperate to achieve environmental and sustainable development objectives on the 13 See Martin Khorr, statement at the High-Level Symposium on basis of the principle of common but differentiated responsibil- Trade and Environment, Geneva, March 15­16, 1999. 481 VII T H E T R A D I N G S Y S T E M A N D D E V E L O P I N G C O U N T R I E S S ince the entry into force of the agree- constraints. Expanding the set of players involved in ment establishing the WTO, it has domestic trade policy formulation and the prepara- become clear that many governments tion of negotiating positions can help achieve this and civil society groups in developing objective, in the process enhancing the "owner- countries have been disappointed with the outcome ship" of eventual agreements. Participation is of the Uruguay Round, in terms both of market another necessary condition--but is not sufficient, access payoffs and of the implementation of certain as noted in Chapter 47, by Diana Tussie and Miguel WTO agreements. There is also a widespread per- F. Lengyel. Many developing countries have inade- ception that efforts to negotiate additional disci- quate (or no) representation in Geneva, which plines on domestic regulatory policies in the WTO impedes their active engagement in negotiations may divert attention from more critical develop- and in the day-to-day functioning of the WTO. ment-related priorities. Options have been identified that would allow poor A major theme of much of the criticism is that countries to expand their representation in Geneva more attention should be focused on ensuring at relatively low cost; for example, Blackhurst, that multilateral rules support the development of Lyakurwa, and Oyejide (2000) propose transfer of low-income countries--i.e., that the rules are not national representatives from other UN bodies to inappropriate for their institutional capacities and Geneva and more intense cooperation by members 483 T H E T R A D I N G S Y S T E M A N D D E V E L O P I N G C O U N T R I E S of regional integration arrangements. Still, only lim- where such aid would be most effective and useful ited expertise is available in most countries. requires analysis and consultations. A first innovative The costs associated with complying with certain attempt to move in this direction was taken with the WTO disciplines such as those on customs valuation revitalization of the Integrated Framework for Trade- can be significant, not so much because of the rules Related Technical Assistance, described in Chapter themselves but because of the ancillary investments 50, by David F. Luke. Although limited to least- that may be required (see Chapter 48, by J. Michael developed countries, the approach is relevant for Finger and Philip Schuler). The norms that are other countries as well in that it aims at identifying embodied in WTO agreements are often those pre- priorities on a country-by-country basis so that assis- vailing in OECD countries, implying not only that tance can be directed to those areas. implementation costs may be significant for poor countries, but also that they are asymmetrically dis- Further Reading tributed. This does not necessarily imply that WTO rules are bad from a development perspective, but Much has been written about developing countries making them work in low-income countries may and the trading system. Robert Hudec provides an require wholesale reform and strengthening of the excellent analysis of the issue, as well as a review of affected institutions. From a development perspec- relevant GATT history, in Developing Countries in the tive, the resources required might be better used for GATT Legal System (London: Trade Policy Research alternative purposes. Centre, 1987). T. N. Srinivasan, Developing Countries One option that is sometimes proposed for deal- and the Multilateral Trading System: From GATT to the ing with potential implementation problems is to Uruguay Round and the Future (New York: Harper- move toward formalization of a two-track multilat- Collins, 1998), provides a detailed analysis of the eral trading system under which not all disciplines developing country dimensions of multilateral trad- apply to all members. This is not in the interest of ing rules and the strategies pursued by developing poor countries. Experience has demonstrated that countries in the GATT and the Uruguay Round. John the payoff to seeking to opt out through a strategy Whalley, "Developing Countries and System of "special and differential treatment"--the tradi- Strengthening in the Uruguay Round," in Will Martin tional approach of developing countries--has been and Alan Winters (eds.), The Uruguay Round and the low. The opt-out strategy did, however, prevent Developing Economies (Cambridge, U.K.: Cambridge countries from being subjected to rules that University Press, 1996), summarizes developing involved significant implementation costs. What is country negotiating stances and objectives during needed is for the rules that emerge from negotia- the Uruguay Round. Rubens Ricupero, the secretary- tions to represent and advance the interests of peo- general of UNCTAD and a former trade negotiator ple in developing countries. A "one size fits all" for Brazil, discusses developing country strategies approach to regulatory policies may not be appro- and concerns during the Uruguay Round in "Integra- priate (see Chapter 49, by T. Ademola Oyejide). tion of Developing Countries into the Multilateral In many of the areas that are being proposed for Trading System," in J. Bhagwati and M. Hirsch (eds.), negotiations, a key need is to assist low-income The Uruguay Round and Beyond: Essays in Honor of countries in enhancing their capacity to trade. Arthur Dunkel (Ann Arbor: University of Michigan Many of the trade-related constraints confronting Press, 1998). Gilbert Winham, "Explanations of low-income countries cannot be addressed through Developing Country Behavior in the GATT Uruguay negotiations. These constraints are often domestic Round Negotiation," World Competition Law and Eco- policy issues that require national action--to nomics Review 21(3) (1998): 109­34, analyzes the improve the investment climate, to strengthen negotiating positions taken by developing countries domestic regulation, and so forth. Such actions in the Uruguay Round. Richard Blackhurst, William could benefit from concerted multilateral efforts in Lyakurwa, and Ademola Oyejide assess the chal- the WTO but in many cases must be complemented lenges and opportunities for enhancing the ability of by additional financial and technical assistance ("aid African countries to benefit from the WTO in for trade"), channeled through the existing institu- "Options for Improving Africa's Participation in the tions for development cooperation. Determining WTO," The World Economy 23 (2000): 491­510. 484 47 D I A N A T U S S I E M I G U E L F. L E N G Y E L Developing est for developing countries, such as market access, the real- ized gains have been more mea- Countries ger than expected. Moreover, many developing countries have confronted serious institutional and economic constraints in Turning Participation into Influence implementing some of the new disciplines. The calamitous Seat- tle ministerial meeting in 1999 added another straw to this already overloaded haystack, illustrating that international T trade relations faced serious he passage from the GATT to the problems of governance. Against this backdrop, it is WTO represented a major turning not at all surprising that developing countries have point for trade policies in developing countries. been pondering anew the dilemmas of their Full-fledged commitments were taken on for the involvement in the WTO and are searching for ways first time, clearly showing these countries' willing- to turn participation into more meaningful and ness to come out of the fringes and play by the new effective influence. It has become evident that rules and marking a major change from their most- increased participation does not always imply more ly defensive pre­Uruguay Round position on multi- effectiveness or, in the end, result in greater access to lateral trade negotiations. In previous rounds the global markets. selected developing countries that had joined the GATT had either remained quiet bystanders or had The Leap from Exclusion to Inclusion concentrated on expanding their rights to free themselves from prevailing rules. The developing For a good 40 years after World War II, most devel- countries' new engagement was not costless. First, oping countries did not perceive the GATT as a they had to accept a new approach to special and friendly or fruitful institution in which to promote differential (S&D) treatment that undermined their their interests. Inward-oriented industrialization former rights. Second, they had to make significant and nationalist ideologies of development pre- offers in order to get their own demands consid- vailed, turning trade relations into the crux of the ered, in part reflecting the "single undertaking" North-South debate. Involvement in the GATT approach of the Uruguay Round. reflected these preferences: developing countries With the benefit of hindsight, it can be argued adopted a "passive" or "defensive" attitude, refrain- that the gains from integration were less than hoped ing from significantly engaging in the exchange of for. The implementation of some commitments was reciprocal concessions. Moreover, many developing delayed and sidetracked. In areas of particular inter- countries were not members, and among those that 485 T H E T R A D I N G S Y S T E M A N D D E V E L O P I N G C O U N T R I E S were, many failed to maintain official representa- parative advantages. Furthermore, as competition tion in Geneva. among the major trading players intensified, the To a large extent the situation was reversed at the continued opening and greater contestability of beginning of the 1990s. The developing country developing countries' markets became a more high- share of GATT membership rose from 66 percent in ly valued goal. Finally, the United States was firmly 1983 to 74 percent by the late 1990s. Of the 44 new determined to extend the GATT into services and members added since 1982 (37 in 1987 alone), 43 other new areas and was no longer willing to accept were developing countries and, more recently, tran- free-riding of developing countries on such issues as sition economies (Michalopoulos 1999a). More sig- intellectual property. To sum up, either out of con- nificant, however, is the relatively active role that viction or because of fears of closing markets and many developing countries played in the Uruguay the implications of conditional most-favored- Round negotiations, not only fully participating in nation (MFN) treatment, developing countries the exchange of concessions but also advancing, on abandoned their former defensiveness and an individual or group basis, a positive agenda of embraced a much more participatory attitude. their own. This historic shift in policy and prefer- Their strategic dilemma turned from whether to ences reflected a myriad of domestic and interna- engage in the multilateral trading system to choos- tional, economic and political, developments. ing an appropriate strategy of participation, focus- Decreasing returns from and fatigue with import ing on what commitments to make and on how to substitution, together with the fall of the Berlin micromanage a bloated trade agenda. Wall, led to the de-ideologization of trade and clos- The challenges of inclusion soon proved to be er integration into the world economy. Developing highly demanding. Developing countries learned in countries began to swim with rather than against the early stages of the Uruguay Round that greater the current. The previously downplayed issue of participation did not translate automatically into market access gained increasing salience, and multi- leverage, as they found it difficult to decisively influ- lateral trade negotiations became more relevant as ence the process of agenda setting and to shape the an instrument for securing such access. A greater final outcome of negotiations. Similarly, with the awareness of the importance of a rules-based sys- expansion of the agenda through the inclusion of tem for anchoring import regimes and protecting very complex and slippery issues (services, intellec- export interests emerged in many developing coun- tual property, technical barriers, and sanitary and tries. Stepping up participation was necessary on all phytosanitary standards), many developing coun- counts. tries' capacity for analysis and for turning such At the same time, industrial countries started to analysis into sound negotiating positions was over- see the engagement of developing countries in mul- taxed (Tussie and Glover 1993). "A pro-active, con- tilateral trade talks through new lenses. The mini- structive approach was frequently out of reach for mal size of developing countries' markets had many countries because of resource and research previously been perceived as not being worth the capacity constraints" (Chadha and others 2000: effort of pressing for greater access. The result was a 432­33). situation in which developing countries had negli- The impact of these difficulties on the results of gible obligations and liberalization in sectors of the Uruguay Round should not be underestimated. export interest to them was disproportionately A balanced negotiating outcome would have small (Tussie 1987; Oyejide 2000; Ricupero 2000). required an agenda that reflected the interests of all GATT Article XVIII, section B, and the Enabling stakeholders as evenly as possible, as well as consis- Clause left developing countries with very little that tent participation by developing countries. Various needed to be done to internalize the results of nego- assessments agree that in such circumstances the tiating rounds into domestic policy. In other words, outcome would have been tilted more against trade negotiations had at best a marginal impact on industrial countries' interests. To be sure, develop- the domestic policy process in these countries. ing countries did not leave the negotiations empty- By the mid-1980s, the picture had changed signif- handed: the inclusion of agriculture, the icantly. Several developing countries became major commitment to phase out the restrictions on tex- exporters of manufactured goods, even in those sec- tiles, and the creation, with the birth of the WTO, of tors in which it had been assumed they lacked com- a much stronger dispute settlement mechanism 486 Developing Countries: Turning Participation into Influence than the one existing under the GATT can be greater value added, and commitments on subsidies deemed important gains. Yet these were more than limit the scope for implementing support policies offset by concessions: a more restrictive approach for growth and exports (Lengyel and Tussie 2000). toward special and differential treatment, commit- These by now rather classic North-South agenda ments made in the intellectual property and ser- items do not exhaust the complex menu on the vices agreements, the binding of many developing table. Following the failure to launch a new round country tariffs, and new disciplines on subsidies and of negotiations in Seattle, the intricate issue of trade customs valuation, to mention the most significant in services became the engine of WTO negotiations ones. Furthermore, the very creation of the WTO in 2000. Complexity is related to several factors: added new challenges for effective participation. negotiations are proceeding simultaneously at the The greatly expanded trade agenda called for addi- multilateral and regional levels; they involve several tional institutional capacity in member govern- sectors of great importance for the domestic econo- ments. In addition, in contrast to the GATT, my and, therefore, different producer interests; and the WTO accommodated ongoing negotiations, the results spill over into other areas such as intel- demanding constant involvement. Finally, while the lectual property rights and foreign direct invest- new dispute settlement mechanism is an asset, it ment. The weaving of negotiating positions has gave rise to a need to finance and develop expertise become highly complex, both technically and polit- on international trade law in order to take full ically. As countries diversify their exports across advantage of it. Given the level of technical exper- products and markets, it becomes more difficult for tise required, questions have been raised about them to concentrate their bargaining resources in a developing countries' capacity to bring cases effi- few selected areas. With the expansion of the negoti- ciently as complainants and to protect their inter- ating agenda, occasions for friction expand as well, ests as defendants. Even though some technical making it virtually impossible to pursue a single- assistance is available from the WTO secretariat, it is issue strategy. Argentina was among the early learn- not intended to assist developing countries on spe- ers of this lesson: it "discovered this [during the cific cases. External legal counsel usually comes Uruguay Round] when, in spite of its efforts to con- from an international law firm or consultant, at centrate on agriculture, it was drawn into bilateral considerable cost, although the creation of the disputes over intellectual property" (Tussie and Advisory Centre on WTO Law in 2001 provides Glover 1993: 231­32), with painful tradeoffs some access to subsidized legal assistance (see between these sectors. Chapter 9, by Delich; see also Weston and Delich At the same time, thinly disguised forms of 2000). administered protection have flourished. Both industrial and developing countries are resorting to safeguards, countervailing duties, and, especially, The Post­Uruguay Round Situation antidumping measures to protect producers operat- Six years after the entry into force of the WTO, ing in their domestic markets. In other words, as developing countries face new challenges and prior- multilateral negotiations reduce conventional barri- ities related to their participation in multilateral ers, trade relief policies are being used and abused. negotiations. There is a broad consensus that mar- This question goes far beyond the North-South ket access has not improved as much as expected. agenda, branching out into South-South relations. Agriculture products continue to face high protec- Indeed, in tandem with regional trade agreements, tive tariffs, as do the classic footwear, clothing, tex- an undercover war of mutually paralyzing trade tiles, and steel sectors, even after allowing for the relief measures has mushroomed. Many developing Generalized System of Preferences. In agriculture countries have actively applied trade relief measures tariffs peaks have proliferated to the point of reach- on "sensitive" sectors precisely against neighbors ing 350 percent in important export products, some that, because of geographic proximity, are able to of them particularly sensitive for developing coun- benefit most from trade liberalization. The South- tries. Industrial countries' trade policies continue to South dimension has also emerged in the context of obstruct export diversification and reduce incen- investment policies, against the backdrop of a poten- tives for processing commodities, with serious tial race to the bottom over incentives. As countries implications for export growth of products with compete to attract investments, transnational corpo- 487 T H E T R A D I N G S Y S T E M A N D D E V E L O P I N G C O U N T R I E S rations shop around for the most "market-friendly" is required. Decentralization may improve the qual- jurisdictions and engage in what has come to be ity of decisionmaking, but it can also lead to called regulatory arbitrage. Multilateral trade negoti- bureaucratic wrangling and reduced effectiveness, ations, which were shunned by many developing with the opening of a wider range of targets for lob- countries, offer an opportunity to strike a balance bying by other countries. As governments turn away between the need to avoid the irrationality of a cost- from their traditional role of regulatory control to ly race and the convenience of retaining measures to one of facilitating investment and trade, the capaci- encourage specific investments. Developing coun- ty to specify the contours of the "national interest" tries would benefit greatly from efforts to increase with a minimally adequate degree of consensus has transparency and limit competition over rules of been undermined. origin or antidumping regulations (Chudnovsky By the same token, these trends make it hard for and López 2000). developing countries to find common ground and These are but a few examples of the multiplicity build joint negotiating positions, even when there is of issues that need to be addressed. Moreover, the agreement that such cooperation could increase emphasis on reciprocity has added a previously their leverage at the bargaining table. As Narlikar absent domestic dimension. With the need to offer and Woods (2002) forcefully argue on the basis of reciprocal concessions, every international negotia- the experience of the Uruguay Round, many factors, tion has necessarily turned into a parallel domestic including the political context, the availability of negotiation whereby the gains of one sector abroad rewards and incentives, leadership, and income lev- require another sector to adjust to heightened els, heavily condition the formation of trade coali- import competition. The sensitivity of domestic tions, alliances, or any other similar collective actors to the distributional impact of trade conces- endeavor (see Box 47.1). The growing heterogeneity sions has tended to generate conflicts and resent- of trade interests within and among developing ments, adding a source of further fragility. Trade countries stemming from these trends only con- issues have acquired a salience in domestic politics tributes to lessening the likelihood of such initia- that is without precedent in the postcolonial era. tives. To be sure, the raw nerve of domestic politics today is not the same as half a century ago. In the Governance era of globalized markets, segments of production chains that used to function within national bound- The failure of the Seattle ministerial meeting in late aries are now internationally integrated. The pace of 1999 has been attributed to several factors: domestic international integration is, naturally, uneven, lead- political considerations and lack of political will to ing to tensions within sectors as different patterns of push for further liberalization on the part of the supply and investment emerge. Although the pat- United States; strong disagreements over the cover- tern may vary from sector to sector and from coun- age of an eventual new round and particularly try to country, there is widespread awareness that about the scope of agricultural liberalization; the residual protection or trade relief measures for one intense opposition of many developing countries to product add an additional cost to the next link in the inclusion of some issues--notably, labor stan- the production chain (Hoekman and Leidy 1992). dards-- in the agenda; and these countries' dissatis- The bid to have access to inputs at international faction with the agenda-setting process. Perhaps prices in order to improve competitiveness coexists even more important than the causes of the events uneasily with the quest to retain domestic market at Seattle were the consequences; it was made shares. apparent that international trade relations face seri- The complexity of the expanding agenda has also ous problems of governance. Indeed, such events led to changes in the locus of responsibility for trade called into question the decisionmaking and rule- bargaining. Traditionally, most countries assigned making processes, the level and nature of actors' this function to foreign affairs or trade ministries, participation, and the transparency and accounta- which spoke for the "national interest." As the inter- bility of the WTO. To some, "the policy differences ests of the nation, and even those of particular sec- probably could have been bridged, if the WTO's tors within it, become less clear, and as the issues decision-making process had not broken down" become more technical, input from other ministries (Schott and Watal 2000: 1). To be sure, the problems 488 Developing Countries: Turning Participation into Influence B O X 4 7 . 1 C A F É A U L A I T D I P L O M A C Y Amrita Narlikar Uruguay; and some, such as Jamaica, whose import dependence in services made efficient The prenegotiation phase (1982­86) of the international provision of services through GATT Uruguay Round marks a turning point in the his- inclusion worthwhile. The second group consist- tory of coalition formation among developing ed of small countries driven by the uncertainty of countries. It remobilized the Informal Group of their own interests and by fears of the costs of developing countries into its most formalized exclusion. incarnation, the G-10. (In its hard-line form, the The G-10 initially participated in the Jaramillo G-10 included the Big Five--Argentina, Brazil, deliberations, where members chiefly took a Egypt, India, and Yugoslavia--along with Cuba, blocking role. Subsequently, the G-10 decided to Nigeria, Nicaragua, Peru, and Tanzania.) This move its deliberations outside the Informal coalition epitomized the traditional bloc diploma- Group, led at the time by Colombia, and pre- cy of developing countries that had found several pared a draft that was presented as a fait accom- avenues through the G-77 in UNCTAD and the pli to the Informal Group and the Preparatory Non-Aligned Movement (NAM) in the General Committee (PrepCom) of the Uruguay Round. Assembly. The negotiating position of the G-10 The draft made no mention of services. Sugges- was simple and clear: members would block the tions by the Jaramillo group that the draft be dis- opening of a new trade round until traditional cussed were dismissed with the argument that issues of standstill and rollback were attended to. the G-10 had not made its submission on behalf Above all, they would resist an inclusion of new of all developing countries and represented only issues (with special reference to services) within the signatories of the draft. In reaction to these the purview of the GATT. events, the participants in the Jaramillo process The prenegotiation phase also catalyzed a sec- came together in the G-20. ond coalition, the so-called Café au Lait group, The G-20, consisting of Bangladesh, Chile, which included both industrial and developing Colombia, Côte d'Ivoire, Hong Kong (China), countries. Both coalitions arose in response to the Indonesia, Jamaica, the Republic of Korea, same issue: whether to include services within the Malaysia, Mexico, Pakistan, the Philippines, GATT. The origins of the Café au Lait group lay in Romania, Singapore, Sri Lanka, Thailand, Turkey, a 1982 decision that urged interested countries Uruguay, Zambia, and Zaire, was quite explicitly to undertake national studies on services and a negotiating coalition. The group realized that exchange information. In the absence of a GATT there was little point in repeating the G-10 exer- program for facilitating such an exchange, some cise of arriving at an independent draft and pre- industrial and developing countries came togeth- senting it as a fait accompli to other countries. If er informally to explore the issue. Colombia's stalemate was to be avoided, negotiation with ambassador to the GATT, Felipe Jaramillo, was the industrial countries would be necessary. This selected to chair the meetings. Jaramillo's posi- recognition led the G-20 to establish liaison with tion as the chairman of the GATT contracting par- the G-9 (Australia, Austria, Canada, Finland, Ice- ties established a de facto linkage of the group land, New Zealand, Norway, Sweden, and with the GATT. The institutional linkage was for- Switzerland). Under the leadership of Colombia mally acknowledged and ratified at the GATT's and Switzerland, and combining the majority of autumn 1985 session. developing and smaller industrial countries, the Two sets of developing countries were active in Café au Lait Group emerged. the "Jaramillo track." The first group, later known Café au Lait members were aware that to move as the Enthusiasts, included services exporters any agenda through the GATT process, the sup- (such as the East Asian newly industrialized coun- port of the Quad (Canada, the European Union, tries); others that hoped to gain from issue link- Japan, and the United States) would be neces- age and tradeoffs, such as Chile, Colombia, and sary. Canadian involvement in the G-9 brought (continued) 489 T H E T R A D I N G S Y S T E M A N D D E V E L O P I N G C O U N T R I E S B O X 4 7 . 1 ( C O N T I N U E D ) the Quad into consultation. On the basis of these culture of the GATT. The Café au Lait group pre- discussions, on July 17, 1986, a draft was submit- sented itself as a bridge-building coalition ted on behalf of Colombia and Switzerland to the engaged in mediation-type diplomacy in the PrepCom. Further discussions indicated that the space provided by the EU-U.S. rift on services and Café au Lait draft had come to command the the extremes of the U.S. position versus the G-10. explicit support of nearly 50 GATT members. A Its alternative name, Friends of the New Negotia- revised draft (W/47/Rev. 2, July 30, 1986) provid- tions, was indicative of this positive stance. It is ed the basis for the Punta del Este declaration and not entirely surprising that the path of issue- the launch of the Uruguay Round. By resolving based diplomacy pursued by the G-20/Café au the impasse on services, the Café au Lait group Lait was subsequently taken up with fervor by the had prevented the GATT bicycle from toppling Cairns Group of agricultural exporters. over and had successfully expressed the voice of While the Café au Lait group successfully smaller developing countries in setting the nego- demonstrated a new pattern of coalition diplo- tiating agenda. macy to developing countries and highlighted In good measure, the success of the Café au the weaknesses of the bloc-type diplomacy of the Lait reflected the strategies that the group G-10, it is noteworthy that its attempts to contin- employed. An issue-based focus gave the group a ue its activities in the Uruguay Round did not simplicity of structure that presented a marked yield any visible results. The group survived in contrast to the traditional blocs of developing various versions--the Hotel de la Paix Group, the countries, with their infinite log-rolling and shop- Friends of Services Group, the Rolle Group--but ping lists of demands. The Café au Lait, in all its it had minimal visibility and minimal successes to versions, also enjoyed a flexibility of agenda that its credit. Whether the Café au Lait represents a equipped it to work as a negotiating coalition new and lasting style of coalition diplomacy or rather than purely a proposal-making one. The merely a one-time success that relied primarily on flexibility of the Jaramillo-led agenda derived external conditions remains an open question. from the investigative process from which it Nevertheless, the sharp policy reversals that the emerged. Common interests and the emphasis group produced in the traditional bloc diplomacy on research created a virtuous cycle that of developing countries and the example it gave strengthened intragroup coherence and won of successful issue-based diplomacy make the external legitimacy for the group. Furthermore, coalition a landmark in the stance of developing in overcoming the North-South divide, the Café countries in multilateral trade negotiations. au Lait was unprecedented. The group was also particularly appropriate for the consensus-based Source: Based on work presented in Narlikar (2000). of the WTO decisionmaking process predated the and objectives currently represented, albeit faultily, Seattle meeting. The process, as during the GATT in the WTO makes it difficult to reach consensus period, moved on the basis of consensus, arrived at over the broad range of issues. through formal and informal consultations, much The grievances of many developing countries like an "old boys' club." It worked well when players with this system mounted following the conclusion were few and issues were fairly straightforward, but of the Uruguay Round, particularly regarding the over the past 15 years it seems to have gone into constraints the agreements imposed on these coun- slow motion. The image of past success, nonethe- tries' real possibilities of influencing the agenda-set- less, led to unrealistically high expectations on the ting process. Although developing countries played part of both new entrants and new activists. The an active role in the period leading up to the Seattle result is that high expectations now coexist uneasily meeting, submitting over half of the more than 250 with decreasing returns. The diversity of interests specific proposals on the agenda, they claimed that 490 Developing Countries: Turning Participation into Influence their proposals were not given due weight. The pursue new trade negotiations on a comprehensive "green-room" practice fueled the disenchantment, basis will probably have to go hand in hand with a underpinning the claims of a "democratic deficit" streamlining of the decisionmaking process that and a lack of transparency during the Seattle meet- pays due attention to the requirements of efficiency ing. The "green room" is the name given to the tra- and legitimacy. Unless these worries are addressed, ditional method used in the GATT/WTO to new negotiations will add to the frustration. expedite consultations; it involves the Director Gen- eral and a small group of members, numbering Enhancing Participation between 25 and 30 and including the major trading countries, both industrial and developing, as well as Improving developing countries' participation a number of other countries that are deemed to be involves two central dimensions. First, improving representative. The composition of the group tends skills and institutional capacity to analyze, take stock to vary by issue, but there is no objective basis for of, and manage the workings of existing agreements participation. This procedure worked when most is a precondition for designing adequate positions in developing countries were quiet bystanders. After follow-up negotiations. Second, a reform of the the significant concessions made in the Uruguay WTO decisionmaking system is needed to allow Round, developing countries felt entitled to be developing countries to increase their voice in the included in the green-room process, and on several affairs of the organization. On the first front, there is occasions they submitted declarations stating that no doubt that efforts must be made at the national they would not adhere to any consensus reached level, particularly in view of the ever-finer hetero- without their effective participation. Although the geneity of developing countries' interests. Gone are subsequent ministerial meeting, in Doha, was more the times when multilateral bargaining could be inclusive and open to all members, the issue of broadly articulated by a grand coalition of develop- effective participation remains a key one. ing countries. The need for each country to do its Perceptions of inequities in the WTO decision- own homework in following issues, to attend all making system implicitly call into question other meetings, and to have teams in capitals doing exten- facets of governance, specifically, the failure to bal- sive background research and providing adequate ance the costs and benefits arising from trade nego- instructions on all matters cannot be dismissed. The tiations. "The end result has been an absence of acquisition of sufficient knowledge about how the `ownership' of many agreements, and a general sus- system works, of technical skills, and of an adequate picion of the WTO" (Chadha and others 2000: 434). institutional capacity should be priorities. This is One reason for this has been the costs associated particularly so given that the WTO is a member- with the implementation of certain WTO agree- driven organization with a very small secretariat, ments (see Chapter 48, by Finger and Schuler, in leaving a great part of the analysis of issues and this volume). development of positions to members. To be sure, the WTO is not an international Enhanced capacity to participate also requires an organization intended to "govern" the global econo- effort to overcome the lack of coordination and the my, or even international trade relations, as a whole. turf wars at the national level that usually plague It does, however, perform some functions of gover- developing countries. Coordination problems stem nance at an international level by providing a forum from various sources,"including differences regard- for trade rule-making (legislative function); pro- ing the location of real compared to nominal tecting trade opportunities; fostering transparency authority with respect to the articulation and in the trading system; and enforcing rules through a implementation of trade policy as well as differ- dispute settlement system (judicial function). In ences in terms of which institution has the responsi- addition, there are other functions not attributed bility for trade policy and which government agency formally to the WTO that are subject to an intense has the power to negotiate and sign international international debate as to whether they should be agreements" (Oyejide, 2000: 23). The need for coor- put under its purview. Examples include the supply dination is pressing not only among government of international public goods and the subjection of agencies but also between them and business. In markets to social objectives. Given the scope of the most developing countries business tends to follow recent questioning of WTO governance, efforts to negotiations at the WTO from a considerable dis- 491 T H E T R A D I N G S Y S T E M A N D D E V E L O P I N G C O U N T R I E S tance, and its involvement is basically ad hoc and which participation would be representative of the limited. Several countries have taken initiatives to broader membership and to which responsibility improve this situation, but results are still incipient for building consensus could be delegated. and have a long way to go.1 Although this would make WTO procedures more Proposals to address these issues have stressed the efficient and equitable by linking expeditiousness need for increasing technical assistance to develop- with fair representation, it falls short of what is ing countries. The WTO itself provides assistance, needed. As Martin Wolf observed some time ago focused on training and the dissemination of infor- (Wolf 1984: 216), "There is little likelihood that mation. An important initiative was the adoption by industrial countries will grant special privileges to a high-level meeting in 1997 of the Integrated some countries without also victimizing others (or Framework (IF) for trade-related technical assis- even the same ones in different circumstances)." tance to least-developed countries. The initiative The proposed steering committee therefore involves six international agencies (the IMF, UNC- requires more "radical" reengineering. Instead of TAD, the World Bank, the International Trade Cen- centralizing the consensus-building process and tre, the United Nations Development Programme, delegating responsibility "upstream," the WTO and the WTO) that work together to help least- process as it stands today could be decentralized, developed countries integrate into the world econo- delegating the task of finding common ground my and benefit from WTO membership. This is a "downstream"-- for instance, to regionally based most welcome initiative, yet it falls short of actual committees that bring together developing country needs (see Chapter 50, by Luke, in this volume). constituencies with the mission of assembling joint Developing countries must continue the search for positions on major issues to be brought to the nego- additional self-generated, homegrown ways to build tiating table. The broadening of participation and their technical and institutional capacity. Only those interest representation would strengthen the legiti- with expertise will normally have the knowledge nec- macy of the WTO decisionmaking process without essary to codify and interpret arcane information. No undermining the quest for efficiency. This path does doubt the task is daunting and seems out of reach for not depend on a notion of reciprocity, on which a many countries if approached on an individual basis. fruitful discussion of present difficulties can only It looks more feasible if scarce financial resources are founder. What is required is to break the logjam by pooled in the context of regional groupings of coun- changing the context for discussions. After all, over tries that share many trade interests, allowing actions the past two decades, and after much soul-search- to be jointly designed, organized, and managed. Such ing, most developing countries have given sufficient an endeavor could result in a more demand-driven proof of their commitment to the world trading and customized process of technical assistance. Par- system, as well as of their willingness to sustain ticipation in this network-like effort must include trade liberalization. institutions such as universities, specialized organiza- tions, and research institutes that have knowledge on Notes trade matters or the capacity to build up knowledge within a short time. The authors wish to acknowledge debates and discussions with Some of these reflections can also be applied to the members of the Latin American Trade Network (LATN) as well the decisionmaking process--the other side of the as the able support of the coordinating unit housed at FLACSO, coin of participation. Proposals to improve the Argentina. WTO decisionmaking process have included the 1 The status in Africa and Latin America is described in Oyejide creation of a management or steering group in (2000) and Lengyel (2000), respectively. 492 48 J . M I C H A E L F I N G E R P H I L I P S C H U L E R Implementation strengthening and reform. Our analysis indicates, however, that existing WTO regulations in of WTO these areas reflect little awareness of development problems and Commitments little appreciation of the capaci- ties of developing countries to implement SPS, customs valua- tion, intellectual property, and The Development Challenge other such regulations. For most of the developing and transition economy members--some 100 countries--significant comple- A mentary investments will be t the Uruguay Round, developing required, raising the question of whether money countries took on unprecedented spent to implement WTO rules in these areas would obligations not only to reduce trade barriers but be money productively invested. This is a question also to implement significant reforms both in trade that must be posed and answered before concluding procedures (e.g., import-licensing procedures and agreements.A lesson from the Uruguay Round is that customs valuation) and in many areas of regulation not doing so can give rise to serious implementation that establish the basic business environment in the problems and to lack of "ownership" of agreements. domestic economy, such as technical standards, san- Whatever the answer to the question for any given itary and phytosanitary standards (SPS), and intel- country may be, undertaking this type of analysis lectual property law. will be ineffective if it is not complemented by This chapter discusses the second type of com- active participation in the WTO. Because of devel- mitment. These are more than policy commit- oping countries' limited capacity to participate in ments; they imply investment decisions. Their the Uruguay Round negotiations, there was little implementation will require that countries pur- support for the reforms to which WTO member- chase equipment, train people, and establish sys- ship obligates them. From their perspective, imple- tems of checks and balances, to name just a few mentation requirements were imposed in an actions. This will cost money, and the amounts imperial way, with little concern for what imple- involved can be substantial. A review of Bank proj- mentation will cost, how it will be done, or whether ect experience in the areas covered by the agree- it will support their development efforts. Thus, ments suggests that an entire year's development many developing countries have neither an eco- budget may be at stake in many of the least-devel- nomic incentive nor the political will to implement oped countries (LDCs). the obligations. Developing country institutions in the three areas In considering implementation of WTO agree- mentioned are often weak and would benefit from ments, it is helpful to ask the following questions: 493 T H E T R A D I N G S Y S T E M A N D D E V E L O P I N G C O U N T R I E S · How much will implementation cost? question as the primary basis for customs value and · What are the development problems in this area? prescribes a hierarchy of methods for determining · Does the WTO agreement correctly diagnose the that value. The first, basic option is to use as the cus- development problems? toms value the transaction value of the imported · Does the WTO agreement prescribe an appropri- merchandise--the price actually paid or payable for ate remedy? the specific shipment. The agreement lists items (add-ins) that must be included in the price actual- "Appropriate," in the last question, refers both to ly paid or payable, such as packing costs and the cost correct identification of the problem and to recog- of tools, dies, and molds provided by the buyer. The nition of the capacities (resource constraints) second alternative is to use the transaction value of of developing countries. To lend specificity to the identical merchandise sold for export to the same discussion of the scope and cost of the investments country of importation, at or about the same time, that may be involved in implementing WTO for which a transaction value can be determined. commitments, we review (primarily) World Bank The third, fourth, and fifth options are also project experience with customs reform, with appli- attempts to come as close as operationally possible cation of SPS standards, and with the installation of to the transaction value of the specific shipment. systems of intellectual property rights (IPRs). In The agreement also contains a rogue's gallery of each of the three areas, we outline basic WTO obli- methods that may not be used, such as the selling gations and examine how implementation might be price of competing domestic products, or the selling managed so as to best help developing countries use price of the goods in the market of the exporting trade as a vehicle for development. On the basis of country or in another export market. this discussion, we offer some recommendations that may help avoid future implementation prob- Presumed Administrative Environment lems. This effort must include binding commit- ments by industrial country members to furnish The valuation process the Uruguay Round agree- technical assistance to developing country members ment imposes is one that complements the customs that request it. In the Uruguay Round developing systems in place in most of the advanced trading countries took on bound commitments to imple- countries (both developing and industrial). That ment the agreements in exchange for unbound com- system is based on the generalized use of electronic mitments of assistance. This should be avoided in information management and on built-in incen- future negotiations. tives for compliance by importers. Trade in these countries takes place in large-scale lots, and duty rates are generally low. In this context, departure Customs Valuation from routine business practice (for example, The WTO Customs Valuation agreement addresses retrieval of additional information in response to a only valuation--only one part of the customs valuation inquiry) is costly. Importers themselves process. In addition to providing information on normally conduct the valuation process, including how much customs reform might cost, we argue in the application of the add-ins and take-outs needed this section that given the initial situation in many to comply with the rules. In Norway a paperless cus- developing countries, changing the valuation process toms declaration system operates around the clock; without undertaking overall customs reform is not clearance takes 15 minutes, on average, and is likely to improve the predictability of the customs almost always completed well before the goods process. Likewise, changing the customs valuation arrive. About 85 percent of declarations pass process would not significantly lessen the possibility through the system without being stopped for fur- of using the customs process as a nontariff barrier. ther investigation (WCO 1999). Investigation and verification of the importer-submitted customs value do not normally cause physical delay of the Scope and Content of the Customs Valuation shipment; instead, the importer posts a customs Agreement bond sufficient to cover the amount at issue. Finan- The Uruguay Round Customs Valuation agreement cial institutions in many developing countries do establishes the transaction value of the shipment in not offer such bonds. 494 Implementation of WTO Commitments: The Development Challenge Developing Countries' Customs Practices and · Computerization, including the introduction of Problems computerized customs systems and of systems for warehouse inventory control and statistical Customs practices in many developing countries reporting differ significantly from those in the more advanced · Improvements in valuation procedures trading nations. The differences often involve basic · Cargo controls to speed up processing and elimi- concepts, not just differences in details or efficiency. nate fraudulent or incorrect valuation Physical Control. Effective customs administra- · Refitting of customs buildings in order to permit tion has both physical and administrative dimen- the use of UNCTAD's ASYCUDA customs soft- sions. Physical control has to do with keeping track ware of what passes into and out of the country. In many · Administrative reforms, including creation of a poorer countries, traditional smuggling--goods new division responsible for customs valuation sneaked across the border, away from recognized and tariff classification; recruitment and training ports--is a significant problem. At a duty rate of 50 of staff; establishment of an appeals tribunal; and percent, the avoided duty on the number of televi- reduction of the discretion exercised by customs sion sets one person can transport on a bicycle-jit- officers ney can come, in a poor country, to a year's wages. · Provision of antismuggling and drug interdiction Where physical control systems are lax, smuggling equipment, ranging from X-ray equipment and need not even involve clandestine overland trails or gas chromatographs to communications equip- secret moonlit beaches; goods often move through ment ports without coming under the supervision of cus- · Training of management and staff in basic man- toms authorities. agement, customs procedures, and computer Administrative processes. Customs processes in operations; establishment of staff training schools poorer countries exhibit many interacting weak- · Screenings for drug interdiction nesses: excessive procedures that are not codified · Legislative reforms, including revision of laws, for- (often, not even a published schedule of current tar- mal accession to the Harmonized System Con- iff rates is available); poorly trained officials; a civil vention, and measures to increase transparency. service system that does not pay a living wage, leav- ing officials dependent on side payments for per- The reforms we reviewed cover 16 major cate- forming their functions; and ineffective provision gories of activities ranging from rewriting legisla- for appeal. Cunningham (1996), in an assessment of tion, through training in auditing procedures, to several least-developed countries that are consider- physical security in customs warehouses and polic- ing customs reform, observed that systems and pro- ing of smuggling and of traffic in illicit drugs. Com- cedures seem to have evolved so as to maximize the binations of these components may involve a cost in number of steps and approvals needed--to create as the neighborhood of US$10 million for one country. many opportunities as possible for negotiation between traders and customs officials. It should be Sanitary and Phytosanitary Standards evident from this brief account of customs prob- lems in poor countries that valuation is only an inch The SPS agreement recognizes the right of govern- in a whole yard of customs operations that need ments to restrict trade when necessary to protect improvement. human, animal, or plant life or health, but it limits exercise of that right to measures that do not unjus- tifiably discriminate between countries with the Reform Experience in Developing and same conditions, and that are not disguised restric- Transition Economies tions on trade. The agreement further obligates We present in this section a digest of our review of members to impose such restrictions only to the World Bank projects bearing on customs reform. extent necessary to protect life and health, and on Table 48.1 contains a tabulation of the cost of cus- the basis of scientific principles. Restrictions are not toms reform projects in a sampling of countries. to be maintained if scientific evidence to support Reform projects have included the following ele- them is lacking. The last point implies that SPS ments (few projects covered them all): measures can be put in place only on the basis of 495 T H E T R A D I N G S Y S T E M A N D D E V E L O P I N G C O U N T R I E S Table 48.1 Costs of World Bank Customs Reform Projects, Selected Countries Cost (millions Country Main elements of U.S. dollars) Armenia, 1993­97 Drafting of new customs laws, staff training, computerization of procedures (component of an institution-building project) 1.6 Lebanon, 1994­2001 Staff training, introduction of new tariff classifications, computerization of procedures (component of a revenue enhancement and fiscal management project) 3.8 Tanzania, 1996­2000 Provision of buildings, equipment, new processes, training 10.0 Tunisia, 1999­2004 Computerization and simplification of procedures 16.2 careful laboratory testing and analysis and if science- Article 3 of the SPS agreement specifies that SPS based concerns about food safety or serious threats measures which are in conformity with relevant to animal or plant health have been identified. international conventions are to be deemed neces- The SPS agreement requires that the process of sary to protect human, animal, or plant health and developing and enforcing SPS regulations be trans- presumed consistent with the agreement. A country parent. Governments must publish proposed regu- may adopt other standards or methods, but if it is to lations in advance and allow comment by the apply them at the border, it is required by the WTO public, including foreign exporters. Governments agreement to demonstrate their scientific merit and must notify the relevant international body of any appropriateness. Industrial countries have been lead- changes to SPS rules and must establish enquiry ers in establishing these international conventions, points so that traders can determine a country's which are, to a significant degree, generalizations of present and planned SPS regulations and processes. industrial country practices and standards. This does Before the SPS agreement came into force, an not imply that these standards are bad standards in a exporter had, in effect, to comply with the import- scientific sense. It does, however, mean that the SPS ing country's SPS measures. With the agreement in agreement lends itself to a more effective assault on force, the exporter must still comply with those SPS developing countries' use of SPS measures against measures, but the importing country is required to imports than on industrial countries' behavior. demonstrate that its SPS measures are in fact based For a developing country to effectively use the on science and are applied equally to domestic and WTO agreement to defend its export rights or justi- foreign producers. The Uruguay Round agreement fy its import restrictions, it has to upgrade its SPS puts the WTO on the side of exporters who comply; system to international standards. Effective use of the exporter now has clearer grounds for challeng- the WTO agreement depends on extensive invest- ing an import restriction. ments: it is not a matter of applying existing systems of standards to international trade but a much broader matter of installing world-class systems. A Heavier Burden for Developing Countries Although the SPS agreement does not require that a Lessons Learned from World Bank Experience country's domestic standards meet the agreement's requirements, it does require that the standards the The World Bank has assisted several countries in country applies at the border meet those require- implementing SPS regulations. Bank projects sup- ments. In this regard, the agreement probably places porting SPS systems have typically placed these a heavier burden on developing than on industrial measures in a general development context of countries because the standards already in place in ensuring food security, increasing agricultural pro- industrial countries have more or less been estab- ductivity, and protecting health, rather than focus- lished as the standard with which the developing ing on the narrower objective of meeting stringent countries must comply. requirements in export markets. 496 Implementation of WTO Commitments: The Development Challenge One SPS-related project that the Bank has sup- The costs of several SPS-related projects that the ported--for export reform in Argentina--did have World Bank has supported are reported in Table improving trade performance as an objective. The 48.2. In addition to such costs to the government, principal goal of the program was to gain interna- producers in the private sector bear other expenses tional recognition of certain zones as disease free or of complying with SPS regulations: vaccinating live- pest free. Argentina's meat, fruit, and vegetable stock, eliminating pesticide residues, guaranteeing exports have been limited by other countries' con- sanitary food processing conditions, and so on. cerns over the presence of, in particular, foot-and- mouth disease and citrus canker. In addition, the Intellectual Property Rights program recognized that if Argentina was to diver- sify into higher value added exports such as The WTO TRIPS agreement covers the seven main processed meats, seeds, and horticultural products, areas of intellectual property rights: copyright, producers would have to meet more stringent qual- trademarks, geographical indications, industrial ity control standards. Among the components of designs, patents, layout designs of integrated cir- the program that related directly to implementation cuits, and undisclosed information, including trade of SPS standards were upgrading of central and secrets. In each area the agreement specifies mini- field-level veterinary services; establishment of lab- mum standards of protection, requires govern- oratories and of quarantine stations; disease and ments to establish enforcement procedures, and pest eradication programs; certification of disease- provides means of dispute settlement.1 The mini- free and pest-free zones; training, facilities, and mum standards are similar for each of the seven equipment for seed certification and registration, areas. In the case of patents, they cover: for quality control, and for ensuring that exported meat is free of chemical residues; creation of a labo- · What is patentable. ratory to bring wool certification up to internation- · What rights flow to the owner of a patent. (The al standards; and staff and equipment for research government is obligated to prevent unauthorized aimed at reducing chemical residues. persons from using, selling, or importing the Table 48.2 Costs of SPS-Related World Bank Projects Project Cost (millions Country description of U.S. dollars) Algeria, 1988­90 Locust control project 112.0 Argentina, 1991­96 General agricultural export reform project 82.7 Brazil, 1987­1994 Livestock disease control project 108.0 China, 1993­2000 Animal and plant quarantine (component of agricultural support service project) 10.0 Hungary, 1985­91 Slaughterhouse modernization (component of integrated livestock industry project) 41.2 Madagascar, 1980­88 Livestock vaccination (component of rural development project) 11.8 Poland, 1990­95 Food-processing facilities modernization (component of agroindustries export development project) 71.0 Russia, 1992­95 Improvement of food-processing facilities and disease control (component of rehabilitation loan) 150.0 Turkey, 1992­99 Modernization of laboratories for residue control (component of agricultural research project) 3.3 Vietnam, 1994­97 Pest management (component of agricultural rehabilitation project) 3.5 497 T H E T R A D I N G S Y S T E M A N D D E V E L O P I N G C O U N T R I E S patent, the patented process, the patented prod- rightsholders can obtain the cooperation of customs uct, or the product or products directly made authorities to prevent imports of infringing goods. from the patented process.) Although it is impossible to predict how the process · What exceptions to those rights are permissible of application and interpretation through the WTO (for example, compulsory licensing may be dispute settlement mechanism will play out, a num- required). ber of legal experts believe that there is sufficient · How long the protection lasts (WTO 1999: 214 f).2 "wiggle room" in the agreement that developing countries could--within a good-faith implementa- The TRIPS agreement, like the SPS agreement, tion of their obligations--strike a balance between builds on standards set forth in relevant interna- the interests of second-comers and the need to pro- tional conventions such as the 1967 Paris Conven- mote innovation and investment (see Reichman tion for the Protection of Industrial Property and 1998 and references cited therein). This would, how- the 1989 Washington Treaty (the Treaty on Intellec- ever, require a considerable departure from the bal- tual Property in Respect of Integrated Circuits, ance that has been institutionalized in the industrial sometimes labeled the ICIP Treaty). countries' IPR law. That balance, many experts argue, is tipped toward the interests of commercial- ized producers of knowledge--and tipped past the Extension of IPR Obligations point of optimality even for the communities of The TRIPS agreement requires each WTO member interests that make up industrial country societies.4 to adhere to the provisions (with a few exceptions) Yet the tendency of the WTO is to give the benefit of the international IPR conventions, whether or of the doubt to established standards. Finding not the member is a party to those conventions. grounds for moving away from such standards is This in itself was a significant widening of obliga- particularly difficult in the area of intellectual prop- tions for many countries. For example, the coverage erty rights, which are, after all, an existential matter of integrated circuits was an extension even for of legal definition, not a scientific matter of empiri- some industrial countries.3 Under the TRIPS agree- cal estimation. ment, WTO members must consider unlawful--if not authorized by the rightsholder--the import, How to Do It sale, or other commercial distribution of the inte- grated circuit design, of integrated circuits contain- Even for an individual country, it would be nigh on ing that design, and of articles that contain such impossible to provide objective guidelines on how to integrated circuits. strike the optimal balance between legal incentives As another example, the Rome Convention, which to create and the costs incurred by users and poten- establishes rights of performers, producers of sound tial second-comers as a result of protecting IPRs. recordings, and broadcasters, has few signatories, Systems in place have to be seen as the outcome of particularly among developing countries. The accepted (e.g., democratic) political processes, not of TRIPS agreement, however, creates obligations for scientific calibration. It would be even more difficult governments to allow recording companies from to adjust this balance to different levels of economic one country to attack unauthorized reproduction development. Analysts have so far built up little and sale of their products within another country. In knowledge about the impacts of various forms of some areas the TRIPS agreement has broader cover- IPRs on economic development, much less about age than the relevant international convention. For the effects of different degrees of any of these forms.5 example, it goes beyond the Berne Convention by Our review of World Bank projects in support of requiring copyright protection for certain computer IPRs again shows a considerable range of needed programs and computerized databases, and it con- reforms--new legislation (e.g., to extend IPR pro- tains the first multilateral obligations concerning tection to plant varieties), improvement of adminis- industrial designs (e.g., textile designs). trative structures (capacity to review applications, The enforcement provisions of the TRIPS agree- including the introduction of computerized infor- ment require that a member provide civil as well as mation systems and extensive training for staff), criminal remedies for infringement of IPRs. They and better enforcement. Some information on the also obligate members to provide means by which associated costs is reported in Table 48.3. 498 Implementation of WTO Commitments: The Development Challenge Table 48.3 World Bank Projects Related to Intellectual Property Rights Project Cost (millions Country description of U.S. dollars) Brazil, 1997­2002 Training for staff administering IPR laws (component of science and technology reform project) 4.0 Indonesia, 1997­2003 Improvement of IPR regulatory framework (component of information infrastructure development project) 14.7 Mexico, 1992­96 Establishment of agency to implement industrial property laws (component of science and technology infrastructure project) 32.1 Lessons Learned versa), the WTO adds no more than an obligation to apply their domestic regulations fairly at the border. The following are the main points that emerge from This includes not discriminating among transactions our review. involving different countries and not unnecessarily A need for reform. In the areas we have covered-- impeding international transactions. Countries that customs administration, sanitary and phytosanitary at present apply their own indigenous standards have standards, and intellectual property rights--there the additional--and far larger--obligation to apply was no shortage of World Bank projects to review. the internationally sanctioned standards in their Developing countries are willing to borrow money domestic economies. Although new WTO areas such to finance improvements in these areas. It is evident as SPS and IPRs aim at the trade-related aspects of that they themselves see a need for reform. their subject matter, their implementation by devel- The message from industrial countries: "Do it my oping countries requires, first of all, the establish- way!" The content of the obligations imposed by the ment of such systems, or the conversion of WTO agreements on customs valuation, SPS, and indigenous systems to those recognized by interna- intellectual property rights can be summed up as tional conventions. the advanced countries saying to the others, "Do it A related lesson is that the scope of what the WTO my way!" The Customs Valuation agreement impos- regulates is narrower than the scope of what must es on all countries a system in use in the leading be done to make development sense out of imple- industrial countries; the TRIPS and SPS agreements mentation. Customs valuation versus customs explicitly establish as the WTO standard interna- reform is an example: it helps little to change cus- tional conventions developed in large part by the toms valuation procedures if containers still sit on industrial countries. the dock for 60 days. Although the SPS agreement appears to allow Inappropriate diagnosis and inappropriate reme- countries to retain indigenous systems, doing so is dies. One effect of the "do it my way!" nature of the not a real alternative. In defending trade-related agreements is to intensify the ownership problem. actions, the systems recognized by international This characteristic also returns us to our initial conventions have the legal benefit of the doubt; an questions. From a development perspective: indigenous system must prove itself. The developing countries do not have the necessary resources to · Do the WTO agreements appropriately identify defend their systems, and so the only effective option the problems faced by developing countries? for a country that retains an indigenous system of · Given countries' needs and their resource bases, standards is not to apply standards at the border at do the agreements provide the most effective all.6 The WTO's free-rider problem has not gone remedies? away; it has been swapped for a forced-rider prob- lem, and the burden has been shifted from the The Customs Valuation agreement provides nei- industrial countries to the developing countries. ther appropriate diagnosis nor appropriate reme- For the advanced countries whose systems are dies. It addresses only a small part of most compatible with international conventions (or vice developing countries' problems with customs 499 T H E T R A D I N G S Y S T E M A N D D E V E L O P I N G C O U N T R I E S administration and provides no remedy for other ultimately be responsible for implementation. How parts. For the small part of the problem it covers, it the developing countries organize their participa- provides an inappropriate remedy, one incompati- tion in WTO affairs needs modification, but per- ble with the resources many developing countries haps so does the WTO process itself. have at their disposal. As discussed in Chapters 47 and 49 in this vol- Our conclusions on the TRIPS agreement are ume, participation in the WTO is necessary but not similar. As to diagnosis, its focus is not on encourag- sufficient for realizing outcomes that are perceived ing innovation or protecting indigenous technology to be in the national interest. Addressing the "devel- in developing countries; rather, it is on industrial opment credibility" deficit requires, first and fore- country enterprises' ability to collect for IPRs on most, that there be a good understanding of the which many countries did not recognize any obliga- relevance and implications of multilateral rule- tion to pay before the Uruguay Round. The default making, as well as proactive participation in the remedy is to imitate industrial countries' intellectu- negotiating process. Better understanding of the al property law. Although legal scholars point out issues in developing countries is required not just by that the TRIPS agreement allows for the possibility government officials but also by the private sector of adopting intellectual property law that is friendly and civil society. Greater explicit attention needs to to users and to second-comers, they also note that be devoted to analyzing the economic relevance and the benefit of the doubt is on the side of copying implications of proposed rules. This is something present industrial country approaches. A major cost that must be done by national stakeholders, with of standardizing on the prevailing industrial coun- due consideration given to how alternative rules try norms is the thwarting of experimentation that might best be applied and enforced, how transition- could lead to legal approaches more appropriate for al arrangements can be linked to national capacities developing countries.7 to implement agreements, and what types of corol- Effective implementation and compliance involve lary investment are needed. investment in development projects, but WTO Each of the three Uruguay Round agreements we negotiations have not supported examination of have reviewed includes a promise of assistance with actual and proposed agreements from this perspec- implementation. In addition, each provides for tive. The dynamic behind the WTO process has delayed implementation and sets forth a procedure been the export interests of major enterprises in the whereby a developing country can request an exten- advanced trading countries. Development min- sion beyond the agreement's deadlines. The latter istries in those countries frequently complain about provision might be interpreted as recognition that how hard it is to get their trade ministries to pay the prescribed or default technology included in the attention to development issues; the development agreements might not be the most suitable for these ministries are junior partners in making trade poli- countries. Although the agreements allow for the cy. Meanwhile, at the WTO many developing coun- possibility that other approaches might be devel- tries have little capacity to organize and to advance oped and recognized, they provide no such alterna- their own interests. tive. As for finding alternatives, it must be recognized that WTO negotiations are propelled by self-interest. Narrowly interpreted, that places the Conclusions burden of designing alternatives that are appropri- Avoiding implementation problems of the type that ate to developing countries' needs and resources on arose after the Uruguay Round requires that there those countries themselves. be much greater ownership of negotiated agree- Determining the costs and benefits of implemen- ments. The absence of instinctive ownership of the tation is clearly of great importance. The project reforms needed to comply with WTO obligations costs presented here are just a first approximation of will make implementation very difficult and is like- the investments needed to implement WTO obliga- ly to push governments toward superficial adjust- tions on SPS, IPRs, and customs reform in order to ments aimed at avoiding clashes with trading ensure that implementation benefits the country. partners. Private and social sector shareholders Implementation costs go beyond what is needed to must be involved in the creation of WTO obliga- simply pass new legislation or ministerial decrees. tions--not just the government agencies that will Often, very significant investment requirements are 500 Implementation of WTO Commitments: The Development Challenge associated with upgrading the capacity of the rele- implementation. They agreed in their declaration vant institutions (see Box 48.1 for estimates for that where negotiations are undertaken, the relevant Jamaica). To gain acceptance for its meat, vegetables, implementation issues would be addressed in those and fruit in industrial country markets, Argentina negotiations and that other implementation issues invested over US$80 million in achieving higher lev- in other areas would be addressed as priority mat- els of plant and animal sanitation. Hungary spent ters by the relevant WTO bodies. In an accompany- more than US$40 million to upgrade the level of ing ministerial decision (WT/MIN(01)/W/10), sanitation of its slaughterhouses alone. Mexico spent ministers addressed 46 individual points, principal- over US$30 million to upgrade intellectual property ly in the regulatory areas reviewed in this paper but laws and enforcement that were already at higher touching as well on the market access commitments levels than is the case in most least-developed coun- of the Uruguay Round. tries. We identified 16 elements in customs reform, Although these documents do convey the minis- each of which can cost more than US$2.5 million to ters' reading of the importance of implementation implement.8 Although the amounts will vary by questions, they also reflect the international com- country, the totals involved in effective implementa- munity's lack of advancement in thinking about the tion of WTO agreements can exceed the annual "development dimension" of the behind-the-border development budget of least-developed countries. or "new" WTO areas in ways beyond the concep- This suggests that in future negotiations, efforts tions built into the existing agreements. Sixteen of must be made to obtain binding commitments to the individual decisions encourage use of special provide adequate financial and technical assistance and differential treatment provisions in the WTO for implementation (Finger and Schuler 2000). agreements, and 10 of them reaffirm phase-in pro- Finally, we should be careful not to be lulled into visions. Ten call attention to unbound provisions the mercantilist ethic of reciprocal negotiations in for technical assistance for developing countries; which delay is itself victory. Nor should we ignore the one decision reaffirms that industrial country potential benefits of pursuing multilateral reform. members have a legal obligation under TRIPS Arti- Where reform is needed, to delay improvements is to cle 66.2 to provide their enterprises and institutions prolong the time that people in developing countries with incentives to promote technology transfer to remain poor. Time will, of course, be needed for LDCs. Eight decisions call for further review to clar- implementation, but implementation periods should ify certain parts of the antidumping, subsidies, and be based on the engineering requirements for accom- TRIPS agreements. Three possibilities are covered plishing the required infrastructure improvements by the decisions: technical assistance to do what the and making the investments associated with comple- WTO agreements mandate; longer transition peri- mentary, supporting institutional strengthening, not ods for doing it; and in some cases, special and dif- handed out as a second prize in a tough negotiation. ferential permission not to do it. There is a need to distinguish between agreements In the upcoming trade negotiations, more imagi- that require real investments--both direct and indi- native thinking will be needed to examine reforms rect--and agreements that can be implemented by in the WTO new areas in ways that make sense from the stroke of a pen (see also Chapter 49, by Oyejide, a development perspective. For example, the decla- in this volume). Both require careful analysis, but ration encourages greater developing country only the first gives rise to legitimate concerns regard- membership in international standard-setting ing the costs of implementation. organizations such as the Codex Alimentarius. The negotiations should go on to devise indigenous alternatives to the default regulatory framework Postscript: Implementation and the Doha that existing international standards embody. Trade Round ministers at Doha (implicitly) called attention to At the WTO ministerial meeting in Doha in 2001, the need for such thinking; the research community trade ministers paid considerable attention to should interpret this as a challenge to provide it. 501 T H E T R A D I N G S Y S T E M A N D D E V E L O P I N G C O U N T R I E S B O X 4 8 . 1 C O S T S O F I M P L E M E N T I N G T H E W T O A G R E E M E N T S : T H E J A M A I C A N E X P E R I E N C E This box provides some (conservative) estimates of the residue analysis, and microbiology, and provision cost to Jamaica of implementing selected WTO agree- of training in lab methodology, quality manage- ments.* The estimates do not take into account the ment, and use of equipment (US$500,000); con- need for and cost of ancillary investments and reforms duct of pest surveys, surveillance, and monitoring that may be required to support implementation. (US$250,000); establishment of an enquiry point (US$150,000); creation and strengthening of Agreement on Trade-Related Aspects of Intellectual inspection facilities at ports of entry and exit to Property Rights (TRIPS) serve all the agencies involved in the certification Implementation of the TRIPS agreement in Jamaica of food imports and exports, with provision for will require an initial public sector investment of at additional staff, training, and equipment to detect least US$1million. This is associated with upgrading high-risk materials in shipments (US$500,000); and modernization of the domestic intellectual prop- and funding for participation in international stan- erty rights (IPRs) system. Modernization involves: dard-setting meetings, working groups, and the Committee on SPS Measures (US$30,000). Many · Revision of existing laws (on, for example, of these costs will be recurring. trademarks and patents) · Enactment of new laws on layout designs, geo- Customs Valuation graphical indications, plant varieties, and so on The estimated initial cost of implementing the · Development of proper administrative structures WTO Customs Valuation agreement is and officers to implement intellectual property US$840,000, most of which is needed for train- (IP) procedures and policies as required by the ing (US$120,000), computing equipment and legislation and the government. databases (US$50,000), and increased staffing A new Intellectual Property Office (JIPO) will (US$600,000). Staffing costs will be recurring. administer the laws required by the TRIPS agree- ment. Approximately US$437,500 was already allo- Conclusion cated to the existing IPR structures; another These are very rough estimates and are quite con- US$875,000­US$1.25 million will be required to servative. Nevertheless, they illustrate that for a develop the JIPO so that it covers all TRIPS areas. The small developing country the implementation of the estimated costs required for establishing the JIPO WTO agreements can be a substantial undertaking come to about US$775,000 a year over five years, that requires both technical and financial assistance. after which the office is expected to be nearly self- Implementation of the TRIPS, SPS, and Customs Val- sufficient. An additional estimated US$250,000 is uation agreements narrowly defined--that is, needed to jump-start the implementation of major excluding ancillary and complementary invest- enforcement programs, including border controls; ments to improve customs and standards institu- these are likely to be recurring costs. tions, and ignoring all costs for businesses--will require at least US$10 million. These financial Agreement on Sanitary and Phytosanitary Measures resources are in addition to what is already budget- (SPS) ed by the government or allocated by donor coun- Implementation of the SPS agreement will require tries as overseas development assistance to Jamaica. a total of US$7.6 million. This includes revision of current laws and regulations to make them WTO- * The cost estimates were originally made in Jamaican compliant (US$200,000); establishment of an dollars and were converted to U.S. dollars at the rate of Agriculture Health and Food Safety Authority to US$1 = J$45. administer and coordinate SPS activities (US$6 mil- Source: Prepared by the volume editors, based on a lion); upgrading and equipping of existing labora- paper prepared by the Ministry of Foreign Affairs and tories in areas such as pest identification, pesticide Foreign Trade, Jamaica. 502 Implementation of WTO Commitments: The Development Challenge Notes 4 Reichman, for example, asserts that "the logical course of action for the developing countries in implementing their obli- gations under the TRIPS Agreement is to shoulder the pro- This chapter draws heavily on Finger and Schuler (2000). competitive mantle that the developed countries have increas- 1 One critic argues that "these [are] not `minimum' standards of ingly abandoned" (Reichman 1998: 606). Templeman (1998) intellectual property protection in the classical sense of the argues that there is no public justification for the level of intel- term; rather, they collectively expressed most of the standards lectual property protection defined by industrial countries' of protection on which the developed counties could agree laws. among themselves" (Reichman 1998: 603). 5 Abbott (1998a: 501), in his introduction and summary in an 2 The TRIPS agreement provided the following transition periods: issue of the Journal of International Economic Law devoted to industrial countries, until January 1, 1996; developing countries the TRIPS agreement, notes this lack of understanding of the and transition economies, up to January 1, 2000; least-devel- impact of IPRs on economic development. oped countries, up to January 1, 2006. The transition period for 6 We are not arguing here that the iron fist imposes the wrong a least-developed country may be extended on a "duly moti- standards. Our concern is to remove the velvet glove of com- vated" request by the country. Developing countries that cur- forting rhetoric from that fist. rently provide patent protection to processes and not to products--for example in the food, chemicals, and pharmaceu- 7 Matthew Stillwell of the Center for International Environmental tical sectors--can delay until January 1, 2005, the application of Law pointed this out to us. the obligation to protect products. Even here, governments 8 The experiences we have reviewed were in fairly large and must ensure that inventions made between 1995 and 2004 will more advanced developing countries; the costs could be high- be able to gain patent protection after January 1, 2005. er in least-developed countries of a similar size that begin far- 3 This treaty is not yet in force; thus far, it has only nine signato- ther from the required standards, and they may be lower in ries, of which only one has ratified. much smaller countries. 503 49 T. A D E M O L A O Y E J I D E Special and have derived as much benefit from the various rounds of mul- tilateral trade negotiations as Differential expected and that very few of them actually participate effec- Treatment tively in the WTO process (Hudec 1987). More specifically, a disproportionate increase in the exports generated by these trade negotiations has accrued to the industrial countries, part- ly because the negotiations have typically reduced tariffs on I products of export interest to n broad terms,WTO provisions relat- industrial countries more sharply than on products ing to various elements of special of interest to developing and least-developed coun- and differential (S&D) treatment constitute a set of tries. The latter groups of countries have typically rights and privileges that apply to developing and been handicapped in negotiating reductions of the least-developed country members and from which high tariffs facing their exports because they are industrial countries are excluded. In effect, these often not principal suppliers. Furthermore, they provisions are meant to grant developing countries may have little to offer in multilateral trade negotia- and least-developed countries (LDCs) more favor- tions because their "essential" imports of capital able access to the markets of the industrial countries and intermediate goods already carry zero or mini- and to give them substantial policy discretion with mal tariffs, while their heavy reliance on trade taxes respect to their own domestic markets. as sources of fiscal revenue often restricts the extent In principle, the existence of S&D provisions in to which they can reduce these tariffs as "conces- the GATT/WTO framework reflects the recognition sions" in the negotiating process. Finally, their that the multilateral trading system consists of capacity to participate effectively in the WTO countries at markedly different levels of develop- process is constrained by their limited leverage, ment. Because of disparities in economic situation which arises from a series of factors, including the and capacities, there are significant differences in small size of their economies, the limited number of the benefits that countries reap from the global their export commodities, their greater vulnerabili- trading system. S&D provisions are aimed at relat- ty to terms of trade shocks, their endemic balance of ing these differences to the obligations and commit- payments problems, and their limited human and ments that different categories of member countries institutional capacity. are expected to undertake. Broadly reflecting these concerns, S&D provisions There appears to be some concern that many are designed to accomplish two objectives: (a) to developing and least-developed countries may not enhance the market access conditions facing the 504 Special and Differential Treatment beneficiary countries, and (b) to exempt them from sions. In particular, it provided permanent legal certain multilateral trade disciplines and thus give cover for the Generalized System of Preferences them some flexibility in the use of various trade and (GSP), identified the LDCs as a separate category of trade-related measures. In operational terms, GATT members meriting more favorable treatment enhanced market access has been implemented than other developing countries, and codified the through trade preferences offered by the industrial "graduation" principle by which developing coun- countries on an individual basis to specific develop- tries would be expected to take on more of the obli- ing and least-developed countries. The right of the gations of GATT membership as their economies developing and least-developed countries to regu- grew stronger. In specifying the S&D provisions late access to their own markets is operationalized applicable under the Tokyo Round codes, the through the maintenance of trade barriers and Framework Agreement identified three special through substantial exemption from several modalities for S&D treatment: (a) the offer of tech- GATT/WTO disciplines. The exemptions enable nical assistance to developing countries to help them to use quantitative import restrictions for them comply with the new rules; (b) the granting of both infant industry protection and balance of pay- the right to weaker disciplines for developing coun- ment reasons; to establish preferential regional tries in certain respects; and (c) the granting of trading arrangements among themselves; and to exemptions from some of the new obligations on benefit from tariff reductions achieved in the the grounds that the developing countries con- process of multilateral trade negotiations, in accor- cerned faced limitations of administrative and dance with the most-favored-nation (MFN) princi- implementation capacity. ple, but without reciprocity. The two sets of S&D Up until the Uruguay Round of multilateral trade provisions are obviously interrelated and comple- negotiations, the key S&D provisions that granted mentary. The derogation from certain rules ensures special trade policy discretion to the developing and that beneficiary countries are not deprived of the least-developed countries were codified in GATT essential tools for strengthening their export supply Article XVIII. Section A of this article permitted capacity, without which they may not be able to take developing countries to modify previously negotiat- full advantage of the offer of preferential access to ed tariff bindings in order to assist the establish- industrial country markets. ment of a particular industry; section B allowed the use of quantitative import restrictions by develop- ing countries as an instrument for dealing with bal- Before the Uruguay Round ance of payments problems; and section C The traditional S&D treatment strategy was devel- permitted developing countries to use quantitative oped between the mid-1960s and the mid-1980s. In import restrictions for infant industry protection. particular, the principle and objectives of a general- Of the three, sections A and C were subject to com- ized and nonreciprocal system of trade preferences pensation or retaliation, and hence section B was for developing countries received approval in 1968, much more widely invoked than the others. Finally, during UNCTAD II. Eventually, by its decision of GATT Article XXVIII bis (3) allowed developing June 25, 1971, the GATT provided legal backing for countries not to offer full reciprocity for negotiating the UNCTAD agreement. In effect, the GATT concessions made by industrial countries, in recog- approved a waiver of the provisions of GATT Article nition of the developing countries' need to use I for a period of 10 years, thus enabling its industri- import duties for general economic development al country members to offer trade preferences to and fiscal purposes. developing countries without offending the MFN Enhanced market access through the GSP was principle. one major component of S&D treatment that was It can be claimed that S&D treatment provisions expected to produce concrete results. The benefits, reached their peak during the Tokyo Round (Wolf however, have been limited by the typically narrow 1984). The 1979 Framework Agreement on Differ- product coverage of the GSP, by restrictive rules of ential and More Favorable Treatment, Reciprocity, origin, and by the application of safeguard measures and Fuller Participation of Developing Countries, by some preference-granting countries. Further- also known as the Enabling Clause, offers a fairly more, the GSP is not a multilateral agreement, and comprehensive statement on the core S&D provi- preference-granting countries have exercised the 505 T H E T R A D I N G S Y S T E M A N D D E V E L O P I N G C O U N T R I E S right to exclude or graduate specific developing oping countries were to implement the new agree- countries from GSP benefits. During the first 10 ments. In other words, the intent of the Uruguay years of the scheme (1968­78), less than 11 percent Round agreements is that developing countries of eligible imports actually received GSP treatment. should eventually meet virtually the same set of A decade later, in 1988, only 27 percent of all standards as industrial countries on a broad range dutiable imports was granted preferential access. of market access issues. The benefits derived by developing countries Thus, many post­Uruguay Round S&D provi- from the GSP have thus been quite small in relation sions are expressed in terms of transition periods to the total exports of developing countries, and and differences in threshold levels. That is, the they have been heavily concentrated in a few benefi- Uruguay Round agreements specify how soon and ciaries. Up to the mid-1980s, three economies-- to what extent industrial and developing countries Hong Kong (China), the Republic of Korea, and should meet their obligations. Some of the agree- Taiwan (China)--accounted for about 45 percent of ments add nonmandatory offers of technical assis- total GSP gains. This concentrated nature of GSP tance to help developing countries fulfill their benefits remained unchanged through the early commitments. The implied eventual convergence in 1990s, as 6 to 12 of the largest beneficiaries claimed standards of behavior for industrial and developing 71 to 80 percent of the total. countries applies, in particular, in such areas as the Both of the key components of S&D provisions use of quantitative trade restrictions, offers of spe- have been criticized in the literature (Wang and Win- cial assistance to producers, tariff binding, and reci- ters 1997). It has been argued that the component procity. which grants greater flexibility to developing coun- For example, the use by developing countries of tries in their use of trade policy instruments is coun- quantitative restrictions for dealing with balance of terproductive because these "market-distorting" payments problems has been constrained by the measures impose a self-inflicted cost on developing imposition of more stringent rules and procedures. countries' own economies, while nonreciprocity may A schedule for removing existing quantitative prelude their use of the GATT/WTO framework as restrictions has to be publicly announced, and there an "agency of restraint." The GSP component does is an explicit preference for price-based measures not provide, it is argued, a stable and reliable basis for for curtailing imports. Where the use of quantita- investment, and it promotes production and trade tive import restrictions is justified, they must be inefficiencies in the beneficiary countries. These crit- limited in duration and be applied on a nondis- icisms foreshadowed developments regarding S&D criminatory basis. Similarly, the right of developing treatment during the Uruguay Round negotiations, countries with per capita income equal to or greater in particular. than US$1,000 to use export subsidies has been sharply curtailed, as they are required to eliminate export subsidies by 2003. The Uruguay Round Several Uruguay Round agreements appear to The launch statement that began the Uruguay preserve some of the old S&D treatment provisions. Round in September 1986 contained an explicit For instance, the Agreement on Technical Barriers understanding that developing countries would be to Trade includes a statement that developing coun- accorded S&D treatment in the negotiations in tries are not required to use international standards accordance with the terms of the 1979 Framework which are not appropriate for their needs or which Agreement. But the adoption of the "single under- may hinder the preservation of indigenous technol- taking" as the guiding principle for the round ogy. Similarly, the provisions on safeguards exempt ensured that the agreements coming out of the a developing country's exports from countervailing negotiations would radically change the form and measures as long as its share of total imports of the content of most of the key elements of the second product is 4 percent or less. Most provisions for dimension of S&D provisions. In particular, the S&D treatment of LDCs survived the changes intro- agreements had the effect of reducing the scope of duced in the Uruguay Round; perhaps the single many of the existing S&D provisions, and the sur- most important S&D provision to survive without viving provisions were reformulated essentially in modification is the GSP. But the Uruguay Round the form of longer time periods within which devel- did nothing to eliminate or even reduce many of the 506 Special and Differential Treatment limitations (including the unilateral nature of the It is obvious that the limited duration of the transi- scheme) that have traditionally curtailed the bene- tion periods used to reflect S&D "concessions" in fits derivable from the GSP. many Uruguay Round agreements renders them It may be concluded, therefore, that in general the both inadequate and inappropriate as a basis for Uruguay Round essentially reduced S&D treatment building capacity for enhanced production and for developing countries to extended transition trade in low-income countries. periods over which developing countries would Redefining S&D treatment also requires multilat- assume the same levels and scope of obligations as eral agreement regarding the classification of WTO industrial countries. But the setting of transition member countries and the measurable develop- periods and threshold levels appears haphazard and ment, trade, and other parameters that should be ad hoc and is not closely linked to objective criteria used in this categorization. Currently, the WTO reflecting differences in levels of development or a appears to recognize (implicitly, at least) three cate- country's institutional and human capacity. In the gories of countries in its membership: industrial, light of experience with implementation following developing, and least developed. The WTO indirect- the Uruguay Round, the transition periods and ly defines the least-developed countries by adopting threshold levels appear to have been excessively the UN list. This list, however, is defective for at least optimistic in many cases. two reasons. First, it is based on income and hence does not necessarily reflect trade competitiveness, with which the WTO is (or should be) concerned. Redefining Special and Differential Second, it excludes several very low income coun- Treatment tries. This may be why the agreement on subsidies The deficiencies associated with post­Uruguay expands the UN list to include other countries with Round S&D treatment provisions suggest the need per capita income of up to US$1,000. As for "devel- for a careful rethinking of the concept--of its justi- oping countries," the WTO has no specific defini- fication, form, and content. The absence of such tion. In practice, it falls back on an implicit rethinking during the Uruguay Round probably led self-designation arrangement that permits coun- to the patchwork nature of the post­Uruguay tries to so describe themselves. Round S&D provisions. For example, the adoption An explicit categorization of WTO member and wholesale use of transition periods appear not countries based on a multilaterally agreed set of to have been carefully thought through. The transi- measurable criteria could also address the questions tion periods are probably meant to reflect the costs of which countries are eligible for S&D treatment of changes in trade policy rules for an economy. But and which countries should be graduated out of there are typically at least three different types of which S&D provisions, and when. The Uruguay cost: the costs of adjustment, of implementation, Round agreement on subsidies offers an example. and of compliance. Some policy changes (e.g., tariff By categorizing beneficiaries according to per capita rate reduction) may be associated with minimal income, it was able to express the graduation implementation and compliance costs, although the threshold in terms of measurable economic indica- adjustment cost could be high if the reduction is tors (such as exceeding a specified per capita large and is implemented quickly. A long transition income over three consecutive years or achieving a (implementation) period could be a way of reduc- specified export share) rather than by assigning a ing (or perhaps spreading out) the adjustment cost. transition period. Thus, a solution to the problems By comparison, a policy change that mandates associated with country categorization and gradua- increased protection of intellectual property rights tion could be the adoption and generalization of the could be associated with high costs of implementa- principle used in the agreement on subsidies. As an tion, compliance, and adjustment, to the extent that alternative, the WTO might consider adopting the it involves human and institutional capacity build- World Bank's classification of countries as low ing for implementation and compliance, in addition income, middle income, and high income. This to the cost of adjustment. In such a case the use of a method has at least two advantages: it is determined transition period may, by itself, be neither wholly in a transparent way, and it is widely accepted. The adequate nor appropriate for taking account of the income-based indicator could be supplemented by full costs associated with the policy or rule change. a measure of trade competitiveness (for example, 507 T H E T R A D I N G S Y S T E M A N D D E V E L O P I N G C O U N T R I E S manufactured products as a percentage of total cially if current limitations regarding product cover- exports) to distinguish between least-developed age, rules of origin, and the unilateral nature of the countries (less than 20 percent, in the example), schemes could be eliminated in the context of the developing countries (20 to 40 percent), and indus- decisions to grant duty-free, quota-free, and multi- trial countries (more than 40 percent). laterally bound access to industrial country markets A redefinition of S&D treatment also requires the for all LDC exports. This could be made more fully identification and negotiation of the multilateral multilateral if the developing countries could also rules from which full or partial derogation should extend multilaterally bound preferential market be granted to the least-developed and developing access to LDC exports--for example, by offering to country categories. LDCs should probably be grant- halve applied tariffs for these countries. The burden ed full derogation (as is essentially the case current- of this special market access scheme on the industri- ly) except for some obligations, such as tariff al and developing countries is likely to be small, binding and negotiated and phased tariff reduction, given the rather low share of the total export market that would commit them to a regime of rational and accounted for by the LDCs. Developing countries' sound trade policy conducive to their own econom- sharing in this "burden" would be an important way ic growth and development. not only to demonstrate "South-South solidarity" Finally, special market access through trade pref- but also to demonstrate the readiness of the multi- erences has historically been an important compo- lateral trading system to accommodate the needs of nent of S&D treatment. Its actual benefits have fallen its different categories of members. far short of the potential because of the many limita- The various S&D provisions that evolved in the tions of the GSP scheme. Negotiated MFN tariff GATT/WTO framework were established in response reductions have also reduced preferential trade mar- to the perceived special problems of the developing gins. Still, the continued importance of special mar- and least-developed countries. Their continued rele- ket access arrangements should not be vance must be shaped by a process of redefinition underestimated. They could provide an important that pays attention to the changing nature and signif- boost to the exports of low-income countries, espe- icance of these special features and problems. 508 50 D A V I D F. L U K E Trade-Related manage a dynamic trade policy and to meet the demands of par- ticipation in the WTO frame- Capacity Building work (see Ohiorhenuan 1998; Blackhurst 1999a). The second for Enhanced issue arises from the declining competitiveness of African African economies as expressed through their falling share in world trade, Participation in down to 2 percent by the end of the 1990s from 4.2 percent in 1985. The region's share of the Global world manufactured exports is almost negligible (World Bank Economy 2000a: 208; AfDB 2000). Building on efforts at policy reform and trade liberalization during the 1980s and 1990s, there has been what may be T described as two waves of his chapter examines issues sur- responses to the need for trade-related capacity rounding trade-related capacity building in African countries. The WTO, UNC- building and technical assistance for the enhanced TAD, and the International Trade Centre (ITC), participation of African countries in the interna- sometimes referred to as the three Geneva trade tional economy. These questions have been the agencies, have been at the forefront in designing focus of ongoing attention from African govern- and implementing the required interventions. Over ments and development partners since the conclu- the years, these agencies have developed a diverse sion of the Uruguay Round and in light of the array of capacity-building tools and expertise with- challenges posed by the broader context of global- in their respective competencies. Bilateral donors ization and liberalization.1 There are two main con- and other development agencies such as the United cerns. The first is institutional capacity building for Nations Development Programme (UNDP), the effective participation in the multilateral trading World Bank, and the IMF have also been involved. system. The second has to do with the supply-side Indeed, interagency cooperation on trade-related constraints of African countries as trading nations. capacity-building interventions and assistance has The first issue arises from identifiable weaknesses proved a major challenge. This is to be expected, on several fronts, including capacity in many given the separate mandates and the different inter- African countries, especially those designated as ests and approaches of donors and agencies, the least-developed countries (LDCs), to formulate and range of initial conditions and needs among coun- 509 T H E T R A D I N G S Y S T E M A N D D E V E L O P I N G C O U N T R I E S tries, and questions of ownership and the setting of of African countries into the global economy and to priorities. their more effective participation in the WTO. The "first-wave" response can be dated to the sec- These goals will require complex interventions, ond half of the 1990s. It focused on three initiatives. demanding a coherent approach from African poli- The first, launched in May 1996 at UNCTAD IX in cymakers, development partners, and other actors. Midrand, South Africa, was the Joint Integrated To the extent that regional markets and regional Technical Assistance Programme for Selected Least- integration constitute a springboard for enhancing Developed and Other African Countries (JITAP). integration in the global economy, there is a need to The second initiative was taken in December that make specific provisions in these interventions to year in Singapore, at the First WTO Ministerial Con- support the development of intra-African trade. ference, which adopted the Comprehensive and Integrated WTO Plan of Action for LDCs. The plan The "First Wave": The JITAP, the IF, envisaged closer cooperation between the WTO and and Assistance with Preparations for other multilateral and bilateral agencies assisting Future Negotiations LDCs in the area of trade. To implement the Plan of Action, a high-level meeting convened by the WTO In April 1994, 30 African countries signed the Mar- in October 1997 launched the Integrated Frame- rakech Agreement that established the WTO, even work for Technical-Related Assistance, Including though only a handful of African countries had Human and Institutional Capacity Building to Sup- been active participants in the Uruguay Round port Least-Developed Countries in their Trade and negotiations and in the multilateral trading frame- Trade-Related Activities (the Integrated Framework, work established by the GATT. By 1998, 10 more or IF). Finally, a separate initiative designed to help African countries had joined the WTO, and 5 oth- African countries prepare for future WTO negotia- ers were at various stages of accession. The 40 tions emerged, encompassing a variety of inter- African members of the WTO constitute nearly a ventions, including UNCTAD's positive agenda pro- third of the entire membership. This surge in the gram. interest of African countries in the multilateral The "second-wave" response, dating from around trading system reflected the fact that several of mid-2000, followed reviews of the experience of the them had undertaken significant liberalization of JITAP and the IF. It apparently has broader objectives their national economies as part of economic and a much wider scope than the initiatives in the first reforms under adjustment programs supported by wave. It is concerned with "mainstreaming trade" as the World Bank and the IMF. It further reflected a an integral part of the overall development and basic perception that with the increasing globaliza- poverty reduction effort. In other words, it is more tion of the world economy, the disciplines of the explicit in its recognition of trade as a major engine of WTO agreements provide a framework for stable enterprise development, diversification, economic and predictable market access and for safeguarding growth, and poverty reduction. Consequently, it is national trading and related interests. focused on assisting the countries concerned to iden- It was clear from the start, however, that partici- tify and prioritize structural supply-side constraints, pation in the WTO framework and in international including insufficient human, institutional, and pro- trade would require building up the necessary ductive capacity and inadequate trade-related infra- capacity. Shortly after the Marrakech meeting, in structure. It is significant that the second-wave October 1994, African trade ministers meeting in response emerged in the context of a renewed effort Tunis adopted a Framework for Action for the by major international financial institutions such as Implementation of the Uruguay Round Agree- the World Bank and the IMF, and by the development ments by African Countries. The framework was community as a whole, to pursue a more inclusive substantially concerned with the identification of policy agenda aimed at addressing entrenched pover- capacity-building needs for the development and ty and the marginalization of the poorest countries in management of trade policy, including the imple- the new global economy. mentation of the Uruguay Round agreements, par- This chapter reviews the main issues encom- ticipation in the WTO framework, and the passed by the first and second waves of responses. promotion of exports. The JITAP, the IF, and UNC- The latter holds the key to a sustainable integration TAD's positive agenda interventions emerged dur- 510 Trade-Related Capacity Building for Enhanced African Participation in the Global Economy ing the second half of the 1990s as responses to interventions also underlies the IF's activities. The these concerns. Each of these responses is briefly Geneva trio, along with the World Bank, the IMF, described below.2 and the UNDP, constitute the six core organizations collaborating in the delivery of trade-related capac- ity-building assistance under the IF. The IF is the The JITAP product of the expressed desire of WTO member The JITAP was established as a collaborative venture states "to foster an integrated approach to assist by the three Geneva trade agencies in cooperation least-developed countries in enhancing their trad- with interested international donors. As has been ing opportunities."4 noted, the Geneva trio had over the years acquired As noted above, the IF was established at the high- substantial expertise in providing technical assis- level meeting organized by the WTO in October tance for various aspects of trade-related capacity 1997 to put into effect the Comprehensive and Inte- building as part of their respective mandates. grated WTO Plan of Action for the Least-Developed Indeed, the JITAP was conceived as a vehicle for uti- Countries adopted in Singapore in December 1996. lizing this expertise by adopting a systematic The main assumption underlying the IF is that each approach and a framework for donor and inter- LDC has a different set of initial conditions and agency coordination. To enhance the sustainability therefore specific trade-related capacity-building of JITAP interventions, much emphasis was placed requirements. Accordingly, it was emphasized right on human resource development, institutional from the start that the interventions envisaged capacity building, and strengthening of export sup- under the IF must be demand-driven to ensure the ply capabilities. relevance and country ownership of the capacity- Eight countries--Benin, Burkina Faso, Côte building process. Each participating country was d'Ivoire, Ghana, Kenya, Tanzania, Tunisia, and therefore required to carry out a needs assessment Uganda--were initially selected for JITAP projects. for trade-related technical assistance. The objectives of the JITAP were put into effect The IF adopted a methodology based on a stan- through a series of interconnected activities aimed dard questionnaire that was designed to help coun- at building national capacity to understand the tries carry out the needs assessment exercise with WTO agreements and their development implica- the assistance of any, some, or all of the six agencies. tions for each beneficiary country, including their Following the completion of the needs assessment implications for trade negotiations; adapting the exercise, the six agencies were to cooperate in national policy and regulatory framework to the preparing a provisional program of trade-related WTO agreements; and enhancing the country's technical assistance that responded to the needs capacity to take advantage of the WTO agreements which had been identified. The provisional pro- through improved export readiness. Although the gram, which became known as the integrated JITAP concept was launched in 1996, it only took response, was to be discussed and agreed on with off two years later, following the establishment of a the LDC concerned. Each of the six agencies would Common Trust Fund to finance program activities then assume responsibility for implementation of and the receipt of pledges to the fund from 13 those aspects of the integrated response that fell donor countries, amounting to US$8.2 million of within its competence and specialization. By mid- the estimated funding needs of US$10.3 million 1999, 40 of the 48 LDCs had completed the needs over four years. The fund is managed by the ITC assessment exercise.5 and is supervised by a steering group of representa- To facilitate implementation of the integrated tives of donors, beneficiaries, and the secretariats of response, it was further expected that the exercise the ITC, UNCTAD, and the WTO. In 1999 the three would culminate in the scheduling of a trade sector agencies implementing the JITAP delivered just roundtable meeting with donors, typically in the under US$3 million of activities.3 context of a World Bank Consultative Group Meet- ing or a UNDP Roundtable Meeting. The purpose of these meetings was to give development partners The IF an opportunity to endorse a multiyear program of The JITAP concept of interagency coordination in trade-related technical assistance and to pledge sup- the delivery of trade-related capacity-building port for elements of the program. By mid-1999, 511 T H E T R A D I N G S Y S T E M A N D D E V E L O P I N G C O U N T R I E S only Uganda had been able to organize such a meet- demand for trade-related capacity-building was ing, although more than 20 countries had expressed designed to respond to the specific conditions of a an interest to the WTO in doing so.6 country--thereby avoiding "one size fits all" solu- tions--and to facilitate preparation for future nego- tiations. It can be argued that this effort has resulted Positive Agenda Initiatives in greater sensitization in African countries as The Singapore ministerial conference revealed seri- regards the requirements for compliance with WTO ous flaws in WTO decisionmaking processes. Basi- membership and for participation in rules-based cally, an inner circle of only 34 of the WTO's then multilateral trade decisionmaking, including nego- 128 members had taken responsibility for negotiat- tiations. The JITAP and the IF, however, suffered ing an agreed text on sensitive issues--including from serious deficiencies. The JITAP was limited to textiles and clothing, labor standards, investment, just a handful of countries and was not able to and competition policy--that remained to be final- deliver projects aimed at enhancing competitiveness ized for inclusion in the ministerial declaration to and overcoming supply-side constraints that the be issued at the end of the conference (see Black- beneficiary countries faced, including inadequate hurst 2001). Few African and LDC representatives investment in production and infrastructure. The IF were part of this inner circle. In addition, African never really took off, although it was generally and LDC representatives had little input in the acknowledged that the needs assessment exercise negotiations that were concluded during the confer- had enabled both governments and development ence to eliminate tariffs on information technology partners to carry on a dialogue on trade policy products and on a number of pharmaceutical prod- issues and priorities and to engage in serious reflec- ucts. Because they lack export capacity in these sec- tion on overcoming the constraints on interagency tors, the WTO's LDC members and a substantial cooperation. Issues related to regional trade policy majority of the African delegations at the confer- were mostly ignored. This is quite a curiosity, as the ence were effectively marginalized during these mid- to late 1990s was also a period during which negotiations. important strides were made toward regional inte- The experience at Singapore raised the question gration, especially in eastern and southern Africa as to how developing countries in general could take and in francophone West Africa. issues of interest to them forward in future WTO By contrast, by the time of the Seattle ministerial negotiations. It was in response to this question that conference in late 1999, the fruits of the positive the positive agenda initiative emerged, principally agenda initiative were clearly in evidence. During under the leadership of UNCTAD acting in concert the preparatory process leading up to the Seattle with other intergovernmental organizations and conference, developing countries submitted well regional organizations such as UNDP, the South over 100 negotiating proposals, more than half the Centre, the Commonwealth Secretariat, and the total (see UNCTAD 2000b: vii). African countries in Organization of African Unity (OAU). The objec- particular exhibited an unprecedented degree of tive of the initiative was to step up research and preparedness and greater awareness of the issues at analysis aimed at assisting the countries concerned stake (see Luke 2000; Oyejide and Njinkeu 2000). to develop a positive agenda for future WTO negoti- This was the result of intensive preparatory events, ations, including the negotiations on agriculture as well as serious efforts to formulate an Africa-spe- and services due to be launched at the Third WTO cific positive agenda, thanks to UNCTAD's technical Ministerial Conference, which was held in Seattle in cooperation activities in this area and to policy late 1999. research capacity-building programs supported by donor agencies. In addition, the difficulties in implementation experienced by all African mem- Assessment bers of the WTO had increased their awareness of The first-wave response (including the regular the impact on their economies of rules emanating trade-related technical assistance activities of the from multilateral trade negotiations. When the Geneva trio and other bilateral and multilateral Seattle conference reverted to Singapore-style agencies outside the framework of the JITAP and inner-circle decisionmaking, African delegations the IF) in the second half of the 1990s to the were at the forefront in denouncing this approach. 512 Trade-Related Capacity Building for Enhanced African Participation in the Global Economy This was one of the factors leading to the break- For IF-eligible countries, the centerpiece of the down of consensus and the eventual failure of the IF-II arrangements is that trade-related technical conference. assistance and associated programs and projects are to be carried out through a country-led process of defining national poverty reduction strategies. This The "Second Wave": Mainstreaming Trade would be ensured principally through such instru- During 2000, mandated reviews of the functioning ments as the national poverty reduction strategy of both the JITAP and the IF were undertaken. papers (PRSPs) and the United Nations Develop- These reviews occurred at a time of renewed effort ment Assistance Framework (UNDAF), which are at the World Bank and the IMF, and in the wider to provide the basis for a program of assistance development community, to address the trade con- agreed on with development partners. This main- straints faced by the poorest countries in a more streaming effort is to be led and coordinated by the comprehensive manner while taking key factors World Bank, according to the principles of its such as ownership, sustainability, market failure, Comprehensive Development Framework, with and institutional resistance to donor coordination participation and input from other core agencies into account (see, for example, Stiglitz 1998; World and other stakeholders. Building on initial needs Bank 1998; Kaul, Grunberg, and Stern 1999). It is assessments and subsequent work, the IF-II therefore not surprising that these were the themes approach would involve formulating country-spe- that provided the subtext for the JITAP and IF cific trade programs as part of the broader poverty reviews. reduction strategies. These activities are expected The overall message from these reviews was the to feed into the World Bank Consultative Groups need to "mainstream" trade as an integral part of and UNDP Roundtable Meetings, where countries the overall national development and poverty will present their medium-term policy frameworks reduction effort (see Rajapathirana, Lusthaus, and and financing needs, including needs for trade- Adrien 2000).7 Consequently, the second-wave related assistance, for support by the donor com- response lays considerable stress on ensuring that munity. trade policy, trade-related technical assistance, and Another change was the decision to seek donor capacity-building needs are articulated in a broad support for and voluntary contributions to an IF development context, not addressed in isolation. Trust Fund (IFTF). The IFTF, which would involve about US$20 million over a three-year period, would be primarily dedicated to helping LDCs Next Steps: "JITAP II" and "IF II" develop the necessary analytical and policy frame- In addition to examining ways of streamlining the work for mainstreaming trade into national devel- JITAP's management process, the midterm evalua- opment strategies and for developing programs and tion carried out during 2000 emphasized the need to projects.9 It was subsequently agreed that imple- strengthen the role of ministries of trade as the focal mentation of IF II would proceed on the basis of a point in the development of trade policy, including pilot scheme to assist countries that demonstrated a the provision of extension services to the private sec- clear choice and commitment to mainstream trade tor and engagement with the WTO. Further empha- as part of their country development strategies as sis was to be placed on building a network of trainers expressed through PRSPs or the UNDAF.10 through the involvement of local universities and Thus, under the new IF-II arrangements, the need business schools. It was also envisaged that the for trade-related assistance was to be assessed along- remodeled JITAP would be extended to an additional side a country's other priorities and supported 10 to 15 African countries, given the demand for its accordingly by the government concerned and the accelerated and integrated mode of delivery of trade- donor community. This approach was expected not related capacity-building interventions. Whereas only to ensure that trade takes its rightful place in regional integration considerations were mostly policy terms but also to create a viable framework ignored under the original JITAP (JITAP I), it was for making available the resources required to foster proposed that one of the criteria for selecting the the necessary skills, institutions, and infrastructure additional countries to be included in JITAP II would for the effective integration of the LDCs into the be their role in regional processes.8 world economy. 513 T H E T R A D I N G S Y S T E M A N D D E V E L O P I N G C O U N T R I E S Assessment tion, on the other. If IF II takes off as expected, the proposed trade chapter of the PRSPs would include It is too early to assess the second-wave response, the identification and prioritization of trade-related since implementation has hardly begun. A few posi- capacity requirements, from infrastructure to human tive elements, however, that have emerged from the resources, within a coherent policy framework. reviews of the JITAP and the IF could potentially It has been suggested that, ultimately, main- bridge the major gaps evident in the first-wave streaming trade will give greater visibility to the response. linkages between trade and all other related policy areas, including health, education, and general JITAP II. The effort that is to be made in JITAP II to social conditions (see Fried 2000). This would strengthen the role of ministries of trade as the focal require governments to reflect on how they can point for trade policy is to be welcomed. The expe- most efficiently use the limited resources they are rience of several developing countries in East Asia, able to devote to trade. In this regard, it is to be Latin America, and elsewhere has shown the need hoped that in the reassessment of the use of for such a corporate framework to manage the trade resources, such constraints as the understaffing of policy process; to oversee policy issues concerned the WTO missions of African countries in Geneva with multilateral and regional trade agreements, and cases of complete nonrepresentation would be including compliance and negotiation; and to facil- resolved in a decisive manner (see Blackhurst, itate coordination with other institutions concerned Lyakurwa, and Oyejide 2000). By the same token, it with national economic management, with a view would require development partners to reexamine to ensuring that supply-side constraints are ade- development assistance priorities to ensure that quately addressed. they are sending a coherent message across their In this regard, human resources development and various assistance mechanisms and institutions. training for trade-supporting services are essential. The World Bank, in particular, as the lead agency The JITAP II proposal for a stronger involvement of in the mainstreaming exercise, faces a major chal- local universities and business schools to comple- lenge to ensure that it becomes more proactive in ment other activities aimed at strengthening the engaging national trade policymakers to deter- network of trainers, if acted on, will be a significant mine the nature and extent of trade-related techni- contribution. There is considerable scope for inter- cal assistance and capacity building needed by a country cooperation in this area in promoting high- country. quality regional centers of excellence to fulfill training, advisory, analytical, and research func- Conclusion tions. These could be developed from within exist- ing institutions. This chapter has revisited the question of trade- JITAP II further proposes to address relevant related capacity building for enhanced African par- regional integration elements by including as crite- ticipation in the international economy. It has ria for the extension of the program to other coun- suggested that the first-wave response to this ques- tries such factors as the role of a country with tion during the second half of the 1990s led to respect to regional integration; the country's poten- greater sensitization in African countries on inter- tial to benefit from the program; proximity to national trade issues, including participation in the regional clusters; and possibilities for regional syn- WTO. The JITAP and the IF, however--as the main ergies and economies of scale at the implementa- instruments of capacity-building interventions-- tion stage. were constrained by serious deficiencies. The JITAP was limited to just a few countries and was not able IF II. The IF was originally an unfunded mandate, to deliver projects to enhance competitiveness and and donors and agencies differed in the priority to be overcome constraints on the supply side, including given it. Situating it at the center of a beneficiary those in production and infrastructure. The IF itself country's program of assistance with development never actually took off. Issues related to intraregion- partners provides a more solid basis for establishing al trade policy were mostly ignored. the link between trade and development, on the one Still, by the time of the Seattle ministerial confer- hand, and development strategy and poverty reduc- ence in 1999, there was much evidence that the 514 Trade-Related Capacity Building for Enhanced African Participation in the Global Economy effort given to technical assistance to prepare Notes African countries for trade negotiations had paid An earlier version of this paper was presented at the African Eco- off. In Seattle African countries exhibited an nomic Research Consortium (AERC) Seminar on Assistance in the unprecedented degree of preparedness and greater Preparation of African Countries for the WTO Trade Negotiations, awareness of the issues at stake. At the next ministe- Geneva, March 9, 2001. The author may be reached at rial meeting, in Doha, they played an active role. . Technical assistance for trade negotiations is an 1 See, for example, the Framework for Action for the Implemen- ongoing activity that will call for improved coordi- tation of the Uruguay Round Agreements by African Countries nation among the various partners involved. adopted at a meeting of African trade ministers in Tunis, Octo- Mandated reviews of the JITAP and the IF during ber 1994 (available from the Organization of African Unity), and OECD (1999b). 2000 resulted in a second-wave response concerned with mainstreaming trade as an integral part of the 2 It should be noted that the regular trade-related technical overall development and poverty reduction effort. cooperation activities of the Geneva agencies and of other bilateral and multilateral donors continued outside the frame- For the JITAP, specifically, the review resulted in a work of the three initiatives. commitment to extend JITAP II to an additional 10 to 15 countries, given the demand for its accelerated 3 See, for example, UNCTAD, "Review of Technical Co-operation Activities," TD/B/47/2/Add.1, 2000, and ITC, "JITAP Mid-term and integrated mode of trade-related capacity- Evaluation, Management Response," November 16, 2000. building interventions. It was further proposed that one of the criteria for selecting the additional coun- 4 See WTO, "Singapore Ministerial Declaration," WT/Min tries should be their role in regional integration (96)/DEC, 1996, para. 14. processes. For the IF, the centerpiece of the new 5 See WTO, "High Level Meeting on Integrated Initiatives for arrangements is to place trade-related technical Least-Developed Countries, Trade Development, Outcome and assistance within the framework of a country's Follow-up, Report of the Director-General," WT/MIN (98)/2, 1998, and WTO, Sub-Committee on Least-Developed Coun- national PRSP process. This is expected not only to tries, "Note on the Meeting of 12 July 1999," ensure that trade takes its rightful place in policy WT/COMTD/LDC/M/16, 1999. terms but also to create a viable framework for mak- 6 See WT/COMTD/LDC/M/16, 1999. ing available the resources required to foster the necessary skills, institutions, and infrastructure for 7 See also WTO, "Review of the Integrated Framework: Commu- the effective integration of IF-eligible countries into niqué from Heads of the Six Core Agencies," WT/LDC/SWG/IF/2, July 12, 2000: 2. the world economy. IF II will require complex inter- ventions, demanding a coherent approach from 8 See ITC, "JITAP Mid-term Evaluation, Management Response," African policymakers, development partners, and November 16, 2000. other actors. 9 See WT/LDC/SWG/IF/2, July 12, 2000. 10 See WTO, Sub-Committee on Least-Developed Countries, "Integrated Framework--Proposal for a Pilot Scheme," WT/LDC/SWG/IF/13, February 16, 2001. 515 VIII T R A D E P O L I C Y Q U E S T I O N S A N D G U I D E L I N E S T he chapters in this part are more norma- Michalopoulos, Maurice Schiff, and David Tarr, out- tive than most of the others in this lines these general rules. Handbook in that they offer rules of thumb for poli- One area of policy advice that is particularly con- cy. These policy recommendations are not abstract troversial--despite many years of research and cross- or absolute; they are based on extensive country country experience--is the design of the tariff experience, and in many instances they do not so structure. Chapter 52, by David G. Tarr, on argu- much give formulas as point toward asking specific ments for and against uniformity in the structure of questions and determining the situation that applies the tariff, comes out strongly in favor of a single tar- in a given context. As emphasized in other parts of iff rate for all goods. Although not all may agree with the Handbook, trade reforms must be accompanied this recommendation, it is important to understand by many other, complementary, actions. What the arguments as to why uniformity makes for good these are will depend on the country concerned. economic policy. It is also important to note that uni- Although each country must determine its own pri- formity does not necessarily mean free trade. A pref- orities, taking into account the prevailing institu- erence for uniformity has implications for the pursuit tional capacity constraints, the rules of thumb that of multilateral tariff negotiations, which can be used are proposed for trade policies are quite robust in both to reduce the dispersion in tariffs and to reduce the sense that they help guide policymakers away the average level of tariffs. Countries that have uni- from narrow and piecemeal approaches that are form tariffs cannot play the WTO game of trading likely to be inefficient. Chapter 51, by Constantine concessions on tariffs for specific goods; they must 517 T R A D E P O L I C I E S Q U E S T I O N S A N D G U I D E L I N E S use a formula approach to lower tariffs. The use of 2001). The CD-ROM on Applied Trade Policy that formulas in WTO negotiations is discussed by Arvind accompanies this volume contains numerous rele- Panagariya in Chapter 53. vant readings. Mechanisms to increase the credibility of trade On modalities for negotiation, see Bernard Hoek- reform can have a powerful effect on the invest- man and Michel Kostecki, The Political Economy of ment supply response. Mexico after WTO accession the World Trading System: The WTO and Beyond, 2d and entry into the North American Free Trade ed. (Oxford, U.K.: Oxford University Press, 2001). Agreement is an example; another is Portugal, The contributions by Robert Baldwin and by Alan which experienced investment booms after its Winters in J. Michael Finger and Andrzej Olechowski accession to the European Union. A major benefit of (eds.), The Uruguay Round: A Handbook for the Multi- WTO membership is that it allows countries to lateral Trade Negotiations (Washington D.C.: World "bind" the level of their tariffs, thus reducing uncer- Bank, 1987), provide succinct treatments of GATT tainty for investors, as described by Joseph F. Fran- negotiating techniques and principles. Constantine cois and Will Martin in Chapter 54. Michalopoulos, Developing Countries and the WTO Regional agreements can have a similar function; (New York: Palgrave, 2001), discusses trade policy indeed, regional integration agreements (RIAs) offer issues and options for future reform of the WTO. a potential avenue for countries to pursue many of Analyses of how international agreements may the behind-the-border reforms that are necessary for serve as mechanisms for governments to reduce improving the investment climate. Such agree- investor uncertainty and risk premiums by making ments, however, can also easily be detrimental for irrevocable policy commitments can be found in longer-term economic growth prospects because Joseph Francois, "External Bindings and the Credi- they can give rise to costly trade diversion. How to bility of Reform," in A. Galal and B. Hoekman (eds.), benefit from regional integration is the subject of Regional Partners in Global Markets: Limits and Possi- Chapter 55, by Bernard Hoekman and Maurice bilities of the Euro-Med Agreements (London: Centre Schiff. In practice, when it comes to day-to-day poli- for Economic Policy Research, 1997), and in Raquel cies, regionalism is as important as the WTO for Fernandez and Jonathan Portes, "Returns to Region- many developing countries, if not more so, as medi- alism: An Analysis of the Nontraditional Gains from um-term strategy is increasingly developed and Regional Trade Agreements," World Bank Economic implemented in the context of regional agree- Review 12 (1998): 197­220. ments. Getting these agreements right in the sense The economics of regional integration are of maximizing their potential to foster economic assessed in Jagdish Bhagwati and Arvind Panagariya growth is therefore extremely important. Well- (eds.), The Economics of Preferential Trade Agree- designed regional agreements can complement the ments (Washington D.C.: AEI Press, 1996). A survey WTO--and membership and participation in the of the theoretical literature may be found in Richard WTO are critical in order to minimize the potential Baldwin and Anthony Venables, "International Eco- downsides of regionalism. nomic Integration," in Gene M. Grossman and Ken- neth Rogoff (eds.), Handbook of International Economics, vol. 3 (Amsterdam: North-Holland, Further Reading 1997). L. Alan Winters, "Regionalism Versus Multi- There is a huge literature on the design of trade lib- lateralism," in Richard Baldwin, Daniel Cohen, eralization policy programs and the need for ancil- André Sapir, and Anthony Venables (eds.), Market lary reforms. An accessible review of the literature Integration, Regionalism and the Global Economy and country experience is John Nash and Wendy (London: Centre for Economic Policy Research, Takacs (eds.), Trade Policy Reform: Lessons and Impli- 1998), is an in-depth discussion of the incentive cations (Washington D.C.: World Bank, 1998). See effects created by RIAs as regards the stance taken also Neil McCulloch, Alan Winters, and Xavier Cir- by members in the WTO. World Bank, Trade Blocs era, Trade Liberalization and Poverty: A Handbook (Washington DC: World Bank, 2000c), assesses the (London: Centre for Economic Policy Research, impact of RIAs on developing countries. 518 51 C O N S TA N T I N E M I C H A L O P O U L O S M A U R I C E S C H I F F D A V I D G . TA R R Rules of Thumb regime and provide guidance for the direction of reforms. for Trade Policy · No licensing or other approvals, except for health, safety, and environmental rea- sons and automatic licensing used for statistical purposes; no other quantitative restric- tions. · Low and uniform tariffs. If the tariff is not uniform, it should show little dispersion, with T only a small number of bands. he effect of trade reform on In particular, having a few sectors with very high growth depends on a variety of tariffs should be avoided (for more on this, see complementary policies and institutions. In low- Chapter 52, by Tarr). income countries the key complementary policies · If tariffs are important for revenue generation, and institutions that need to be analyzed fall into uniformity implies that the overall level of the the following general areas: (a) macroeconomic, tariff should be such as to generate the revenue and especially exchange rate, policy; (b) the opera- required. However, some items--such as alcohol tion of the market for labor, since the poor are often and tobacco products--may be subjected to high concentrated in the informal sector; (c) the opera- duties to raise revenue, as long as equivalent tion of the markets for agriculture, which is a major excise taxes are imposed on domestic production. source of income and accounts for a large portion of · An efficient customs clearance process, with little the household expenditures of the poor; (d) access red tape, that ensures tariff-free access to interme- of the poor to trade-related services such as credit, diate imports for exporters. marketing, and transport; and (e) access to safety · Only one instrument of contingent protection--a nets. There are, of course, other issues, such as gov- safeguard provision. No antidumping measures. ernance, that are important here as well as in other · Contestable service markets: measures to ensure reform efforts. What follows is a checklist of ques- that competition prevails, that there is no dis- tions and issues that can be considered in the design crimination against foreign suppliers seeking to and pursuit of trade reform. establish a presence in the market, and that The basic elements of a good trade policy regime appropriate regulation is in place. are predictability, transparency, and uniformity. The following list describes succinctly what constitutes a A number of questions are relevant in determin- liberal trade policy regime. They provide a bench- ing whether trade reform should be a priority from mark against which to judge the prevailing trade a poverty-reducing growth perspective: 519 T R A D E P O L I C I E S Q U E S T I O N S A N D G U I D E L I N E S What is the impact of status quo trade policies on How efficient is the customs service? How long does the poor? It is important to determine the effect on it take to clear a container or air freight shipment? the poor of the existing pattern of protection and How does this compare with the norms in neigh- subsidization. Such effects may be positive or nega- boring countries and with best practice? How large tive and may affect particular products consumed are "unofficial" trade facilitation payments? Is there by the poor or the incomes of a significant number a functioning drawback and temporary admission of the poor, throughout the country or in a particu- mechanism? lar region. Taxes or supports for important food sta- The overall analysis of the trade regime should ples or inputs to agriculture, in particular, should be yield a preliminary judgment on the desirability of identified and their incidence examined. In those trade reform. This judgment should then be cases where the structure of protection is not bene- reviewed in light of the potential short-term effects ficial to the poor, there is a prima facie case for of trade reform on the poor and on other groups in reform. In cases where some of the poor benefit, an society that are likely to experience serious adverse assessment should be made of the relative magni- effects. If there are possible negative effects, it is tude of their potential losses and the economywide important to identify the relevant products and sec- gains from reform. tors early on, to make it possible to design arrange- Are there nontariff barriers for reasons other than for ments for dealing with the adverse impacts of the health, safety, and the environment? To the extent that reform and to develop strategies for building con- significant nontariff barriers are present, there is sensus in support of reform. It must be recognized again a prima facie case for reform, starting with con- that there will always be groups that lose from trade version to tariffs. This reform is likely to benefit the reform; by definition, trade policy is an instrument poor more than the nonpoor, since license recipients through which income gets redistributed in the typically collect rents and (almost by definition) are economy. In many cases those that lose will not be unlikely to be poor, and the competition for licenses the poor but those who benefit from the scarcity wastes resources that could be used productively. rents that are created by restricting trade. What is the average tariff, and how dispersed is it? Strong government commitment to the reform is The more the dispersion, the greater the difference critical. The government should attempt to explain in treatment of different sectors and segments of the desirability of the reform and obtain support society is likely to be, and the greater the urgency of from some parts of civil society. Achieving a broad reform. Dispersion, which is often generated by consensus may be a difficult task. The benefits from exemptions and tariff escalation, will lead to high reform are likely to be dispersed, uncertain, and effective rates of protection and is likely to entail spread over time, whereas the private costs of sec- significant inefficiencies. tors that will be facing increased competition from Is there discrimination against agriculture? The imports will be obvious, near term, and likely to be overall policy stance affecting agriculture should be concentrated in powerful political groups. determined, starting with an assessment of the It might appear tempting to design a trade reform effective rate of protection for this sector compared by identifying sectors that are important to the with manufacturing. (See Schiff and Valdés 1992 for poor--either on the consumption side or on the a description of a methodology for doing this.) income side--and singling out these sectors for dif- Agriculture is of great importance for poverty ferentiated cuts in protection. If, for instance, many reduction because the rural poor are likely to poor people produce maize, as in Mexico, it might account for a large share of the country's poor. seem sensible to exclude this product from a tariff How well do critical service markets function? Do reduction. There are at least two problems with this the poor have access to important ancillary services approach; one is fundamental, and the other has to such as transport? Do policies discriminate against do with political economy. The fundamental prob- foreign suppliers and lead to high-cost, low-quality lem is that trade policy is a single instrument, and a domestic supply? Is entry possible in labor-intensive basic principle of economic policy formulation is sectors such as tourism and back-office services? that a single instrument cannot be expected to Does competition prevail in key backbone sectors address multiple targets. The political-economy such as transport, finance, and communications? Is problem is that once a highly differentiated trade appropriate procompetitive regulation in place? regime is adopted, it is essentially impossible to stop 520 Rules of Thumb for Trade Policy special interests from building a case that their sec- fied. Questions that should be posed include the fol- tor deserves special treatment for one reason or lowing: another. Returning to the example of maize, if we decide to maintain or raise protection, we are likely · In agriculture, are prices passed on to farmers, or to find that there is another important group of are there government or private intermediaries poor people for whom maize is an important that make large profits on the sale of farm prod- expenditure item. ucts or farm inputs? A better approach is to concentrate on developing · More generally, is the reform likely to destroy two different sets of instruments--one (trade poli- existing markets that are significant for the poor? cy) focused on providing the incentives appropriate Will it make it possible for poor consumers to for the efficient production and use of goods and obtain new goods? (For further discussion, see services, and another (distributional policy) Winters 2000a; McCulloch, Winters, and Cirera focused on alleviating poverty. With this assign- 2001.) ment of instruments, trade policy can be designed · Are there serious impediments of various kinds using the simple, comprehensible guidelines for (legal, cultural, and infrastructural) to labor mobil- trade policy formulation. A set of distributional ity? Labor market restrictions, such as prohibitions instruments will necessarily have a much wider on firing workers, often spawn an informal labor range of dimensions, including expansion of access market sector, with the poor concentrated there. to education, the provision of safety nets, and Reduction in the restraints in the labor market, investments in the infrastructure needed to provide especially combined with trade reform, can result people in poorer regions with access to the markets in an expansion of the formal sector and an and other amenities enjoyed by relatively advan- increased demand for labor in that sector. This can taged people. have a strong impact on poverty reduction, since it will allow the poor to move out of the informal sector and into the formal sector. Complementary Policies · Are there serious financing obstacles to participa- As has been emphasized in other chapters in this tion in trade? In the long run developing an effec- Handbook, trade policy reform and institutional tive financial system is a key to development. In strengthening must be implemented in the context of the short run trade-focused instruments such as a variety of complementary policies. Some of these the use of back-to-back letters of credit may alle- are general, and some are focused on making the viate some of the most pressing obstacles to trade. trade policy reform more likely to benefit the poor. · Are there serious obstacles to setting up a busi- ness? Competition may be impeded because it is difficult to obtain a license to start a new business Macroeconomic Policy or to make an investment. Or foreign investment Macroeconomic stabilization and a competitive may be impeded. The reduction in barriers to exchange rate are essential for supporting greater entry, especially those imposed by governments at integration into world markets. Exchange rate various levels, can be expected to improve compe- depreciation at the outset of major trade reform tition and allow entrepreneurs to sell new, cheap- programs will facilitate adjustment. Great care must er imports to the poor or to provide services be taken to avoid real exchange rate appreciation at necessary to bring goods to market. The latter is a time of import liberalization. As appreciation is particularly important in cases in which reform largely determined by macroeconomic and fiscal affects entities that provide ancillary services to policies, it is important that these are managed so as the poor and that may cease to do so after reform to be consistent with the trade reform. occurs. An example would be the provision of transport, storage, and distribution services to farmers by a marketing board. If there are barriers Markets to entry into such services activities, or if entry is If markets are not competitive or are missing, trade unlikely to occur because the market cannot sus- reforms may not benefit the poor. Critical obstacles tain operations, continued government involve- to the operation of market signals must be identi- ment may be necessary. 521 T R A D E P O L I C I E S Q U E S T I O N S A N D G U I D E L I N E S · Are there transport obstacles to trade? High effective institutions--takes both money and time. transport costs make it difficult to engage in International assistance is the appropriate response trade. Governments of landlocked economies in many cases. that are far from markets may have few variables that they can influence, owing to the absence of Timing and Sequencing direct links and the difficulties of establishing transit arrangements, but sometimes public poli- Existing interest groups--often, entrenched elites cies or institutions can be modified in a way that benefiting from the status quo--may oppose would lower transport costs. Here again, a key test reform, and if reform is agreed on, there will be should be whether entry is feasible. pressures to postpone adjustment as long as possi- · Are there cases in which analysis indicates that the ble. While recognizing that this is the likely environ- market will not supply the needs of the poor or ment in which trade reforms are usually proposed, those located in outlying regions? If so, there may it is nonetheless extremely important to investigate be a need for universal services regulation. Expe- in advance the impact of reforms on specific groups rience suggests that explicit subsidies to achieve of the poor and to design programs to address the such objectives can be efficient. effects. The timing of the implementation of trade · Are there serious trade obstacles to entering reforms needs to be closely linked to the establish- major external markets or competing abroad? ment of programs that deal with the impact of Most low-income developing countries face rela- reforms on the poor. Some points about sequencing tively low traditional trade barriers in external are worth noting: markets, given a variety of preference schemes, but in many markets there are specific problems · If a reform is preannounced to be implemented related to sanitary and phytosanitary controls, the over several years and it is a credible reform, nor- threat of contingent protection, and competition mal market adjustment and attrition can be used from subsidized exports in third markets. to eliminate or greatly reduce adjustment costs. This may, however, come at the cost of the threat of reversal of the trade reforms, as entrenched Trade-Related Institutions interests will be granted time to mobilize opposi- The success of trade policy reforms involves a vari- tion. A staged reform that is scheduled to take ety of institutions, both public and private. On the more than five years is not likely to be credible government side, an effective and noncorrupt cus- unless it is anchored in WTO commitments or a toms authority is critical to the success of reform. far-reaching regional trade agreement. Other institutions to which particular attention · Nontariff barriers and high tariff peaks should be needs to be paid to ensure that trade reforms bene- addressed earlier rather than later. fit the poor include those in marketing and export · Tariffs should be reduced across the board during finance, both of which are necessary for export each stage of a gradual reform. If, instead, a target expansion. In order for the poor to benefit, it may is set on the basis of the tariff average, the tenden- be useful to establish organizations such as cooper- cy will be to cut tariffs only where they cause no atives that can put together large enough shipments immediate difficulty and to leave all the adjust- from individual producers to supply foreign mar- ment to last. kets and that are able to obtain financing linked to · Broad trade reforms frequently meet with much their exports--which individual poor farmers can- less political resistance than cuts in protection to not. The extensive international experience on individual sectors. Broad reforms help the win- these issues can be accessed through the Interna- ners from reform recognize their potential gains tional Trade Centre and UNCTAD in Geneva. and tend to reduce the costs even for industries Bilateral donors and multilateral development that lose protection on their output. banks can provide assistance in the design of such · Waiting for the completion of some important programs, as well as in financing. Overcoming infrastructure project, such as a port facility or a some of the constraints discussed above--for road, is not usually a good reason to delay example, inadequate infrastructure and lack of reforms. 522 Rules of Thumb for Trade Policy Coping with Uncertainty perfect at the start, but some minimum conditions, especially macroeconomic stability and a competi- Because the success of any reform depends on so tive exchange rate, should be in place. Also impor- many other complementary policies and institu- tant are mechanisms to increase the credibility of tions, it is bound to be uncertain. One needs to be reform. This is an important area in which well- especially careful regarding the effects of any reform designed multilateral and regional trade agreements on the poor, as they are least able to bear risks. They can play a beneficial role (see Box 51.1; see also may be unwilling to take risks to increase their Chapter 55, by Hoekman and Schiff, in this volume). income if by doing so they also increase their It must be recognized that the poor are least able chances of incurring losses--which would have dire to bear risks, and that in the short run there will be consequences, since the poor do not have the losses for some groups. Analysis of the status quo resources to carry them over a bad spell. The issues and of the likely impact of reform on the poor is that arise in this connection are discussed at greater therefore very important. In many low-income length in Chapter 5, by Winters, in this volume. countries, general safety nets may not exist or may be inadequate. In such situations reforms should not be postponed but should be implemented gradually, Conclusion following a preannounced schedule, and comple- Critical ingredients in trade reform are the necessary mented by actions to minimize adverse conse- complementary policies and institutional develop- quences to the poorest in society. This can be done, ment. The former include the removal of obstacles in many cases, by directly targeting trade policies to labor mobility, the introduction of competitive that are currently clearly detrimental to the interests exchange rates, and the reduction of barriers to of the poor and by taking care that the reform entry and competition. Institutional issues include process considers the need for action in ancillary the operations of the customs authority and country areas to improve the functioning of services markets marketing and distribution. Not everything will be and ensure universal supply of essential services. B O X 5 1 . 1 E X P O R T D I V E R S I F I C AT I O N I N A F R I C A : E L E M E N T S O F A " B U S I N E S S P L A N " A recent joint research effort involving African restrictions. It can learn from Chile's success in economists and international organizations iden- influencing both the type and the size of pri- tified six policy actions as the most pressing from vate capital inflows. an export development and diversification per- 2.Make trade reforms credible and effective. spective. Taken together, they illustrate the type Although considerable trade liberalization has of approach that is consistent with the rules of been achieved, an unfinished agenda remains. thumb proposed in this chapter. African trade taxes and restrictions are still higher than in most other developing regions, 1.Anchor export orientation on competitive and sta- and the antiexport bias is considerable. Given ble real exchange rates. International evidence the continued importance of trade taxes for suggests that the real exchange rate is even government revenue, further liberalization will more powerful for export growth than trade not be possible in many cases without major policy. Recently, some countries have seen a fiscal reforms. In the interim, there is no choice surge in capital inflows, mainly driven by high but to focus on getting compensatory mecha- real interest rates, which has led to misaligned nisms--export-processing zones, duty draw- or unstable real exchange rates. Given Africa's backs, exemption schemes, and value-added need for higher private capital flows, it cannot tax rebates--to work efficiently. Key services afford to reimpose sweeping capital account such as customs also require urgent attention. (continued) 523 T R A D E P O L I C I E S Q U E S T I O N S A N D G U I D E L I N E S B O X 5 1 . 1 ( C O N T I N U E D ) Liberalization often is not yet locked in, is sub- flights in other regions. International telephone ject to reversal, and is often seen as donor driv- charges and Internet connections are among en, without strong national ownership. The the world's costliest. Ugandan firms lose an consequence is that private agents remain hes- average of 91 days a year because of power itant to invest. WTO tariff bindings are general- outages. ly much higher than actual rates, so there is 4.Mainstream regionalism in a new way. Despite room for much greater use of this commitment past failures and the lackluster implementation mechanism. Reciprocal free trade agreements of existing schemes, the case for Africa's eco- with countries outside Africa and harmoniza- nomic integration remains compelling. Alterna- tion of trade and investment policies along tive approaches are needed, however. One is to subregional lines would also help. stress an outward orientation, or open region- 3.Integrate further trade reforms with national alism, and a flexible design for jointly imple- "business plans" for economic diversification. menting specific projects. Since African Trade reforms need to be accompanied by economies are very small, both individually and measures that lay a stable base for investment as subgroups, the potential welfare gains from and production: effective and honest tax and freer trade within Africa may be limited. Creat- customs administrations, functioning commer- ing an economic space in which investors can cial courts, reliable infrastructure, and a work- produce for regional as well as global markets ing financial system. Failure to effect these may provide better growth opportunities than measures has blunted the investment response simply removing barriers to trade among to first-generation trade reforms in most African countries themselves. This approach African countries. With Africa ranked among may serve as a building block for eventual mar- the world's riskiest places to do business, even ket integration while being less political. retaining domestic savings becomes a chal- 5.Create a platform for effective African participation lenge. In 1990 flight capital accounted for in multilateral forums such as the WTO. Such almost 40 percent of private wealth in Africa, forums are essential for underpinning the credi- compared with 6 percent in East Asia and 10 bility of reforms and for enforcing "appropri- percent in Latin America. Investment booms ate" global standards, which are becoming everywhere have been led by domestic capital, prerequisites for accessing markets in industrial so the first order of business is for governments countries. But African countries also need to to provide a safe and profitable environment help shape these standards and to negotiate for that will persuade their citizens to invest at the dismantling of restrictive trade practices home. While some countries, such as Mauritius that inhibit export diversification. Agriculture, and Uganda, have steadily improved their risk processed goods, and textiles and clothing are ratings, others, including Kenya and Zimbab- particularly important. In order to be effective at we, have seen sharp declines. the WTO, it will be essential to pool expertise. The cost of doing business in Africa is high 6.Base all this on consultative processes. Any busi- partly because of its economic sparseness and ness plan requires a supportive, mutually the distance of much of its production from the accountable relationship among business, sea, but weak business services are also a major labor, and government. Much of East Asia's impediment. An efficient Nacala rail line and success has been attributed to active interac- port could save Malawi the equivalent of 3 per- tions between the state and business. Poor cent of its GDP. Road transport in Africa may be communication, and outright mistrust, twice as costly as in Asia, in part because of between these groups limit growth in many unofficial tolls. Restrictive agreements help African countries. Consultative groups should explain why air transport tariffs and handling begin by focusing on regulations that affect the charges can be twice those for comparable entire private sector, but they can also serve as 524 Rules of Thumb for Trade Policy B O X 5 1 . 1 ( C O N T I N U E D ) agents of restraint on government behavior, to needed in the new era of globalization, democ- ensure support for proexport policies. The gov- racy, and participatory politics. ernment needs to be accountable for services standards, and business for performance. In Source: Prepared by the volume editors based on World many countries closer consultation will be Bank (2000a): ch. 7. 525 52 D A V I D G . TA R R Arguments for An effective trade policy is central to the integration of developing countries into the and against international economic system and the growth that integration Uniform Tariffs will generate. Trade policy, together with the exchange rate, forms the transmission mecha- nism through which interna- tional trade affects domestic resource allocation, the efficient and competitive restructuring of industry and agriculture, access W to new and diverse technologies, orldwide experience in the past 50 improved incentives to exporters, and reduction of years demonstrates the benefits of smuggling, rent-seeking, and corruption in customs. open trade regimes. The OECD countries brought Tariff policy is the centerpiece of trade policy in a trade barriers down through successive WTO nego- market system. Tariffs are, with very few exceptions, tiations and experienced sustained growth in trade the only acceptable policy tool for protection under and incomes. Many developing country govern- the GATT/WTO. They are superior to alternative ments initially took a different stance and attempt- instruments of protection such as nontariff barriers ed to promote industrialization behind high (NTBs)--that is, quotas, licenses, and technical bar- protective barriers. But in the past 10 years or so, riers to trade (TBTs)--because they are less likely to the balance of opinion has shifted in these coun- lead to rent-seeking and corrupt practices, and tries as well, with the growing evidence that high because tariffs limit the exercise of domestic monop- rates of protection significantly depress economic oly power where it exists, whereas NTBs do not. development, and that open trade regimes are This chapter examines the arguments for and more conducive to growth. Moreover, virtually all against a uniform tariff structure. Arguments recent development success stories have been based against uniformity have to do with terms of trade; on strong growth of industrial exports. Exporters promotion of "strategic" or infant industries or have not been disadvantaged in these successful restructuring of industries; revenue or balance of economies, either because there were low barriers payments considerations; and the utility of tariffs to imports--as in Chile, Hong Kong (China), and as a negotiating tool at the WTO. Arguments in Singapore--or because regimes were developed to favor of uniformity include political-economy con- provide incentives to exporters comparable to siderations, administrative convenience, and import-competing sectors despite the protection, reduction of smuggling and corruption in customs. as in the Republic of Korea, Mauritius, and Taiwan We maintain that tariff uniformity is the best (China).1 choice, in practice. 526 Arguments for and against Uniform Tariffs In many circumstances in which tariffs are sec- "infant"or restructuring industry, (d) to raise budg- ond-best policy instruments--for raising public etary revenue, (e) to reduce imports as a remedy for revenues or coping with balance of payments prob- balance of payments problems, and (f) as bargain- lems, for example--a uniform tariff rate is the most ing tools to extract concessions from trading part- practical and efficient alternative. If a country is ners. First-best policies in the pursuit of these interested in using the tariff as a bargaining instru- objectives are subsidies (objective c), indirect taxes ment in multilateral negotiations, it is immaterial (objective d), and devaluation and other macroeco- whether the tariff is uniform or differentiated; the nomic policies (objective e).3 When the first-best issues are the country's capacity to use the tariff as a policies are not available, we argue below that low bargaining instrument and what it will bargain for. and uniform tariffs are preferable to a high and var- Differentiated tariff protection in support of infant ied tariff structure. or restructuring industries is typically ineffective at addressing the alleged market failure problem: gov- Tariffs to Exploit Monopsony Power ernments are not very good at picking winners, and there are serious dangers that the policy will be One generally accepted theoretical reason for a overwhelmed by requests for protection from vested country to impose tariffs on individual products is interests irrespective of economic merits. in order to exploit its monopsony power and there- A uniform tariff conveys a number of advantages, by improve its terms of trade. If a country is large the most important of which is that if the tariff is enough that it imports a significant share of the uniform, the gains to industry lobbying are much world's supply of a particular product, a tariff on smaller (and may be negative), creating a free-rider that product could lower the price the country must problem for the lobbying industry and dramatically pay to world suppliers. Consistent with this argu- reducing the incentive to lobby for protection. This ment, the government could impose tariffs at differ- means that (a) the level of protection is likely to be ent levels on different products to exploit the lower (the recent experience of Chile is a dramatic monopsony power it possesses, and the "optimal" case in point); (b) there is a direct saving of tariff on each product would be different. resources because of the reduced lobbying; (c) the Although the theory is valid, in practice there are reduction in the gains from lobbying for protection very few products in which the typical developing provides a vastly improved signal to valuable entre- country possesses sufficient monopsony power for preneurial talent, which will thus be encouraged to this argument to be relevant. Even then, the tariffs create better and cheaper products; and (d) the would typically be quite small (1 to 10 percent) reduction in resources devoted to lobbying will because the share of world imports would have to result in less corruption in government, and this be large indeed to allow high tariffs. Furthermore, may have positive spillover effects on other dimen- the actual tariffs for most countries are typically sions of government activity. larger than the values optimal tariffs could reason- ably be expected to take. For all practical purposes, tariff policy can be established without reference to Arguments for Tariffs and for Nonuniform this basically theoretical issue. Tariffs There are several arguments in favor of government Tariffs to Gain Strategic Advantage intervention through tariffs. Some of these support tariffs as first-best policies and call for nonuniform In recent years a number of arguments have been tariffs. The reasons given are (a) to exploit a developed justifying tariffs on the basis of strategic monopsony position and thereby improve the considerations in industries with excess profits that terms of trade, and (b) to maximize benefits from a are highly concentrated globally. Among others, "strategic" application of protection. Often, govern- Brander and Spencer (1985) developed models ments wish to pursue other objectives than the pur- showing that in the presence of an international oli- suit of real income, or there may be constraints on gopoly, tariffs could increase a country's welfare by the use of first-best instruments to achieve those enabling excess profits to be shifted from foreign to objectives.2 Other justifications for tariffs are (c) as domestic firms. Krugman (1992) showed that given instruments for temporary protection of a specific the existence of increasing returns to scale in the 527 T R A D E P O L I C I E S Q U E S T I O N S A N D G U I D E L I N E S firm, protection to allow domestic firms to gain ini- arguments have been made in the context of tial competitive advantage at the expense of foreign economies in transition, where it is argued that if firms could be reinforced by internal economies of some firms in the process of restructuring are given scale and would allow domestic firms to appropri- protection for a time, they will be able to increase ate excess profits. their productivity and become viable in the longer Despite its popularity among theorists in the term. A firm may be faced with imperfections in its 1980s, today strategic trade theory is not regarded as markets for inputs that raise its costs--for example, a significant policy choice. There is doubt as to because an inefficient banking sector prevents it whether excess profits really exist (except in the very from obtaining credit. short term) in many industries worldwide and The argument, then, is that temporary tariffs may whether they are not easily dissipated by new be necessary to protect these infant and restructur- entrants or by utilization of excess capacity. Eaton ing industries so that they can generate benefits for and Grossman (1986) have shown that the policy the economy as a whole. Under the infant industry conclusions of strategic trade policy models can be argument for protection, the optimum tariff struc- completely reversed, depending on assumptions ture would not normally be uniform because pro- about which little is known. For example, whether tection would be accorded only to specific firms compete in price or in quantity will reverse industries affected by market failure or externalities the optimal policy conclusion from subsidizing an but would not be warranted for other industries. industry to taxing the industry. One of the principal But as Baldwin (1969) has explained, a tariff will authors of this literature has concluded that the not typically address the market failure problem risks of following strategic trade theory far out- and therefore is not better than laissez-faire policy. weigh the possible gains (Krugman 1989, 1992); a Consider, for example, the case of the inability of country might make small gains in some circum- the firm to appropriate the gains from investment stances, but the measures are more likely to be mis- in technical know-how. A duty raises the domestic applied and to lead to large losses. Given that a price of a product and, from the viewpoint of the typical developing country seldom has a manufac- domestic industry as a whole, makes some invest- turing firm competing in international markets ments in knowledge more profitable. But the indi- dominated by only a few firms with excess profits, vidual entrepreneur still faces the same externality the potential scope for strategic trade policy is problem as before--the risk that other firms in the inherently limited. same industry will copy, without cost to them- selves, any new technology discovered by the firm and will then drive the product's price or factor Infant Industry and Restructuring Protection prices to levels at which the initial firm will be The use of tariffs to effect welfare improvements unable to recover the costs of acquiring knowledge over the longer run has been defended most often (Baldwin 1969: 298). on infant industry grounds. The infant industry Thus, a tariff does not correct the problem. argument posits that certain industries are initially Indeed, it has been shown more generally that the uneconomic but that they may become competitive best intervention is a policy which attacks the prob- (at world prices) in the long run because costs may lem at the source (Bhagwati and Srinivasan 1969). decrease over time through learning-by-doing In this case, the appropriate interventions directed effects. Market failures stemming from gains that at the source of the distortion, which could be are external to the firm may prevent the develop- imperfect appropriability, labor turnover, or capital ment of industries that exhibit positive discounted market imperfections, are not tariffs but rather present values. For example, a firm may be unwill- measures such as provision of information, patent ing to invest in technical know-how that may protection, or more effective use of instruments to become freely available to other firms; that is, the allow collateral.4 Sometimes a government may activities of an individual firm could generate exter- argue that the whole manufacturing sector is an nalities that cannot be captured by the firm. (If infant. Although protection is unlikely to be the there were no externalities, the firm would be will- appropriate response, if any protection is offered for ing to make the investments, and there would be no this purpose, a uniform tariff would be called for, need to depart from laissez-faire policy.) Similar not a diverse structure. 528 Arguments for and against Uniform Tariffs Revenue Considerations lem is a macroeconomic problem, and the optimal response is to attack the problem directly through Trade taxes are not optimal instruments for achiev- macroeconomic tools--that is, through a combina- ing a revenue objective because they significantly tion of actions to reduce domestic spending (expen- distort production and consumption choices. Pre- diture reduction) and to encourage exports and ferred instruments for raising revenue are taxes discourage imports (expenditure switching). such as income taxes or commodity taxes (excise Expenditure reduction can be achieved through fis- taxes, value-added taxes, and so on), which, since cal or monetary tightening, which reduces domestic they are applied neutrally to domestically produced absorption for any given level of output. Expendi- and imported goods, impose less distortion or inef- ture switching, which is best accomplished through ficiency costs. The use of tariffs to raise revenue pre- depreciation of the real exchange rate, raises the supposes that other trade-neutral tax instruments domestic price of tradables in relation to nontrad- are not available or cannot be used beyond existing ables, thereby encouraging exports and discourag- levels; in other words, domestic taxes have to be ing imports. Across-the-board import surcharges taken as given either because the tax base cannot be are often applied for balance of payments reasons; enlarged rapidly enough or because marginal costs this has the same effect in reducing imports as does of increased domestic tax collection are very high exchange rate depreciation, but it fails to achieve the (Corden 1974; Balassa 1989; Mitra 1992). beneficial effects on the export side. The optimal One of the best-known arguments for a nonuni- tariff structure, given that it is a surrogate for a form structure is the inverse elasticity rule. If the devaluation (without export incentives), must be economy is characterized by only final goods uniform, inducing resources to flow into import- (ignoring rent-seeking, administrative, and smug- competing industries in general rather than into gling costs), the most efficient way to generate tax any particular import-competing industry. revenue is to impose higher tariffs on the goods with the lower elasticity of demand (Ramsey 1927). Tariffs as a Negotiating Tool and WTO Accession This causes the least distortion, since it diverts the least resources. The simplicity of this rule has appeal Finally, countries may use tariffs as a bargaining for theoretical economists. The rule becomes tool to extract concessions from trading partners in exceedingly complex in practice, however, because multilateral trade negotiations within the context of of substitution effects between goods and the pres- the WTO. For countries applying for WTO mem- ence of intermediate goods and of goods that can- bership, the future level and structure of tariffs are not be effectively taxed. Application then requires important elements of the accession negotiations. not just the own-elasticities but a complete set of The WTO, however, does not focus on the actual cross-elasticities of demand, including the substitu- level of tariffs (the "applied rate") but on the tion elasticities with untaxed goods such as house- "bound" rate. The bound rate is the maximum legal hold leisure and underground economy goods. In level of the tariff for each individual tariff line that a fact, this information is never available, so that country may not exceed without either renegotiat- information requirements make the application of ing the tariff or providing compensation by reduc- Ramsey-type rules impractical. We know of no ing the tariff level for other products. country that has actually tried to implement them. (Michalopoulos treats this issue in some depth in A still more important reason to avoid Ramsey-type Chapter 8, on WTO accession.) In summary, it is rules for diverse tariffs is that lobbying is likely to unwise for acceding countries to use tariffs as a bar- lead to the application of tariffs which depart from gaining tool, since experience has shown that this is uniformity in economically inefficient ways. We likely to delay accession. More generally, if a country return to the lobbying argument below. were to succeed in negotiating a structure of high bound rates, it may have gained a pyrrhic victory: by negotiating such a structure, it would create an Balance of Payments Considerations opening for domestic interests to exert political Tariffs are sometimes employed to deal with a bal- pressure for additional protection in the future. The ance of payments problem, but, again, they are not government would lose the "political cover" that the best instrument. A balance of payments prob- legally binding WTO commitments offer against 529 T R A D E P O L I C I E S Q U E S T I O N S A N D G U I D E L I N E S domestic protectionist interests, which may other- iff. They would prefer that someone else do it for wise succeed in subverting the trade regime and them, allowing them to free-ride on the efforts of making it far more protective, to the detriment of similarly minded individuals. The result is that typ- long-term efficient industrialization. ically only the industry that gains from the tariff Regarding the use of the initial tariff offer as a lobbies the government, and governments some- bargaining tool at the WTO, the initial tariff offer times yield to this one-sided pressure. could be uniform without loss of bargaining power. The advantage of a uniform tariff is that it makes There is no reason to believe that an initial diverse the gains to the industry much smaller, creating a tariff offer that is based on political-economy or free-rider problem for the lobbying industry, and so other considerations would be superior to a uni- dramatically reduces the incentive to lobby for pro- form tariff offer as a bargaining position at the tection. If a country employs a uniform tariff, an WTO. Thus, the bargaining power argument is not industry will not receive concentrated gains from its an argument against uniformity but an argument lobbying; if it did succeed in raising the uniform tar- about the level of the tariff. iff, it would have to bear the costs of raising the tariff In sum, for an economy acceding to the WTO, for all the other products. These costs would include binding tariffs at low uniform levels is likely to be the higher cost of imported intermediate inputs and helpful to the longer-term development of an inter- the lower price of the industry's exports as a result of nationally competitive and efficient industrial induced changes in the real exchange rate. Any gains structure and will, in addition, facilitate the acces- from raising the tariff would be dispersed and would sion process. have to be weighed against the dispersed costs that the higher tariff in other industries would impose.5 A strong case in point is Chile, which has had a Arguments in Favor of a Uniform Tariff uniform tariff since 1979. In 1998 the legislature Most policy-oriented analyses (Harberger 1988; Bal- considered a progressive reduction of the uniform assa 1989; Thomas, Nash, and associates 1991; Mitra tariff from 11 to 6 percent, to be accomplished 1992) start with the assumption that if protection is through reductions of 1 percent per year though to be granted at all, it should not favor any specific 2003. The lobbying and testimony of Chilean indus- industry or set of industries. In other words, uniform try groups supported a reduction of the tariff, which effective protection should be provided to all indus- passed the Chilean legislature. Evidently, uniform tries. The arguments supporting a uniform tariff take tariffs led the industrialists to conclude that a reduc- into account political-economy considerations, lob- tion was in their interest. bying, administrative and smuggling costs, and nega- The fact that a uniform tariff reduces the gains tive experiences with picking winners. from lobbying for protection conveys several advan- tages: Political-Economy Considerations · Most important, the level of protection is likely to Political pressures from constituents frequently be lower, for the reasons discussed above. induce governments to depart from trade policies · Lobbying for protection is unproductive activity that are preferred on grounds of economic welfare. and a waste of resources, whereas there is a direct Diverse and inefficient tariffs typically arise as a saving of resources when lobbying is reduced. result of a free-rider problem in political lobbying. · The reduction of the gains from lobbying for pro- Political interests that want tariffs are usually com- tection gives a vastly improved signal to entrepre- panies or unions engaged in the affected industries; neurs. Entrepreneurs need to believe that they the gains are concentrated in relatively few hands, have more to gain by creating better and cheaper and the interests are able to capture enough of the products or production processes than by lobby- gains to warrant devoting resources to lobbying for ing their government. Entrepreneurial talent is the tariff. By contrast, those who lose from a tariff scarce and valuable, and if it is diverted into rent- are the consumers of the product, and although seeking, the growth rate of the economy can be they greatly outnumber the protariff interest, their adversely affected. costs are too small to induce them to spend · The reduction in the resources devoted to lobby- resources to lobby the government to avoid the tar- ing will result in less corruption in government, 530 Arguments for and against Uniform Tariffs and this may have positive spillover effects in my than a structure that has a 20 percent tariff on other dimensions of government activity. half the products and zero tariffs on the rest.6 Turkey had succeeded in lowering its overall average tariff to less than 15 percent, but the bulk of the dis- Administrative Convenience tortion costs came from the relatively few sectors Uniform tariffs have a number of administrative where tariffs or export subsidies were quite high. advantages. First, if tariffs are uniform, there is no In another example, Chile has a uniform tariff of incentive to misclassify goods. This enables customs 11 percent and has elected to participate in a free authorities to concentrate on ensuring that the trade area with the Common Market of the South, value of the imported goods is not understated and MERCOSUR, which comprises Argentina, Brazil, will reduce corruption related to customs clearing. Paraguay, and Uruguay. Chile has refused to join In addition, the transparency and administrative MERCOSUR, in part because the customs union simplicity of uniformity in customs clearance pro- employs a common external tariff that is not uni- cedures will lower the administrative costs of trad- form. Harrison, Rutherford, and Tarr (1997) have ing. For example, traders will not have to obtain estimated that the free trade area with MERCOSUR information on the category under which their is substantially better for Chile than the customs products will fall. union because the nonuniformity of MERCOSUR's common external tariff would impose considerable distortion costs on Chile. Similar results, although Reduced Smuggling not as strong, were found for the Philippines Diverse tariffs provide an incentive to smuggle (Clarette 1989) and India (Mitra 1994). products that are subject to a high tariff--the "out- liers" of the tariff structure. If the tariff is uniform, International Experience with Tariff these strong incentives for smuggling are consider- Uniformity ably reduced. Unfortunately data do not exist that would allow precise international comparisons of either tariff Empirical Evidence on the Benefits of Uniformity uniformity or escalation. Although most-favored- Empirical evaluations using computable general nation (MFN) ad valorem tariff rates are available, equilibrium (CGE) models of actual economies have estimates of the ad valorem equivalents of either found that, holding government revenue constant, specific tariffs or nontariff barriers are not. The lat- movements toward uniformity of the tariff structure ter data are not available because estimating tariff increase real income. Martinez de Prera (2000) eval- equivalence is very time- and data-intensive and uated the consequences of moving to uniform tariffs involves thousands of tariff lines in many countries. from the actual tariff structures in CGE models of 13 Calculations have been performed based on the separate countries and found that in all 13 countries ad valorem rate, and these estimates should provide there would be welfare gains from tariff uniformity. a lower-bound-biased estimate of actual tariff dis- Evidently, tariffs do not differ from uniformity in persion when the tariff equivalence of NTBs is taken these economies, for Ramsey-type reasons. into account. Despite the arguments in favor of rel- A study of Turkey by Harrison, Rutherford, and atively uniform protection, actual experience with Tarr (1993) found that uniformity in the incentives tariffs worldwide suggests that most countries dif- to importers and exporters would provide more ferentiate their tariffs substantially. (Country exam- than two-thirds of the gains to the economy of ples are discussed below.)7 Typically, the protection going to full free trade. That is, quantitatively the pattern involves low tariffs for unprocessed com- most important distortion in Turkey's trade regime modities and raw materials, as well as capital goods, was not the overall height of the tariff but, rather, and much higher tariffs for processed final goods. nonuniform import tariff and export subsidy rates. The basic reason for this is the influence of vested The reason for these results is that the distortion interests that lobby for protection to the final goods costs to the economy increase more than propor- produced in the country but also for tariff-free tionately with the height of the tariff--a uniform access to their inputs. When there is no domestic tariff of 10 percent is much less costly to the econo- intermediate goods industry, or it is small, there is 531 T R A D E P O L I C I E S Q U E S T I O N S A N D G U I D E L I N E S no effective opposing lobbying influence advocating didates that are most likely to meet the conditions tariffs on these intermediates. The result is low tar- justifying intervention--and to choose and main- iffs on intermediates and high tariffs on selected tain the appropriate level for the tariff, subsidy, or final goods--a situation known as tariff escalation. other policy variable; (b) be immune to pressures The tariff escalation that characterizes the trade from vested groups that inevitably arise once will- regimes of many countries, both industrial and ingness to grant special status is established; and developing, causes problems of inefficient resource (c) prevent any protection granted from becoming allocation. In the long run this escalating tariff permanent. The empirical evidence in both indus- structure that tends to favor production of final trial and developing countries during the past three goods at the expense of intermediates encourages decades casts doubt on most governments' ability assembly-type activities while discouraging produc- to meet these conditions. Moreover, the economy tion of intermediate goods. The result is that must provide its most talented members with the because an intermediate goods industry does not incentive to engage in entrepreneurial activities exist today to lobby for equal protection, incentives such as starting or expanding firms, developing are established that hinder its eventual creation. new products, and lowering costs. If tariffs, subsi- It is well known that Chile has a uniform tariff, dies, or tax exemptions differ greatly by sector, tal- but there are quite a few other economies with tariff ented people will find it more profitable to engage structures that are uniform or at least close to uni- in the socially wasteful activity of lobbying the gov- form. Estonia and Hong Kong (China) have uni- ernment for these privileges. Endorsement of a form tariffs because they practice free trade. Bolivia more general approach--with little differentiation and the Kyrgyz Republic have virtually uniform tar- in the level of assistance--thus emanates from a iff schedules of 10 percent. Singapore has a simple wider skepticism about the practical merits of tar- tariff average of 0.5 percent and a standard devia- geting of any kind (Krugman 1989, 1992; Westphal tion of less than 3 percent. A number of other coun- 1990). Experience therefore indicates that the best tries, including Brunei Darussalam, Ecuador, industrial policy is for the government to provide a Honduras, and Mexico, have tariff averages under stable macroeconomic and regulatory environment 13 percent and small variances (under 6 percent). conducive to business development, with neutral At the other end of the spectrum are countries incentives for all firms and industries. Some practi- such as Bangladesh and India, with tariff averages of cal steps suggested by the above arguments are dis- 84 and 56 percent and tariff variances of 26 and 24 cussed next. percent, respectively. Korea, Mexico, South Africa, Concertina Approach. If a uniform tariff structure and Turkey have more than 10,000 tariff headings, is to be put in place sequentially, priority should be while most other countries average about 6,000 tar- given to reducing the highest rates. (For further dis- iff headings. A large number of economies have cussion, see Chapter 53, by Panagariya, in this vol- granted exceptional levels of protection to a limited ume.) The costs in the form of inefficiencies in number of products. The list--Cameroon, Canada, resource allocation rise more than proportionately China, Egypt, the European Union, Hungary, India, with the height of the tariff. As a consequence, the Indonesia, Israel, Nepal, Nicaragua, Norway, Saudi greatest gains will come from reductions in the Arabia, the Solomon Islands, Turkey, and the United maximum rates. In addition, very high tariffs may States--includes some of the poorest countries but be prohibitive of imports, so that there will be rev- also some of the most prominent OECD members. enue gains from reductions in the rates. Reductions of the high rates will also reduce smuggling, corrup- tion, and rent-seeking disproportionately. Conclusions and Practical Steps Simultaneously raising the low rates (as suggested The above analysis strongly suggests that there is by Hatta 1977) is more controversial, especially with little economic justification and there are many respect to intermediate and capital goods (see Neary dangers in providing differentiated tariff protec- 1997). If there is no duty drawback or its equivalent tion to various sectors of industry and agriculture. in place, increasing the tariff on intermediate goods At the practical level, the arguments for a diverse imposes a tax on the exportable goods that use the tariff structure rest on the ability of governments to intermediate, so that because of the tariffs, too few (a) "pick the winners"--that is, to identify the can- resources are devoted to exports.8 Moreover, it may 532 Arguments for and against Uniform Tariffs be argued that raising tariffs on imports of raw addition, any scheme that exempts intermediates materials and intermediates penalizes "technology" from tariffs could reduce the incentive for real imports that are critical for increasing productivity. import liberalization, which is the first-best policy Against these potential costs, one must weigh the choice (see Cadot, de Melo, and Olarreaga 2000). fact that permitting tariff-free imports of intermedi- Low uniform tariffs, in general, are the best policy ates and capital goods penalizes the development of and are best combined with duty drawback or tem- intermediate and capital goods industries in relation porary admission, depending on administrative to final goods sectors and represents a forgone competence and the level of the tariff. In many opportunity to raise revenue. And with a diverse tar- countries it will also be important to obtain techni- iff structure, the political-economy, corruption, and cal assistance for institutional development of duty smuggling problems persist.9 drawback and temporary admission mechanisms. Duty Drawback and Temporary Admission. To Multiple Tariff Bands (Tariff Simplification): Not reduce the added antiexport bias created by raising Identical with Tariff Uniformity. It is sometimes tariffs on intermediate and capital goods, many argued that for administrative convenience the tariff countries employ mechanisms that allow exporters structure should be simplified into three to five tar- duty-free access to imported intermediates. These iff bands. For example, with five tariff bands, tariffs mechanisms include duty drawback procedures, could be 0, 10, 20, 30, or 40 percent (or 0, 5, 10, 15, temporary admission, and export-processing zones. or 20 percent), but values in between would be pro- Coupled with effective duty-free access to imported hibited. It should be clear that tariff simplification is intermediates for exporters, the welfare tradeoff not tariff uniformity and that simplification will from raising tariffs on intermediate and capital allow very high rates of effective protection. An goods is much more likely to be positive. important point is that such a system suffers from As noted in Chapter 18, by English and De Wulf, virtually all the problems of a diverse structure, in this volume, the principal problem with duty including encouragement of lobbying for high pro- drawback schemes is that their administration can tection by industry groups, and will encourage mis- be very costly and can lead to cumbersome proce- classification by customs authorities, in comparison dures and delays. Moreover, these schemes do not with a uniform system.11 remove all the antiexport bias of tariffs.10 Given the If, however, tariff simplification is used as a vehi- recognized need to provide duty-free access to cle for moving toward low and uniform tariffs by imported inputs to exporters, in countries where limiting the number of tariffs and reducing both the capacity to administer duty drawback schemes tariff levels and the dispersion of the tariff structure is weak "temporary admission," should be offered in the process, then it is a very useful step. More (as opposed to zero tariffs on intermediates). Tem- generally, a tariff structure that is low and has a porary admission guarantees duty-free access to small standard deviation will convey many of the imported intermediates, and to the extent that the same benefits of a low uniform structure. For exam- government bureaucracy can administer the pro- ple, with a sufficiently small standard deviation, gram, it imposes tariffs on imported inputs destined there will be little gain from lobbying or incentives for the domestic market, thereby encouraging the for corruption and, in customs, for misclassifica- development of domestic intermediate goods tion. But tariff simplification by itself, without industries. Temporary admission and zero tariffs on reduction of the level or dispersion of the tariff intermediates both provide tariff-free access to structure, will convey relatively small benefits from intermediates for exporters, but temporary admis- lower administrative costs. sion diminishes tariff escalation in comparison with The Tariff Level. OECD countries have, on aver- zero tariffs on intermediates. age, reduced their tariffs on manufactures to less In general, raising tariffs on intermediates poses a than 5 percent, although a few peaks, notably in conflict between the need to provide balanced pro- agriculture, textiles, and leather products, remain. tection to intermediates and final goods (that is, to The main problem in most countries is agriculture, reduce effective protection on final goods) and the where, as a result of the tariffication following the need to reduce antiexport bias. Duty drawback Uruguay Round, tariff schedules are quite high, appears to resolve the conflict, but it does so at the reflecting the previously high supports and protec- expense of creating administrative complexity. In tion. If revenue from a tariff is needed, countries 533 T R A D E P O L I C I E S Q U E S T I O N S A N D G U I D E L I N E S could aim at a 10 percent uniform tariff. Anything 4 For a similar view, see Krueger (1984). If, for purely political higher could result in significant effective rates of rather than economic reasons, a government wishes to protection for selected industries, especially when achieve a minimum output level in a given industry, the best intervention is a production subsidy, which avoids the con- tariff escalation is taken into account. sumption distortion costs of the tariff. A production subsidy also has the advantage of gearing the infant industry to attain international competitiveness by avoiding discrimination Notes between sales to the domestic and export markets. I would like to thank Harry Broadman, Tatsuo Hatta, and Constan- 5 In addition, since some imported inputs are used by import- tine Michalopoulos for their comments. An earlier version of the competing sectors, a uniform tariff leads to a lower level of paper appeared as Tarr (1998). lobbying because it raises the costs of the import-competing See Bhagwati and Srinivasan (1999) for a discussion of the evi- sectors. See Panagariya and Rodrik (1993) for details. dence on the benefits of openness and a review of the criticism; Sachs and Warner (1995) and Dollar (1992) for estimations of the 6 The distortion costs of the tariff increase with the square of the gains from openness; and Edwards (1993) for a review of the tariff rate. See Morkre and Tarr (1980): ch. 2, for a derivation. empirical literature on the link between open trade regimes and 7 Based on unpublished calculations, using the TRAINS data- economic growth. base, by Ulrich Reincke, World Bank. 1 To offset the antiexport bias of import protection, these latter 8 The Lerner symmetry theorem has shown that a tax on regimes involve more complicated economic management imports imposes a tax on exports. One of the principal reasons (including very competitive exchange rates) and require a is that tariffs cause the real exchange rate to appreciate, mak- technically efficient government bureaucracy that is relatively ing exporting less attractive. immune to corrupt practices. Very few developing countries meet the conditions necessary to manage high protection 9 Tariffs on intermediates also convey the benefit of reducing the regimes that provide good incentives to exporters. For these incentive to produce import-competing products that use the reasons, relatively open regimes are recommended for most intermediate. This is beneficial because, due to tariffs, the developing countries. economy allocates too many resources to these activities. A more complete treatment of this subject is found in Panagariya 2 The technical term used by economists for real income is eco- (1992). nomic welfare. The two terms are used interchangeably in this chapter. 10 To avoid antiexport bias, duty drawback schemes would also have to be extended to indirect exporters (firms that do not 3 Income distribution objectives would call for higher tariffs on export themselves but that sell to exporters). Administration of products with higher luxury content. Income taxes would be such mechanisms is quite complicated in practice, and they best for the purpose of income redistribution but might not be are not recommended for most countries. available. Thus, the use of tariffs for income distribution pur- poses entails efficiency and growth losses. Commodity taxa- 11 Hatta (1986), however, has shown that if commodities that are tion, such as a sales tax that does not discriminate by source, close substitutes are grouped together, this reduces distor- would also involve less efficiency loss than tariffs. tions. 534 53 A R V I N D PA N A G A R I YA Formula ing partners by sectors, as was the case in the earlier rounds, or agree on a general formula Approaches to applicable uniformly to all members, as in the Kennedy and Reciprocal Tariff Tokyo Rounds, or pursue a com- bination of the two, as in the Liberalization Uruguay Round. Multilateral liberalization under GATT auspices differs from unilateral liberalization in one key respect: it brings gains to a country not merely from its A own liberalization (assuming rticle XXVIII bis of the GATT calls that the country is small, or abstracting from the on WTO members to periodically terms of trade effects) but also from the liberaliza- sponsor negotiations aimed at substantial reduc- tion of its partners.1 The more a country is able to tions in the general level of tariffs on a reciprocal get its trading partners with large markets for its basis. Paragraph 2(a) of this article states: products to liberalize, the more it stands to benefit. From the viewpoint of developing countries, this Negotiations under this Article may be carried means opting for an approach that leads to maxi- out on a selective product-by-product basis or mum opening by industrial countries in products of by the application of such multilateral proce- interest to developing countries. dures as may be accepted by the contracting In trade negotiations, countries view their own parties concerned. Such negotiations may be liberalization as a cost to be paid for obtaining directed towards the reduction of duties, the expanded access to the markets of partner coun- binding of duties at then existing levels or tries. These costs may take at least three forms. First, undertaking that individual duties or the aver- if the country is large in world markets, its own lib- age duties on specified categories of products eralization may increase world demand for the shall not exceed specified levels. The binding goods it imports sufficiently that these goods against increase of low duties or of duty-free become more expensive. For developing countries, treatment shall, in principle, be recognized as a which are generally small in world markets, this risk concession equivalent in value to the reduction is minimal. At the aggregate level, the proportion of of high duties. income spent by industrial countries on developing country goods is less than 2 percent. Second, trade Thus, GATT gives member countries substantial liberalization makes import-competing goods flexibility with respect to how tariffs are lowered. cheaper in relation to exportable goods, which in They can engage in bargains with their major trad- turn leads to a redistribution of income from the 535 T R A D E P O L I C I E S Q U E S T I O N S A N D G U I D E L I N E S owners of factories specialized in the production of identified, and all members may liberalize imports import-competing goods to those specialized in within those sectors. This is sometimes called a export goods. This may be politically costly. Finally, "zero for zero" approach. Second, each member the reallocation of resources induced by trade liber- may seek liberalization from its major trading part- alization may itself bring with it short-run real ners in its sectors of comparative advantage in costs. return for its own liberalization of sectors of com- The presence of these costs automatically gives rise parative disadvantage. to the notion that tariff reductions should be bal- The first approach has guided much of the liber- anced according to some criterion: the cost of one's alization of trade in services during and since the own liberalization must be balanced by the benefits Uruguay Round. Member countries have identified from liberalization by one's trading partners. The broad sectors such as financial services and obvious criterion for achieving this balance with telecommunications and have bargained market respect to the terms of trade cost is to ensure that the access within them. The approach was also applied terms of trade are unchanged in the postliberaliza- to trade in goods during the first WTO ministerial tion equilibrium. This would effectively allow each meeting, held in Singapore. At that meeting, a group country to benefit from the efficiency gains resulting of countries signed the Information Technology from its own liberalization without redistribution of Agreement (ITA), under which they committed income across countries. From the viewpoint of themselves to complete free trade in a set of tech- small developing countries, this may not be a major nology products. issue, since their own liberalization has little impact For liberalization of trade goods, this sectoral on the terms of trade. But for large economies such approach is not particularly attractive. There are as the United States, the European Union (EU), and two problems. First, there is some concern that Japan, it is important and plays out in the form of under this approach, the sectors that are liberalized demands for reciprocity. first will be those in which economically powerful As for the long-run income distribution effects, nations have export interests. Sectors in which unless one is faced with the conservative welfare developing countries enjoy comparative advantage, function such that any movement away from the such as textiles and clothing, will be taken up last. existing equilibrium is seen as harmful, there is no This view is certainly supported by the fact that, on presumption that the effects of trade are harmful. balance, industrial countries are exporters and Indeed, to the extent that most developing countries developing countries importers of the products are exporters of labor-intensive goods, trade liberal- covered by the ITA. In addition, even in services, ization is likely to improve income distribution by negotiations in sectors in which developing coun- raising the return to labor at the expense of capital. tries have an overwhelming advantage have not Finally, phasing in liberalization over a number of been opened. years can minimize the adjustment costs. Starting The second drawback of this approach is that ini- with the Kennedy Round, this has been the tially it is likely to lead to a lower level of welfare. approach under the GATT and the WTO. In the case The sectors that will be picked initially are likely to of multilateral liberalization, these costs are likely to be those with lower tariffs to begin with. Economic be especially low, since the simultaneous liberaliza- theory suggests that under plausible assumptions, tion by partner countries allows export industries to the elimination of relatively low tariffs without a expand rapidly to absorb the resources released by simultaneous reduction in the high tariffs leads to import-competing industries. reduced welfare globally, as well as in individual In this chapter I discuss the main approaches to countries. Such a policy change engenders a reallo- trade liberalization that have been taken in past cation of resources from less distorted to more dis- multilateral negotiating rounds and their relative torted sectors. merits. Under the second sectoral approach, all industrial products are made part of the negotiation, but members negotiate with their partners sector by Sectoral Approaches sector. In principle, this approach can be expected Conceptually, two types of sectoral approaches can to yield an efficient outcome. Each member will be distinguished. First, one or more sectors may be seek liberalization in sectors and countries where its 536 Formula Approaches to Reciprocal Tariff Liberalization exports face the highest barriers. In turn, its trading unlikely to rise because input tariffs decline propor- partners will seek access to its most protected sec- tionately more than output tariffs. If the countries tors. Thus, the bargain is biased in favor of lowering engaging in liberalization have equal levels of tariffs the highest tariffs. and are also of equal size, this approach will lead to The approach may become administratively com- balanced reductions in tariffs across the countries. plex if trade patterns happen to be such that each The bargain will be unbalanced, however, if either country exports goods to one set of countries but of these conditions is violated. For instance, if two imports them from an entirely different set of coun- countries are of equal size but one has an average tries. For instance, if India's exports go mainly to the tariff of 50 percent and the other an average tariff of United States while its imports come from the EU, 5 percent, a 50 percent reduction in the tariff leads bilateral negotiations between India and its two to a 25 percentage point reduction by the former trading partners become difficult. One suspects that but only 2.5 percentage points by the latter. Such under the current structure of trade and tariffs, this liberalization leads to a deterioration of the terms of is not a significant problem (although this has not trade of the first country. Likewise, if both countries been empirically verified). Goods subject to high have 50 percent tariffs but one country is 10 times tariffs in developing countries are largely imported the size of the other, a 50 percent reduction by the from industrial countries and evenly distributed former leads it to give greater market access than it over the latter. Conversely, major exports of devel- receives and hence results in a deterioration of its oping countries face high barriers in all industrial terms of trade. countries; examples include agricultural products There are various approaches that attempt to cor- and apparel.2 rect for these differences in initial tariffs and coun- try size. The simplest approach is to define liberalization in terms of the tariff revenue forgone. Across-the-Board Approaches and Tariff For example, we can set the proportionate reduc- Reduction Formulas tion in tariff equal to Rather than negotiate on a sector-by-sector basis, dti f f member countries may adopt an across-the-board (53.2) = ti (ti pi Mi) = (ti Vi) approach such that all tariffs are rolled back accord- ing to a prespecified formula. Of course, even in a where f is a constant, pi is the border price, Mi is the sector-by-sector approach, some previously agreed quantity of imports, and Vi is the value of initial rule must be used to ensure a balance in the bargain imports at the world price. We can think of f as a (that is, reciprocity). The formulas discussed below measure of market access given by the country are therefore relevant for the sector-by-sector undertaking liberalization. An advantage of tariff approach as well. reductions according to (53.2) is that the formula Perhaps the simplest rule to follow is the reduc- takes into account both the initial tariff rate and the tion in all tariffs by a fixed percentage. For example, size of the country in the world market. To achieve the member countries may agree to roll back all (or the same level of liberalization, f, a country that a subset of) existing tariffs by 50 percent. Formally, imports larger volumes of good i and imposes a we can set higher initial tariff on the latter has to liberalize dti proportionately less to achieve the same level of lib- (53.1) = b eralization. Alternatively, if the initial level of tariff ti in a sector is low, the credit given for a given per- where b is a constant between 0 and 1, ti is the initial centage reduction in the tariff is also low. tariff rate in sector i, dti is the change in ti, and dti/ti Tariff reductions according to (53.2) are defensi- is the proportionate reduction in the initial rate. ble from the viewpoint of balancing the bargain This approach has the advantage that within a between member countries, but for a given level of country it leads to a larger absolute reduction in imports this formula implies a lower proportionate high tariffs and a smaller absolute reduction in low reduction in the tariff whenever the initial tariff is tariffs, which must broadly promote efficiency. high. From an efficiency standpoint, this may not be More protected sectors are liberalized more, and a desirable outcome. For instance, if final goods are effective protection of any particular sector is subject to high tariffs and inputs to low tariffs, this 537 T R A D E P O L I C I E S Q U E S T I O N S A N D G U I D E L I N E S pattern of tariff reduction may lead to increased The EEC argued that as a part of the Tokyo effective protection to certain sectors.3 Thus, there is Round, countries should apply this formula four necessarily a tension between the twin objectives of times, bringing tariff rates in excess of 50 percent balancing a bargain and preserving efficiency across (but less than 100 percent) to below 13 percent. The sectors. United States suggested that all tariffs equal to or An alternative measure that compromises some- above 6.67 percent be cut by 60 percent and that what on the bargain-balancing objective but is more those below this number be cut according to the defensible on efficiency grounds is formula dti f f dti (53.3) = . (53.3b) =1.5ti +0.5. ti (pi Mi) = Vi ti This measure controls for country size in that the This formula is less progressive than (53.3a). larger is the Vi, the smaller is the proportionate Given that the U.S. tariff structure before the Tokyo reduction in the tariff to achieve the given level of Round was characterized by much greater escala- liberalization, f. But the measure is independent of tion than the European tariff structure, it is not sur- the initial level of the tariff. prising that the United States wanted a tariff-cutting Yet another approach goes a step farther toward formula that was much closer to the linear cut. efficiency by rolling back the high tariffs propor- Observe that according to the U.S. approach a tariff tionately more and the lower tariffs proportionately of 0.02 (or 2 percent) would be cut by 0.53 (53 per- less. From the reciprocity angle, the argument made cent), and a tariff of 0.06 would be cut by 59 per- is that linear cuts leave the countries that have high cent, while all tariffs equal to or higher than 6.67 initial tariffs more protected. This argument is percent would be cut by 60 percent. There is only a clearly based on achieving reciprocal concessions in small variation in cuts according to the initial level terms of final outcomes rather than additional mar- of the tariff rate. ket access. Put differently, it relies on the assump- The formula actually applied in the Tokyo Round tion that tariff levels must be harmonized across was the one suggested by Switzerland. According to countries. If participating countries share the objec- the "Swiss formula," tive that the eventual goal is to achieve free trade dti ti everywhere, the approach clearly makes sense. (53.3c) = Nonetheless, it must be recognized that if countries ti r +ti are of equal size but some are initially more protect- where r is a positive constant. This formula has a ed than others, in the transition toward zero tariffs progressive element, the degree of which depends on this approach will result in the latter group's experi- the value of r. The higher is the value of r, the more encing deterioration in their terms of trade. The progressive is tariff reduction in the sense of lower- explanation is similar to that presented earlier and is ing the higher tariffs to a greater degree. In the Tokyo understood simply by assuming that one country Round agreement, the value of r was set between has a 50 percent tariff on its imports while the other 0.14 and 0.16. Letting r = 0.14, tariff rates of 0.14 (14 has only 10 percent tariffs. Elimination of these tar- percent) are reduced by 50 percent. Rates exceeding iffs will result in the first country's opening its mar- 0.14 are reduced more, and those less than 0.14 are ket more at the margin and hence will lead to a reduced less. For instance, a tariff rate of 0.06 (6 per- deterioration in its terms of trade. cent) is reduced by 30 percent, while a tariff rate of The simplest rule for achieving harmonization, 0.36 (36 percent) is reduced by 72 percent. Clearly, suggested by the European Economic Community this formula has much greater potential for harmo- (EEC) during the Tokyo Round, is to lower each tar- nizing tariff rates both across sectors within a coun- iff by the same percentage as its initial ad valorem try and across countries within the same sector. rate. Thus, if the tariff on a good is 70 percent, it is Finally, countries may agree to some average tariff cut by 70 percent, while a tariff of 20 percent is cut reduction without specifying the reductions in spe- by 20 percent. Formally, we have cific, sectoral tariff rates. The average may be simple or weighted. This approach can result in both dti f f (53.3a) = . unequal and inefficient tariff reductions ex post. ti (pi Mi) = Vi The Uruguay Round Agreement on Agriculture 538 Formula Approaches to Reciprocal Tariff Liberalization required industrial countries to reduce tariffs by 36 them (see Appendix A, in this volume). Therefore, percent, on average, with the rate on each item the Swiss-formula-like approach will induce signifi- reduced by at least 15 percent. Assuming that there cant liberalization in the products of interest to are four items in all, with 100 percent tariffs on them. Of course, given high tariffs in many develop- three and a tariff of 1 percent on the fourth, a 15 ing countries themselves, this approach will require percent reduction in the 100 percent rate and elimi- those countries to liberalize more as well. But since nation of the 1 percent rate yields (15 + 15 + 15 + they have minimal market power, their own liberal- 100)/4 = 36.25 percent average reduction. Thus, ization is likely to result principally in efficiency rules calling for an overall average reduction can be gains, without deterioration of their terms of trade. easily manipulated to minimize liberalization.4 A similar argument applies even more to trade in agriculture, since many developing countries are, in fact, potential exporters of agricultural products. Choosing among Approaches A choice among various approaches depends on the Notes underlying objective. If the objective is to achieve This chapter has benefited greatly from the excellent discussion of maximum liberalization worldwide, an across-the- tariff formulas in Hoekman and Kostecki (2001), as well as from board approach that lowers higher tariffs more, such valuable comments by Bernard Hoekman on an earlier draft. as that based on the Swiss formula, would be the right 1 If a country is sufficiently small in the world market, its trade choice. The across-the-board approach minimizes the liberalization increases its imports, but this increase is too small room for successful lobbying by political powerful to have any effect on world prices. In that case, the country sectors, which often happen to be the most protected necessarily benefits from its own liberalization, since it reduces sectors in the first place. Moreover, a formula that the domestic output of goods in which it lacks comparative lowers high tariffs to a greater degree reduces the dis- advantage and expands the output of the goods in which it enjoys comparative advantage. By contrast, if a country is a persion in tariffs and hence lowers effective protection large buyer of imports in the world market, increased demand in all sectors. A formula approach also has the advan- resulting from its liberalization can lead to a rise in the price of tage that it does not tie up negotiating resources in a the imports. This terms of trade effect is harmful to the coun- major way, as do sector-by-sector negotiations. try and, in principle, may be large enough to outweigh the From an individual country's viewpoint, the efficiency gains resulting from increased specialization in the goods of comparative advantage. answer may be different. Assuming that welfare maximization is the objective, a large country with 2 See Hoekman, Ng, and Olarreaga (2001) for an analysis of the high tariffs is likely to prefer equal proportional tar- incidence in least-developed countries of tariff peak protection in Quad markets, net of preferences. iff reductions, while one with low tariffs may prefer reductions that follow the Swiss formula. For the 3 The concept of effective protection refers to protection provided high-tariff large country, proportional reductions to value added in a sector by the entire structure of tariffs. Sup- pose that the cost of auto parts in world markets is 90 percent will yield better terms of trade than the Swiss-for- of the fully assembled automobile. Under free trade, domestic mula-like reductions, while for the low-tariff large assembly of the automobile will be profitable up to the point country, the opposite holds true. that the marginal cost of assembly is $1,000. The introduction Governments that are driven by domestic lobbies of a 10 percent duty on the final automobile without a similar are likely to prefer the sector-by-sector approach, tariff on the imported inputs raises the domestic price of the automobile to $11,000 and the domestic assembly industry can which lends itself most easily to selective offers or expand up to the point that the marginal cost is $2,000. This is requests of sectors for liberalization. Import-com- an expansion of the domestic value added per automobile from peting lobbies with political clout can ensure that $1,000 to $2,000 and amounts to 100 percent protection on their sectors are not offered for liberalization. Simi- the assembly operation. Thus, the effective protection to the larly, powerful export lobbies may pressure their automobile industry turns out to be 10 times the nominal pro- governments to seek market access on their behalf tection! If, in addition, a 10 percent tariff were imposed on auto parts, the domestic value added could expand from $1,000 to from their major importers. $1,100 (since parts now cost $9,900 and the automobile sells In my judgment, in the present scenario, from the for $12,000), which is exactly 10 percent. viewpoint of developing countries an approach like 4 Specifying the reduction in terms of a weighted average does that of the Swiss formula would make the most not solve the problem, since the country can achieve a high sense. This is because industrial countries apply average by liberalizing more where its imports are already high rather high tariffs on goods of export interest to while leaving the most protected sectors virtually untouched. 539 54 J O S E P H F. F R A N C O I S W I L L M A R T I N Binding Tariffs: investment measures. In the Uruguay Round market access bindings were also introduced Why Do It? for the services sectors. Because the multilateral sys- tem centers on bindings, tariff negotiations in the WTO are not actually about applied tariff rates but, rather, about the underlying bound rates. For this reason, an important feature of the policy landscape left by the Uruguay Round agreements is B tariff bindings on industrial and orn out of the experience of esca- agricultural goods. Because tariff bindings are com- lating tariffs in the period between mitments not to raise tariffs above a certain level, the two world wars, the multilateral trading system is their actual relevance depends on how far bound a set of rules that restricts the damage governments rates are above actual applied rates. In the case of can impose through unbridled use of the range of OECD industrial tariffs, there is a close correspon- policy instruments otherwise available to them. In dence between applied and bound rates. This is not most cases multilateral trade rules do not prescribe the case for developing countries. In particular, for precisely what countries must do. Rather, they tend many developing countries (and also for Australia), to operate by imposing limits on the values and the industrial tariff landscape now features bound types of protection that are allowed and by forcing a rates that are often well above applied rates. For the degree of transparency in trade regimes. Like Ulysses poorest developing countries, tariffs are often still tied to the mast, governments are then able to listen completely unbound. to the siren call of protectionist lobbies while plead- Bindings are vital to the process of securing trade ing an inability to actually respond. Tariffs are agreements. If an agreed tariff reduction could easily prohibited from varying across suppliers by most- be reversed unilaterally, any liberalization offer favored-nation (MFN) requirements, and their vari- would have to be weighed against the probability of ation across commodities and over time is limited by backsliding. Exporting firms, which provide much of schedules of tariff bindings (concessions). The the political support for multilateral trade liberaliza- application of nontariff barriers is limited, or even tion, are likely to be unenthusiastic about tariff cuts prohibited, by GATT rules. Contingent protection, they expect to be short-lived. Bindings themselves are through fair trade and safeguard actions, is also (at considered so important that countries agreeing to least in theory) limited by related WTO/GATT disci- bind previously unbound tariffs are given "negotiat- plines. Other rules apply to balance of payments ing credit" for the decision. This is true even if the actions, licensing requirements, and trade-related tariff is bound above the currently applied level. 540 Binding Tariffs: Why Do It? In this chapter we examine the issue of tariff occurred since the end of the Uruguay Round, they bindings and its relationship to the assessment have been undertaken for reasons unrelated to WTO of tariff negotiations. We first examine the commitments. post­Uruguay Round tariff landscape to get a gen- More detailed data are presented in Table 54.2, eral sense of the differences between bound and which presents the number of unbound tariff lines applied rates and the implications that new reduc- (as a percentage of the within-category total) by tions in bound rates (especially for developing economy and GATT/WTO multilateral tariff negoti- countries) will have for applied rates. This is fol- ation (MTN) category. These data are drawn from lowed by a discussion of the implications of bind- the WTO Integrated Data Base (IDB). What is clear ings, in terms of real income gains and security of is that for many individual product categories, sub- market access.1 stantially all trade remains unbound across develop- ing countries, except in Latin America. This applies, for example, to transport equipment, miscellaneous Industrial Tariff Bindings: An Empirical manufactures, fisheries products, metals, and wood Overview products. For many developing economies in the Under the Uruguay Round, the share of developing IDB, over 60 percent of trade remains unbound in country imports of industrial products subject to most product categories; these economies include tariff bindings rose from 13 to 61 percent (Black- Cameroon, Chad, Gabon, India, Macau (China), hurst, Francois, and Enders 1996). This rise was Malaysia, the Philippines, Senegal, Sri Lanka, Thai- mainly the result of commitments by Latin Ameri- land, Tunisia, Turkey, and Zimbabwe. can countries to apply ceiling bindings on 100 per- What is important for future industrial tariff cent of their tariff lines, and commitments made by negotiations is the current level of ceiling bindings Asian developing economies. Some tariff bindings vis-à-vis applied rates and the limited coverage of predate the end of the Uruguay Round. Chile was bindings. Taken together, the combination of bound the only developing country that offered to bind and applied rates means that developing countries 100 percent of its tariff lines in the context of the will, collectively, be able to reduce bound rates, or Tokyo Round, while Costa Rica, El Salvador, Mexi- introduce them for the first time, while having to co, and the República Boliviariana de Venezuela make only modest (and in many cases no) changes bound 100 percent of their tariff lines on their in applied rates. Hence, for industrial tariffs the rel- accession to the GATT during the period 1986­91. evant scenarios for many developing countries in Among Asian developing economies, Indonesia the next round are likely to involve little or no real bound more than 90 percent of its tariff lines dur- reductions in applied tariffs. This will be true ing the Uruguay Round. India, the Republic of whether or not developing countries take an active Korea, Malaysia, the Philippines, Singapore, and part in future industrial tariff negotiations. As in Thailand bound between 60 and 89 percent; Sri previous rounds, there is a good chance that only Lanka and Zimbabwe bound less than 15 percent. OECD countries and some middle-income devel- Table 54.1 presents summary data on bindings and oping countries will actually be forced to reduce applied rates for industrial products in 29 economies. industrial tariffs as part of any upcoming industrial These data reflect the tendency of developing coun- tariff negotiations. The only way to avoid such an try tariffs to be unbound, or to be bound well above outcome may be to make particularly large cuts in applied rates. Where developing economies had the highest tariff bindings, in order to close the cur- bound all or a significant portion of tariffs prior to rent gap between bound and applied rates, or to the end of the Uruguay Round (as had Chile, Costa otherwise redefine the negotiation benchmark (per- Rica, El Salvador, Mexico, and the República haps focusing directly on applied rates) for the pur- Boliviariana de Venezuela), the Uruguay Round tariff pose of defining negotiation parameters. commitments often reflected a decline in ceiling rates rather than necessarily in applied rates. For these rea- Bindings and the Security of Market Access sons, implementation of Uruguay Round tariff com- mitments by developing countries has involved Although the perception is that new bindings did virtually no further declines in current applied tariffs. not have much effect on the applied rates of many This means that, to the extent that reductions have developing countries during the Uruguay Round, 541 T R A D E P O L I C I E S Q U E S T I O N S A N D G U I D E L I N E S Table 54.1 Industrial Applied and Bound Tariff Rates (percent) Share of GATT imports Mean industrial tariff Unbound Bound or bound above above Current applied applied applied Bound Economy Bound rates rates rates rates Argentina 100.0 99.9 99.9 14.1 33.5 Australia 96.9 31.7 34.8 4.4 12.1 Brazil 100.0 91.0 91.0 18.3 27.7 Canada 99.8 45.7 45.9 3.2 4.3 Chile 100.0 99.7 99.7 10.9 24.9 Colombia 100.0 97.7 97.7 10.5 35.2 El Salvador 97.1 96.0 98.9 3.8 31.7 European Union 100.0 17.7 17.7 3.5 3.2 Hungary 93.6 3.3 9.7 9.8 6.1 India 69.3 14.8 45.5 29.5 34.2 Indonesia 92.3 86.6 94.3 14.9 36.9 Japan 95.9 0.1 4.2 1.5 1.5 Korea, Rep. of 89.8 3.4 13.6 7.8 6.9 Malaysia 79.3 31.0 51.7 9.4 8.9 Mexico 100.0 98.4 98.4 12.0 33.3 New Zealand 100.0 46.5 46.5 4.0 11.9 Norway 100.0 36.5 36.5 2.4 2.6 Peru 100.0 98.5 98.5 12.5 29.4 Philippines 67.4 15.5 48.1 9.1 21.3 Poland 92.8 44.6 51.8 14.2 8.5 Singapore 36.5 11.7 75.2 2.7 6.9 Sri Lanka 9.2 1.4 92.2 19.8 17.9 Thailand 67.4 8.9 41.5 43.7 27.3 Tunisia 67.9 41.5 73.6 23.5 38.4 Turkey 49.3 0.0 50.7 5.7 16.3 United States 100.0 14.0 14.0 2.7 3.5 Uruguay 100.0 96.3 96.3 10.7 27.9 R. B. de Venezuela 100.0 90.3 90.3 10.9 31.3 Zimbabwe 13.6 3.9 90.3 20.5 23.6 Sources: Finger, Ingco, and Reincke (1996); World Bank, World Development Indicators 1996. ceiling bindings were nonetheless considered will call this binding overhang), why are bindings, important enough that countries which agreed to even when above applied rates, regarded as valuable bind previously unbound tariffs during the round during negotiations? Is it in a country's own interest were given "negotiating credit" for the decision even to impose such bindings? In this section and the next if the tariff was bound at a level well above the cur- we develop a formal (and admittedly stylized) repre- rently applied level. This is likely to be true in future sentation of rules-based bindings on trade policy. rounds as well. For clarity of exposition, we limit ourselves to a sin- Given our observations in the previous section gle instrument operating as a tariff. Although we about the gap between bound and applied rates (we confine the discussion to a price-based instrument, 542 11 Fish and fish 0.00 0.00 12.99 7.005 5.359 1.49 7.75 6.928 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 products (continued) not 01 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 Manu- factured 18.11 16.01 46.10 72.77 31.33 60.71 articles elsewhere specified 9 precious and 0.00 0.00 3.05 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 Mineral stones 34.92 63.32 85.43 56.17 28.44 products precious metals and y 8 0.00 0.00 8.40 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 Electric 22.51 39.71 46.77 43.01 12.66 machiner y 7 0.00 0.00 7.47 7.62 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 10.98 28.40 38.04 11.76 Nonelectric machiner y 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 ransportT 67.18 60.51 65.13 88.41 75.71 29.51 equipment Categor 56 and 0.00 0.00 3.33 1.32 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 photo- (MTN) graphic 27.15 31.56 45.44 11.23 Chemicals supplies 4 Metals 0.00 0.00 6.34 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 50.15 71.58 35.93 45.66 43.54 Negotiation , , , ariffT 3 travel Leather rubber goods 0.00 0.00 0.72 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 13.00 60.11 66.47 54.39 51.43 footwear and Multilateral TT extilesT and 0.00 0.00 0.38 5.19 2.93 5.61 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 clothing 22.01 74.01 GA by 12 and, ood, W pulp, 0.00 7.25 1.18 3.96 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 80.04 54.08 14.15 38.46 Lines paper furniture ariffT unbound) lines Unbound tariff America of Union Asia States Latin 54.2 of centage Indonesia Malayasia Philippines Singapore Thailand Argentina Chile Colombia Mexico Peru enezuelaV ableT (per Region European Japan Southeast India United Brazil Rest 543 11 Fish and fish 56.10 84.38 95.80 66.70 61.72 95.33 products 100.00 100.00 not 01 2.06 4.65 4.29 Manu- 96.20 56.05 93.00 factured 100.00 100.00 articles elsewhere specified 9 precious and 6.11 2.39 0.60 Mineral stones 94.10 89.35 97.30 products precious metals 100.00 100.00 and y 8 2.50 0.45 9.22 0.00 54.76 96.60 Electric 100.00 100.00 machiner y 7 0.00 1.39 0.59 14.80 47.82 79.80 100.00 100.00 Nonelectric machiner 0.27 88.81 27.50 51.23 80.80 27.50 42.62 ransportT 100.00 equipment 56 and 0.29 3.81 0.37 photo- graphic 97.00 62.25 93.90 Chemicals supplies 100.00 100.00 4 0.17 0.00 0.00 Metals 99.00 74.58 94.00 100.00 100.00 , , , 3 travel 2.29 0.00 0.00 Leather rubber goods 74.20 60.00 80.00 footwear 100.00 100.00 and extilesT and 0.24 6.65 1.34 0.62 clothing 99.64 64.40 93.20 100.00 12 and, ood, W pulp, 0.82 0.54 0.00 83.00 64.77 86.70 paper furniture 100.00 100.00 Base. Data (continued) Union Inegrated African y Europe 54.2 WTO Customs nisia ce: Cameroon Chad Senegal South Tu Zimbabwe Hungar Poland ableT Region Africa Eastern Sour 544 Binding Tariffs: Why Do It? other rules-based constraints on protection can be protection. Because it eliminates the highest rates of analyzed qualitatively in a similar way. protection, which are the ones that inflict the great- We begin by representing the underlying distribu- est costs on the economy, even a tariff binding that tion of protection in the absence of a tariff binding is substantially above the average rate of protection by a distribution such as that depicted in Figure can greatly reduce the cost of protection, as is 54.1. In general terms, we motivate our representa- shown in the next section. tion by the political process, which leads to uncer- tainty about the pattern of protection over time. In The Welfare Implications of Bindings this context, there are numerous political models that can be invoked here to drive the underlying What are the welfare implications of bindings? To probabilities related to a particular government ori- explore this issue, we use Figure 54.2 for the case of entation toward trade (see, for example, Stahl and a small country. (See Francois and Martin 2001 for a Turunen-Red 1995). What is important for the more formal treatment.) In the figure, the line present discussion is not the choice of a particular labeled "Import demand" represents the compen- political submodel but the resulting characteriza- sated aggregate import demand curve. With a fixed tion of trade policy as subject to uncertainty. For- tariff, the welfare cost of protection is defined by the mally, therefore, we simply assume that the expected Harberger triangle cab under the excess demand level of the tariff is m0 in the absence of a binding curve. Alternatively, consider symmetric variations on the tariff rate applied and that the distribution of around this tariff level, with a higher tariff yielding protection can be characterized by its mean and a higher domestic price, Ph, in one time period and variance. a lower one, PL, in another. The welfare cost of the Under these circumstances, a tariff binding higher tariff is cfg, and that of the lower tariff is cde. reduces both the average rate of protection and the The reader can verify that the average of these areas variability of protection. The exact relationship (the expected cost of protection) exceeds the cost of between changes in tariff bindings and actual pro- protection under a fixed tariff. Clearly, therefore, the tection depends on a number of factors, including expected cost of protection for a given average tariff the underlying variability of protection and the gap level is higher in our example than when that tariff between the tariff binding and the average rate of is certain. This benefit of bindings is separate from 54.1 Implications of a Tariff Binding for the Applied Rate of Protection 545 T R A D E P O L I C I E S Q U E S T I O N S A N D G U I D E L I N E S 54.2 Welfare Implications of Tariff Bindings for a Small Country f Ph a P d Import demand Pi c P* g b e 0 Imports the investment-related benefits of reduced uncer- What are the implications of price stability tainty (discussed briefly below). It follows from the through bindings? Clearly, if market access can be geometric aspect of the welfare costs of price distor- secured at the lower level of protection, PL, the move tions. The expected benefits of reduction in uncer- is welfare improving. As the current example of tainty will be magnified, for small countries, when China's accession to the WTO and the countersitua- preferences reflect risk aversion. tion of unsecured MFN treatment in the U.S. market By similar arguments, benefits related to secured have highlighted, secure MFN access (that is, secured market access conditions in export markets can be access at the "best available rate") is better than unse- identified. These benefits depend critically on the cured access. Consider also, however, a stabilized level nature of the security. Consider again a small of protection at the mean level P. In the present exporter, with the excess import demand curve example, if we compare the welfare effects of the var- again represented in Figure 54.2. We now assume ied states (in terms of shifting terms of trade effects) away home import policy variance and focus with the fixed state P, in the absence of risk aversion, instead on uncertainty in foreign market access variable terms of trade are preferred. The reader can conditions. Free trade is represented by price line verify this by adding the relevant squares and trian- P*. We again assume symmetric variations in pro- gles under the excess demand curve. Again, this is tection, this time as reflected in market access con- analogous to well-known results in the price stabi- ditions for exports. lization literature, this time for demand agents (that Exports are determined by the terms of trade and is, the importer). The actual welfare implications of by import demand, as reflected in the intersection bindings on the part of trading partners will depend of the world price line for importables. If protection critically on the elasticity of demand, possibilities for in export markets is low, terms of trade are relative- consumption smoothing, and relative risk aversion. ly favorable, and trade occurs along world price line Commercial policy stability in both import and PL. Alternatively, with high protection in export export markets is analogous to commodity price markets, terms of trade are given by Ph. The welfare stabilization. For a small country, the benefits follow costs of these two states, compared with free trade, from the imposition of own-security. There may are PhP*cf and PLP*cd, respectively. also be benefits from foreign market access security, 546 Binding Tariffs: Why Do It? but whether this happens hinges on the nature of key feature of multilateral negotiations is the focus commercial policy security on the export market on tariff bindings. Since the inception of the GATT side (and the rents that may be generated in export system in 1947, trade liberalization under the GATT markets), as well as on the relative risk aversion of has been based on the introduction of tariff bind- home economic agents. Thus, the national welfare ings that have increasingly constrained the range benefits of reducing commercial policy uncertainty and variability of protection rates. The ramifica- are much more evident for securing one's own poli- tions of this process are clearest for the OECD cies than for securing partners' policies. countries, where tariffs on industrial products are Yet another set of effects is likely to follow from a now relatively stable, at rates dramatically below the generally improved commercial policy environment levels observed through the mid-20th century. The (see Francois 1997). In the case of Mexico this is process is not as far along, however, in the case of emphasized by Kehoe (1999) and by Young and agriculture, where effective GATT disciplines are Romero (1994), who stress the linkages between the new, or in that of industrial tariffs of developing terms attached to foreign capital and the conditions countries. for access of the domestic economy to the global For many developing countries, tariff bindings on economy. Basically, the conditions for international industrial tariffs are still well above applied rates capital lending reflect a number of factors, includ- (and the same is true for almost all countries in the ing the security provided by outside obligations case of sensitive agricultural products). High tariff (such as Mexico's accession to the GATT in 1986 bindings allow substantial room for applied tariff and to the North American Free Trade Agreement rates to vary below the level of the binding. Even so, in 1993). As elements are added that reduce the bindings may reduce both the average applied tariff underlying sources of commercial policy uncertain- and the variability of the applied rate of protection. ty, the price of capital for a given country on world In this sense, bindings increase the security sur- markets (inclusive of risk premiums) shifts accord- rounding market access conditions. While the ingly. Arguably, when we view bindings at a macro importance of the security and certainty of market level, this effect may be the most important vis-à-vis access has long been recognized in the policy the expected costs of protection discussed above. It process, little attention has been devoted to these serves to reduce uncertainty for foreign investors issues in analyses of the benefits of WTO negotia- about the ability of an economy to link itself with tions. These analyses are therefore likely to under- the global economy and hence to generate returns state the gains from binding tariff commitments that can ultimately be repatriated. under the WTO. Conclusions Note This chapter has examined the issue of tariff bind- 1 A more formal examination of the relationship between bound ings and their relationship to applied tariff rates. A and applied rates is found in Francois and Martin (2001). 547 55 B E R N A R D H O E K M A N M A U R I C E S C H I F F Benefiting from ments as instruments of foreign policy and, sometimes, nation- building, the WTO does not Regional prohibit RIAs. Instead, it im- poses on members disciplines Integration designed to minimize "oppor- tunistic" behavior that is aimed primarily not at integration but at discrimination against non- members. This is done by requiring that WTO members notify new RIAs to the WTO, that they do not raise barriers to T trade against the rest of the he world has witnessed a veritable world, and, most important, that they eliminate explosion of regional integration barriers to trade on substantially all intraregional agreements (RIAs) in the past 15 years (see Box trade in goods.1 The last requirement is intended to 55.1). More than half of world trade now occurs ensure that the RIA has integration as its goal. within actual or prospective trading blocs, and The WTO disciplines can be regarded as rules of nearly every country in the world is a member of thumb intended to minimize the negative implica- one or more RIAs. RIAs take many forms. The most tions of regional integration for the multilateral trad- common are the free trade area (FTA)--where trade ing system. For historical reasons related to the restrictions among member countries are removed creation of the European Economic Community in but each member maintains its own trade policies the 1950s, GATT/WTO rules on regionalism have toward nonmembers--and the customs union, an never been effectively enforced. The Doha Develop- FTA whose members adopt a common external ment Agenda of 2001 launched negotiations on the trade policy. Deeper forms of integration include rules applying to regionalism. The issues that arise in common markets--customs unions that also allow this context, which are discussed extensively in the for the free movement of factors of production-- literature, include technical trade policy measures and economic unions, which involve some degree of such as the use of rules of origin (discussed elsewhere harmonization of national economic policies. in this volume) and the procedures used to ascertain Because, by definition, RIAs involve discrimina- whether individual RIAs satisfy WTO rules. tion against nonmembers (they result in preferen- This chapter takes the perspective of an individ- tial liberalization among partner countries only), ual country and asks what types of RIAs are likely to they are inconsistent with the most-favored-nation be most beneficial to developing nations. To a large (MFN) rule, the fundamental principle of the WTO. extent, this depends on the form and coverage of the Given the prevalence of regional trade arrange- RIA and on the identity of the partner countries. ments and the historical importance of trade agree- Since many, if not most, developing countries pur- 548 Benefiting from Regional Integration B O X 5 5 . 1 S E L E C T E D M A J O R R E G I O N A L I N T E G R AT I O N A G R E E M E N T S A N D D AT E S O F F O R M AT I O N Europe Caribbean Community but not of the Common European Union (EU). Formerly European Eco- Market). nomic Community (EEC), European Community (EC). 1957, Belgium, France, Germany, Italy, Lux- Africa embourg, Netherlands; 1973, Denmark, Ireland, Cross-Border Initiative (CBI). 1992, Burundi, United Kingdom; 1981, Greece; 1986, Portugal, Comoros, Kenya, Madagascar, Malawi, Mauritius, Spain; 1995, Austria, Finland, Sweden. Namibia, Rwanda, Seychelles, Swaziland, Tanza- European Economic Area (EEA). 1994, EU, Ice- nia, Uganda, Zambia, Zimbabwe. land, Liechtenstein, Norway. East African Cooperation (EAC). 1967 (formerly Euro-Mediterranean Economic Area (Euro- East African Community, EAC; broke up in 1977 Maghreb). Bilateral agreements: 1995, with and recently revived), Kenya, Tanzania, Uganda. Tunisia; 1996, with Morocco. Economic and Monetary Community of Central EU Bilateral Agreements with Eastern Europe. Africa (CEMAC). 1994 (formerly Union Douanière 1994, with Hungary, Poland; 1995, with Bulgaria, et Economique de l'Afrique Centrale, UDEAC); Czech Republic, Estonia, Latvia, Lithuania, Roma- 1966, Cameroon, Central African Republic, Chad, nia, Slovak Republic, Slovenia. Congo, Gabon; 1989, Equatorial Guinea. Economic Community of West African States North America (ECOWAS). 1975, Benin, Burkina Faso, Cape Canada-U.S. Free Trade Area (CUSFTA). 1988, Verde, Côte d'Ivoire, The Gambia, Ghana, Canada, United States. Guinea, Guinea-Bissau, Liberia, Mali, Mauritania, North American Free Trade Area (NAFTA). Niger, Nigeria, Senegal, Sierra Leone, Togo. 1994, Canada, Mexico, United States. Common Market for Eastern and Southern Africa Latin America and the Caribbean (COMESA). 1993, Angola, Burundi, Comoros, Dji- Andean Pact. 1969 (revived in 1991), Bolivia, bouti, Egypt, Ethiopia, Kenya, Lesotho, Malawi, Colombia, Ecuador, Peru, the República Boliviari- Mauritius, Mozambique, Rwanda, Somalia, Sudan, ana de Venezuela. Swaziland, Uganda, Zambia, Zimbabwe. Central American Common Market (CACM). Indian Ocean Commission (IOC). 1984, 1960 (revived in 1993), El Salvador, Guatemala, Comoros, Madagascar, Mauritius, Seychelles. Honduras, Nicaragua; 1962, Costa Rica. Southern African Development Community Common Market of the South/Mercado (SADC). 1980 (formerly known as the Southern Común del Sur (MERCOSUR). 1991, Argentina, African Development Coordination Conference, Brazil, Paraguay, Uruguay. SADCC), Angola, Botswana, Lesotho, Malawi, Group of Three (G3). 1995, Colombia, Mexico, Mozambique, Swaziland, Tanzania, Zambia, Zim- República Boliviariana de Venezuela. babwe; 1990, Namibia; 1994, South Africa; Latin American Integration Association 1995, Mauritius; 1998, Democratic Republic of (LAIA). Formerly Latin American Free Trade Area the Congo, Seychelles. (LAFTA), 1960 (revived in 1980), Argentina, Economic Community of West Africa (CEAO). Bolivia, Brazil, Chile, Colombia, Ecuador, Mexico, 1973 (revived in 1994 as UEMOA), Benin, Burkina Paraguay, Peru, Uruguay, República Boliviariana Faso, Côte d'Ivoire, Mali, Mauritania, Niger, Senegal. de Venezuela. West African Economic and Monetary Union Caribbean Community and Common Market (UEMOA, or WAEMU). 1994, Benin, Burkina (CARICOM). 1973, Antigua and Barbuda, Barba- Faso, Côte d'Ivoire, Mali, Niger, Senegal, Togo; dos, Jamaica, St. Kitts and Nevis, Trinidad and 1997, Guinea-Bissau. Tobago; 1974, Belize, Dominica, Grenada, Southern African Customs Union (SACU). 1910, Montserrat, St. Lucia, St. Vincent and the Botswana, Lesotho, Namibia, South Africa, Grenadines; 1983, The Bahamas (part of the Swaziland. (continued) 549 T R A D E P O L I C I E S Q U E S T I O N S A N D G U I D E L I N E S B O X 5 5 . 1 ( C O N T I N U E D ) Economic Community of the Countries of the 1995, Vietnam; 1997, Myanmar, Lao People's Great Lakes (CEPGL). 1976, Burundi, Rwanda, Democratic Republic; 1999, Cambodia. (ASEAN Democratic Republic of the Congo. members created a free trade area in 1992.) Greater Arab Free Trade Area (GAFTA). 1998, Middle East and Asia Bahrain, Egypt, Iraq, Jordan, Kuwait, Lebanon, Asia-Pacific Economic Cooperation (APEC). Libya, Morocco, Oman, Qatar, Saudi Arabia, 1989, Australia, Brunei Darussalam, Canada, Sudan, Syria, Tunisia, United Arab Emirates, Indonesia, Japan, Republic of Korea, Malaysia, Yemen. New Zealand, Philippines, Singapore, Thailand, Gulf Cooperation Council (GCC). 1981, Bahrain, United States; 1991, People's Republic of China, Kuwait, Oman, Qatar, Saudi Arabia, United Arab Hong Kong (China), Taiwan (China); 1993, Mex- Emirates. ico, Papua New Guinea; 1994, Chile; 1998, Peru, South Asian Association for Regional Coopera- Russian Federation, Vietnam. tion (SAARC). 1985, Bangladesh, Bhutan, India, Association of Southeast Asian Nations Maldives, Nepal, Pakistan, Sri Lanka. (ASEAN). 1967, Indonesia, Malaysia, Philippines, Singapore, Thailand; 1984, Brunei Darussalam; Source: World Bank (2001b). sue regional integration strategies in parallel with Trade diversion takes place when imports from the WTO membership and multilateral negotiations, ROW are replaced in Country 1 by more expensive the question of how such parallel paths can be best imports from Country 2. Why would Country 1 pursued is also addressed. import more expensive goods from Country 2? Because goods from Country 2 do not pay the import tariff, while ROW goods do. Trade diversion Economic Aspects is typically harmful. The development impacts of RIA membership Transfers occur between member countries of the depend importantly on the countries that are trade bloc because removal of tariffs between them involved, the type of agreement, and its substantive means that exports obtain better prices in the part- coverage. Recent research suggests that of particular ners' markets (a positive transfer), while the costs of importance to developing countries are whether imports net of tariffs increase (a negative transfer). large industrial countries are members of the agree- ment, the extent of sectoral exclusions, and the Trade Creation and Diversion and the Type of RIA degree to which the RIAs involve "deeper integra- tion"--that is, extend beyond preferential elimina- Assume that Mexico imports either from the United tion of barriers to trade in goods. In order to States or from Japan and applies a 20 percent tariff understand the economic effects of RIAs, a few def- on all imports. For a given product, the United initions are in order. States sells at a price of $100 and Japan sells at a Trade creation takes place when a member country price of $110. With the 20 percent tariff, the cost to of the RIA (Country 1) increases its imports from its the Mexican consumer is $120 for the good from partner country (Country 2) without a reduction in the United States and $132 for the product from Country 1's imports from the rest of the world Japan. Consumers import from the cheaper (ROW). This occurs because with the removal of source--from the United States, not Japan. Now tariffs between member countries, Country 2's assume that Mexico forms an RIA with the United products become cheaper than those of Country 1. States, along the lines of the North American Free The increase in (cheaper) imports results in an Trade Agreement (NAFTA). Tariffs on imports from increase in consumption and a reduction in output the United States fall to zero, and consumers now in Country 1. Trade creation is beneficial. pay $100. This is a case of trade creation because the 550 Benefiting from Regional Integration United States was already the cheapest source and RIA is ambiguous a priori and depends on whether has become even cheaper, replacing more expensive trade creation or trade diversion dominates. Of domestic output and raising consumption as well. course, even an RIA that is dominated by trade Note that the consumer price must fall for trade cre- diversion and has a negative welfare effect can be ation to take place. Since the United States was the turned into a beneficial RIA by lowering external only source of Mexican imports to begin with, hav- trade barriers sufficiently. ing an RIA with the United States is identical to uni- In the example above, imports were from either lateral trade liberalization by Mexico and is Japan or the United States but not from both. This is beneficial. relevant for a small country facing imports of a sim- Assume now that the good is exported by the ilar good from two large exporters. Countries, how- United States at $110 and by Japan at $100. With the ever, may import the good from more than one 20 percent tariff, the cost to consumers is $120 for source; this is particularly likely to be the case when the good from Japan and $132 for that from the a small country imports from another small coun- United States. The good is imported exclusively try and from the ROW. For instance, assume that from Japan at $120. After Mexico forms an RIA with Mali imports from Côte d'Ivoire and from the the United States, the cost of imports from the Unit- ROW. Say that the world price for a given product is ed States is $110, which is lower than the $120 con- $100 and Mali charges an import tariff of 20 per- sumer cost for imports from Japan. Consumers cent, or $20. Thus, the import price paid by con- switch their source of imports from Japan to the sumers in Mali (and faced by its producers) is $120. United States. But although the United States is the Côte d'Ivoire's supply of exports to Mali is not hori- cheapest source for consumers, it is not the cheapest zontal at $120, as is the case for the ROW. Rather, it source for Mexico as a whole. The reason is that is upward sloping: an increase in Côte d'Ivoire's although the consumer cost of imports from Japan supply of exports to Mali implies a higher (margin- is $120, the true cost for Mexico is only $100. The al) cost of production because Mali is a large buyer $100 is the cost in scarce foreign exchange that Mex- of Côte d'Ivoire's exports (although it is a small ico pays for imports from Japan, while the addition- buyer of ROW exports). In equilibrium, Côte al $20 associated with the 20 percent tariff is simply d'Ivoire also sells at $120 because it cannot sell at a a transfer from consumers to the government and is price above that of the ROW. (The analysis that fol- not a cost to the nation. To put it differently, before lows is presented graphically in the appendix to this entering the RIA, Mexico paid $100 to the foreign chapter.) supplier, and after entering the RIA, it pays $110. Assume now that Mali forms an RIA with Côte Thus, Mexico loses $10 per unit of import. This is d'Ivoire. As long as Mali continues to import from the cost of trade diversion from a cheaper to a more the ROW, the consumer price remains $120. Since expensive source. Another way of showing this the consumer price remains unchanged, no trade result is that the consumer price fell by $10 (from creation takes place. Before the RIA, the Mali gov- $120 to $110), but the government lost $20 in tariff ernment obtained $20 in tariff revenue for each unit revenue, so the net loss is $10. imported from Côte d'Ivoire, and Côte d'Ivoire Note that trade diversion occurred because the 20 exporters obtained $100. After the RIA, Mali no percent tariff is higher than the 10 percent differ- longer charges a tariff on Côte d'Ivoire imports and ence in cost between the U.S. and the Japanese Côte d'Ivoire obtains $120 for its exports. In other good. If the tariff were lower--say, 5 percent--the words, Mali now loses tariff revenue and pays $20 cost of Japan's imports would be $105, still lower more for its imports from Côte d'Ivoire. This than the U.S. price of $110 free of tariff. Consumers amounts to a transfer of income from Mali, the would have continued to import from Japan, the importing country, to Côte d'Ivoire, the exporting cheapest source, and trade diversion would not have country. Note that this occurs even if the quantity of occurred. This example points to an important les- imports from Côte d'Ivoire remains unchanged son: RIAs should lower their external trade barriers (that is, even if Côte d'Ivoire's supply of exports is because that reduces the extent of trade diversion. vertical or perfectly inelastic), in which case there is Another reason for so doing is that reducing exter- no trade diversion.2 nal trade barriers makes possible the classical gains It is likely, however, that Côte d'Ivoire's exports to from trade. Note also that the welfare impact of an Mali will increase after Mali's import tariff is 551 T R A D E P O L I C I E S Q U E S T I O N S A N D G U I D E L I N E S removed. In that case the loss of tariff revenue for The foregoing suggests at least three reasons why Mali will be larger. The increase in Côte d'Ivoire members of RIAs should lower external trade barri- exports amounts to trade diversion because total ers: (a) to generate classical gains from trade; (b) to imports are not affected by the RIA. Since the con- lessen the chances that trade diversion will occur; sumer price is unchanged, the increase in imports and (c) to reduce income transfers between mem- from Côte d'Ivoire comes at the expense of reduced ber countries resulting from the RIA and the ten- imports from the ROW. The (marginal) production sions that can arise from such transfers. cost of the increased imports from Côte d'Ivoire rises from $100 to $120, which is higher than the Political-Economy Issues $100 price paid for ROW imports. Since there is only trade diversion and no trade creation, forming Politically sustainable RIAs (that is, those acceptable such an RIA lowers the income of the bloc as a to interest groups) are likely to be trade-diverting, whole. Côte d'Ivoire gains by obtaining a better not trade-creating (Hirschman 1981a). The reason price in Mali's market, and Mali loses--and it loses is simple: trade creation occurs when domestic pro- more than Côte d'Ivoire gains. The reason is that duction is replaced by cheaper imports from part- Côte d'Ivoire's additional transfers are obtained by ner countries. Although this is beneficial for the increasing its exports to Mali, and the increased economy as a whole, since it reduces the output of exports are more expensive to produce than the inefficient industries, it is damaging for domestic ROW imports they replace. The difference between producers, who will therefore resist it. Trade diver- Côte d'Ivoire's gain and Mali's loss is the net loss sion, by contrast, means that imports from the due to trade diversion. It is an inefficiency loss and ROW are replaced by imports from the partner is not captured by anybody. country, which is less damaging for the domestic Developing countries are more likely to lose by industry. Consequently, resistance will be strongest forming a South-South RIA, as opposed to joining a against preferential liberalization that results in North-South agreement, because the South-South trade creation. A corollary is that RIA formation arrangement entails little or no beneficial trade cre- will lead to requests for exceptions. The outcome ation. Moreover, there is a high probability that one may be an exchange of exceptions, whereby the of the members of a South-South RIA may gain weakest industries in the various countries are while the other loses. It is the poorest member coun- excepted from the preferential liberalization. This tries that are likely to lose the most, as the most would be harmful for the member countries advanced country will usually have a more developed because preferential liberalization would mostly manufacturing sector. Its manufactures are typically result in trade diversion. Exceptions may be sectoral highly protected, and once the RIA is formed, the (agriculture is often excluded from preferential lib- more advanced country can export them free of tar- eralization) or may take the form of continued use iffs to the poorer member countries. The RIA thus of instruments of contingent protection or highly entails an income transfer from the less to the more restrictive rules of origin. Such exceptions tend to advanced member countries. Moreover, because the strengthen vested interests that oppose further lib- poorer country had more to gain from trading with eralization. As Jagdish Bhagwati has noted, these the North (relative to the more advanced country), it groups may argue that "the region is our market" will lose more from creating a South-South RIA. A and that "our markets are large enough." South-South RIA discriminates against the North Thus, in negotiating a RIA, it is important to ensure and so helps those countries within the bloc that are that member countries do not trade exceptions but, the closest competitors with the North. Such asym- rather, minimize the degree to which exceptions are metric distribution of the gains and losses of integra- granted. Accordingly, one of the primary conditions tion has led to tensions between members, breakup imposed by the WTO on members that form RIAs is of RIAs, and even wars. One way to reduce the asym- that RIAs must liberalize substantially all trade metry in the distribution of gains and losses, and between members of the agreement. The other major thus to lessen potential tensions between member condition is that the level of external protection not countries, is to lower the external trade barriers. The be increased during the formation of the RIA. lower the tariff levels with the ROW, the smaller the The political-economy dynamics that determine transfers associated with an RIA. the external trade policy which emerges are likely to 552 Benefiting from Regional Integration be complex and to depend on the type of RIA that is emphasized that this advantage of FTAs over cus- adopted. In the case of customs unions, a variety of toms unions may be offset by other characteristics effects have been identified (de Melo, Panagariya, of FTAs. The primary example is the need for rules and Rodrik 1993). One is a preference-dilution of origin, which may be used by import-competing effect: because the region implies a larger political industries both to restrict significantly internal lib- community, each of the politically important inter- eralization and to increase the effective protection est groups in member countries will have less influ- confronting nonmembers. ence on the design of common policies. Another is a preference-asymmetry effect: because preferences North-South versus South-South RIAs on specific issues are likely to differ across member countries, the resulting need for compromises may The static welfare effects presented above constitute increase the probability of more efficient outcomes. the basis of the economic analysis of RIAs. RIAs, The creation of a customs union may also disrupt however, can have other economic effects and the formation of rent-seeking interest groups, as objectives--for example, reduction of policy uncer- these would have to reorganize at the regional level, tainty and of the associated risk premiums demand- establishing an institutional structure that allows ed by investors; increased foreign direct investment them to agree on a common position. But customs (FDI) inflows; the creation of larger markets, allow- unions may also facilitate the adoption of less liber- ing beneficial procompetitive effects and economies al policies. Consumer interests may be harder to of scale; liberalization of service industries; and defend in an RIA than at the national level, whereas deeper integration through regulatory cooperation producer interests are more likely to be strength- and harmonization. Moreover, RIAs may be driven ened than weakened (Tumlir 1983). Each national by noneconomic objectives, such as improved gov- producer group may face less opposition when ernance or democracy, or reducing foreign policy seeking price-increasing policies and may indeed tensions with partners (cementing a peace). From find support from other producer groups in other the viewpoint of developing countries, North- countries that pursue their own interests. (The South RIAs are likely to be superior to South-South European Union's Common Agricultural Policy is RIAs for the attainment of such objectives. an example.) The need to strike compromises may Reduction in policy uncertainty (`increased credi- then result in a less liberal trade regime. Moreover, it bility') can be realized by locking in reforms as a may be in the interest of national politicians to let a result of binding commitments in the RIA. North- regional organization satisfy national pressure South RIAs can provide more credible lock-in groups, as this is less transparent for domestic vot- mechanisms and do more to improve policy credi- ers and can be justified as being necessary to main- bility. Whereas an RIA between two small develop- tain the agreement (Vaubel 1986). ing countries is unlikely to exercise much restraint The external trade policy bias toward protection on the member countries' behavior, commitments that may arise under a customs union will be weak- made with a large high-income country or region er in an FTA. Because there is no common external (such as the United States or the European Union) trade policy, member countries compete in their are likely to be more binding--or to be perceived as external trade policies. Import-competing firms in such. If so, the policy commitments may result in member countries must lobby their own govern- greater FDI and lower risk premiums. It must be ments. The required coordination and cooperation kept in mind, however, that RIAs mostly entail com- may be more difficult to sustain than in a customs mitments about trade policy, not macroeconomic union, where the centralization of trade policy policy or other aspects important to domestic or requires firms to present a common front. In any foreign investors. For instance, Greece joined the particular instance some member country govern- European Union in 1981, and Spain and Portugal ments will award protection and others will not. If joined in 1986. The latter two countries comple- industries in member countries are all competing mented accession with deep economic reforms and against third suppliers, protection by one member benefited from increased FDI and growth. Greece, may benefit industries in other member states. Such by contrast, ran large budget deficits, did not free-riding can result in less protection than in the reform, and did not obtain increased FDI or achieve absence of the FTA (Deardorff 1994). It should be higher growth. Thus, although an RIA with the 553 T R A D E P O L I C I E S Q U E S T I O N S A N D G U I D E L I N E S North may increase policy credibility, investment, framework within which Franco-German wars and growth, this is unlikely to happen in the absence would no longer be possible. With this motive in of complementary domestic policy reforms. mind, the Economic Coal and Steel Community The benefits from RIAs in terms of larger mar- (ECSC) was founded in 1951, and the European kets, greater competition, and the realization of Economic Community (EEC) in 1957. European economies of scale will depend in part on the integration has continued to deepen and expand extent to which the RIA involves deeper integra- over the years. Other RIAs based in part on political tion; that is, to what degree it extends to services objectives include the Common Market of the markets and regulatory regimes that determine the South (MERCOSUR) and the Association of South- conditions of competition prevailing in the region- east Asian Nations (ASEAN). al market. Since industrial partner institutions are An RIA can enhance security because it increases generally superior to those found in developing the level of trade between member countries and, in countries, a North-South RIA is likely to provide so doing, increases familiarity between the people of more benefits from deep integration.3 However, in the member countries and lessens the degree of the case of services industries and services sector misconceptions. Increased economic integration regulation, as well as institutions that protect com- also makes wars more costly. Thus, security issues petition and property rights, a good case may exist provide a rationale for discriminating against non- for regional harmonization and cooperation members and limiting trade preferences to member between neighboring countries. Thus, South-South countries. If this is the case, external tariffs should agreements can also be effective mechanisms for be chosen optimally in order to maximize the secu- regulatory reform and cooperation and may gener- rity objective. The optimal level of external tariffs ate more appropriate standards. Examples include will fall over time, especially if the RIA results in common competition, regulatory, or intellectual deeper integration. The reason is that with the for- property agencies (such as patent offices)--all mation of the RIA, trade between member coun- institutions with the potential for economies of tries increases, and so does security. As the level of scale or scope if a regional approach is taken, as security rises, the security benefits from additional opposed to the creation of multiple regulatory trade decrease, and so do the optimal tariffs. The agencies in individual countries. But this does not same dynamics arise if the RIA involves deep inte- normally require a preferential trade arrangement. gration--this type of cooperation will also increase Two final arguments in favor of North-South security and, over time, reduce the optimal level of RIAs can be mentioned. First, endowment differ- external protection. ences are usually larger between members of a Other RIAs with a "security" dimension are those North-South RIA than between members of a that may have been created as a form of protection South-South RIA. Developing countries are there- against a regional hegemon. An example is the Gulf fore likely to exploit their comparative advantages Cooperation Council (GCC), established in part to better in a North-South RIA than in a South-South counterbalance regional powers (Iran and Iraq). one. Second, a crucial determinant of long-term One reason why Central and Eastern European growth is the absorption of knowledge and technol- countries wanted to conclude free trade agreements ogy. Since these assets are mainly produced in high- with the European Union was a desire to reduce the income countries, liberalization that occurs in the possible exercise of Russian hegemony. Latin Amer- context of a North-South RIA is likely to generate ica provides another example: in 1996 a rumor of a more growth than opening up through a South- coup in Paraguay led to a joint statement by the South RIA. presidents of the four MERCOSUR countries that democratic institutions were necessary for main- taining membership in the group. This statement is Political Aspects said to have eliminated the threat. Democracy is an Many RIAs have been driven by political rather than explicit condition of the association agreements economic goals. These political objectives include between MERCOSUR and both Chile and Bolivia. security, governance, democracy, and human rights. RIAs can therefore be vehicles for pursuing The clearest example is that of European integra- democracy and governance objectives. Using RIAs tion, whose founding fathers sought to create a to lock in changes in political institutions is likely to 554 Benefiting from Regional Integration be more effective in North-South than in South- relation to that of West Pakistan. Finally, war South RIAs. The reasons include the demonstrated ensued, and Bangladesh became an independent track record and the high value placed on democra- nation. cy and human rights in most high-income coun- Such tensions and potential conflicts can be min- tries; indeed, high-income nations may make entry imized by lowering external trade barriers. This can into the club conditional on the reform of political be done independently in a free trade agreement institutions. The likelihood of effective enforcement but not in a customs union, where the country that of governance criteria is also higher. In the case of benefits from the large transfers may not want to North-South agreements the costs of a breakdown agree to a reduction in the common external tariff. of cooperation will be higher for developing coun- try members than under a South-South agreement, WTO Rules as the preferential access to industrial markets is worth more. In addition, there may be financial and As mentioned above, the WTO imposes several dis- technical assistance transfers associated with North- ciplines on RIAs between member countries, South agreements that are tied to the pursuit and requiring that agreements liberalize substantially all attainment of governance objectives. trade and do not result in higher external barriers. Although an RIA can generate political and secu- Despite the creation of a single Committee on rity benefits, especially when these were part of the Regional Trade Agreements (CRTA) to review RIAs objectives for setting up the RIA, it may also worsen notified to the WTO, a major problem has been security. This is most likely to happen where the dis- inability to achieve a consensus in the WTO on tribution of transfers is asymmetric. For instance, in whether specific RIAs comply with WTO rules. The the United States in the 19th century, the North was reason for this is that the CRTA, like all WTO bod- more developed than the South. It produced manu- ies, operates by consensus. In addition, there has factures and "exported" them to the South, while been relatively little use of dispute settlement proce- the South exported agricultural products. The dures to contest the operation or design of RIAs, North dominated the Congress in the 1830s and although a 1996 case brought by India suggests that was able to pass a law raising tariffs on imports of there is scope to do so.4 manufactures, resulting in a large income transfer The Doha Development Agenda provides for the from the South to the North. This was called the launching of negotiations on WTO rules, including "tariff of abominations" in South Carolina, which those relating to preferential trading arrangements. refused to collect it. The issue played an important This raises the question as to what might be done to role in the genesis of the Civil War (Adams 1993). strengthen WTO disciplines. Economists some- In the 1960s Kenya, Tanzania, and Uganda times argue that a necessary condition for preferen- formed the East African Community (EAC). Kenya tial liberalization to be deemed multilaterally had a more developed manufacturing sector than acceptable is that the volume of imports by member the other two, and the RIA resulted in large income countries from the rest of the world not decline on a transfers from Tanzania and Uganda to Kenya. The product-by-product basis after the implementation EAC ultimately fell apart. Similarly, in the Central of the agreement. The evidence to date suggests that American Common Market (CACM) in the late although the intensity of intraregional trade 1960s, the more developed El Salvador benefited increased in the last century, the propensity of from income transfers from Honduras. Honduras regions to trade with the rest of the world, expressed asked to renegotiate its share of tariff revenues, but as a percentage of regional GDP, has also expanded. El Salvador refused, and Honduras left the CACM. Global integration, measured by trade flows and Another example is that of Pakistan before East capital flows, does not appear to have been affected Pakistan secured its independence (as Bangladesh) negatively by regional integration efforts since in 1971. West Pakistan was more developed and had World War II (World Bank 2000c). highly protected industries. East Pakistan had to pay No one knows, however, what would have hap- high protected prices for imports, but the revenues pened without RIAs. More important, the trade vol- were spent mostly on West Pakistan. Between 1948, ume test is a flawed one in that it does not guarantee when Pakistan became independent, and 1971, East that nonmembers will not be hurt by a RIA. A more Pakistan's income per capita fell by about a third in appropriate measure of the impact of a RIA on non- 555 T R A D E P O L I C I E S Q U E S T I O N S A N D G U I D E L I N E S members is to focus on what happens to the prices additional economic and political benefits from nonmembers get for their exports to RIA members reducing external trade barriers and minimizing after the agreement is formed. Chang and Winters exceptions to preferential trade liberalization. (2001) show that Brazil's membership in MERCO- Although a North-South RIA is likely to be superior SUR was accompanied by an improvement in its to a South-South RIA, unilateral trade liberalization external terms of trade--a fall in the relative price of may dominate both. One reason is that what can be imports from nonmember countries. This study obtained in a North-South RIA depends very much suggests that assessments of the impact of RIAs on the coverage and type of agreement that can be should take into account not just volume but also obtained. The greater are the exceptions and the terms of trade effects. greater is trade diversion, the lower is the payoff to RIAs impose costs on nonmembers even if exter- RIA membership. Unilateral liberalization does not nal levels of protection do not rise. Nonmember generate the payoff associated with better (and pref- suppliers become less competitive because they erential) access to partner countries; as Winters continue to pay tariffs while competing producers emphasizes in Chapter 5 of this volume, the gains from member countries do not. Where there are from trade reform are predominantly generated by economies of scale, RIAs may help lower member the country's own reform. Unilateral reform is also country firms' costs by expanding their home mar- under a country's own control and is not condition- ket. It is impossible to devise a realistic rule that will al on what can be negotiated with partners. Further- ensure that the trade policy stance of a RIA will be more, many countries (for example, in Central Asia welfare improving for members and the rest of the and South Asia) may not have the option of joining world (Winters 1999). What matters is that WTO a North-South RIA. Finally, unilateral liberalization members encourage RIAs to reduce external barri- in the regional context--lowering barriers against ers; this is the best approach toward reducing the the rest of the world--should be an important com- potential downside of RIAs for both members and plement to preferential liberalization to reduce the nonmembers alike. The types of tests and criteria potential costs of trade diversion. suggested by economists to limit the negative impli- Trade diversion may be less of an issue if RIAs cations of RIAs for nonmembers are arguably not include deep integration--that is, if they involve very useful for defining ex ante WTO disciplines. regulatory cooperation and services sector liberal- This is because those that are relevant (that have ization. If the RIA entails the adoption of improved teeth) are unlikely to be feasible. Examples include policies that are applied on a nondiscriminatory proposals to adopt disciplines with respect to pref- basis (to both foreign and domestic suppliers equal- erential rules of origin in the WTO, requirements ly), there will be no trade diversion. The same is true that RIAs be open to new members, or rules requir- if RIAs are used as vehicles to attain a reduction in ing that discrimination be eliminated (i.e., mem- real trade costs by eliminating policies that simply bers would have to extend free trade to the rest of raise the cost of doing business without generating the world) within a certain period of time (Srini- any government revenue, private rents, or social vasan 2001). Instead, the primary source of multi- benefits. (Examples include redundant inspection lateral discipline on RIAs is periodic rounds of or licensing schemes and duplicative testing negotiation aimed at reducing trade barriers. This requirements for goods.) Diversion costs also do could be complemented by strengthened multilat- not arise if the RIA is used to pursue joint regula- eral surveillance of the economic effects of RIAs by tion of particular activities, in the process allowing the WTO secretariat. the realization of economies of scale and scope. RIAs represent both an opportunity and a chal- lenge for the multilateral trading system. The Conclusion opportunity is to use them as experimental labora- Regionalism has become a major component of the tories for cooperation on issues that have not (yet) trade policy strategies of most countries. A large been addressed multilaterally or where there is a body of research--much of it summarized in World clear case for a regional approach (such as a region- Bank (2000c)--suggests that North-South RIAs are al patent office). The more that regional agreements generally preferable to South-South RIAs for devel- are applied on a nondiscriminatory basis, the less oping countries. Member countries will obtain likely it is that they will have detrimental effects on 556 Benefiting from Regional Integration nonmembers. The challenge is to control the dis- partner country equal Q1, and imports from the crimination that is inherent in preferential trade ROW equal Q3 ­ Q1, with total imports of Q3. agreements. The inability of the Committee on WH MFN = surplus ABF + tariff revenue BDEF and is Regional Trade Agreements to determine whether lower than WH under free trade by triangle BCD. RIAs satisfy WTO rules is a problem in this regard. Assume that the home country now forms an RIA It suggests that multilateral negotiations have to with the partner country. As the partner country no play a major complementary role by allowing stake- longer pays the tariff T, its export supply curve shifts holders who are negatively affected by RIAs to use to Sp. The ROW still pays the tariff T, so the home the WTO process to reduce discrimination. country price remains PW. Hence, the partner coun- try imports increase from Q1 to Q2, while the Appendix imports from the ROW fall from Q3 ­ Q1 to Q3 ­ Q2. This trade diversion results in a worsening in the Assume that a home country and a partner country home country's terms of trade. Welfare is WH RIA = form an RIA. Both are assumed to be small relative ABF + BDIG. In other words, the RIA has no impact to the rest of the world (ROW), so they take the on the consumer surplus because the price is not price of goods from the rest of the world, PW, as affected (there is no trade creation), but there is a given (Figure 55.A1). loss of tariff revenue. WH is lower than WH RIA MFN by DH represents the home country's demand for EFGI, the tariff revenue lost on imports from the imports, SP represents the partner's supply of partner country after forming the RIA. Note that the exports facing the home country, and SROW is the welfare loss to the home country would occur in the supply from the ROW. Under free trade, imports absence of trade diversion as well (for example, the equal Q4, and home country welfare WH = triangle loss would be EFGI if SP were vertical at level Q2). ACE. Assume now that the home country imposes The home country welfare loss from the RIA is an MFN tariff, T. Then the price of imports from proportional to the level of partner country the ROW faced by home country producers and imports. Note also that if the MFN tariff were lower consumers rises to PW= PW + T, and SROW shifts to than T, the welfare loss from the RIA would be SROW. Similarly, SP shifts to SP. Imports from the smaller than area EFGI both because of lower 55.A1 A RIA Involving Small Countries A S'P SP PW' = PW + T F B G S'ROW PW E C SROW J I D DH Q Q1 Q2 Q3 Q4 557 T R A D E P O L I C I E S Q U E S T I O N S A N D G U I D E L I N E S imports from the partner and because of the lower the absence or elimination of substantially all measures violat- tariff rate. ing national treatment in sectors where specific commitments were made in the GATS; and that they do not result in higher The partner gains EFGJ, which is less than the trade barriers against third countries. home country loss of EFGI by the triangle GIJ. The reason for the net loss of triangle GIJ for the RIA 2 Many analysts equate trade diversion with income transfers. As this example shows, they are not necessarily the same, and members as a whole is the trade diversion, Q2 ­ Q1. income transfers can occur even in the absence of trade diver- This amount was previously imported from the sion. ROW at a cost of PW but is now produced at a high- er marginal cost. 3 By deep integration is meant measures beyond the border that increase the degree of competitiveness of each of the economies with respect to suppliers from the other members. Notes This includes harmonization or mutual recognition of stan- dards and regulations (for example, regarding production, This chapter draws on the results of a World Bank research project sanitary standards, and so on). on regional integration, much of which was summarized in World 4 As noted by Hoekman and Kostecki (2001), a rare example was Bank (2000c), and on parts of chapter 10 of Hoekman and Kostec- a case brought by India against Turkey in 1996. India contested ki (2001). See both of these sources and the Website for additional references to the literature. products, which were required because Turkey had entered 1 Article XXIV of the GATT allows FTAs and customs unions if (a) into a customs union with the European Union. A WTO panel trade barriers after integration do not rise, on average (Article found that Turkey's measures were inconsistent with GATT Arti- XXIV.5); (b) all tariffs and other regulations of commerce are cles XI and XIII and rejected Turkey's assertion that its measures removed on substantially all intraregional exchanges of goods were justified by GATT Article XXIV. On appeal, the Appellate within a reasonable length of time (Article XXIV.8); and (c) the Body upheld the panel's conclusion on the illegality of the quo- arrangements are notified to the WTO Council. Article V GATS tas but found that the legal interpretation of Article XXIV by the similarly requires that integration agreements have substantial panel was erroneous; the Appellate Body stated that a panel sectoral coverage, in terms of the number of sectors, volume should first ascertain whether an RIA complies with Article XXIV of trade affected, and modes of supply; that they provide for before considering other GATT provisions. 558 A P P E N D I X E S 559 blank Appendix F R A N C I S N G A Patterns of Trade and Protection: Selected Country Data 561 7.7 9.0 1999 11.0 22.2 13.6 13.0 13.6 10.0 9.0 9.7 9.2 9.7 1998 24.2 13.5 23.8 12.6 11.0 14.6 17.6 31.1 11.0 16.8 1997 9.7 7.0 15.9 11.3 24.6 12.3 12.3 11.8 16.6 32.2 20.0 15.8 11.0 17.6 1996 4.0 9.8 9.7 11.2 32.0 27.4 13.1 11.1 32.2 35.0 18.1 24.1 15.8 23.6 1995 9.7 12.0 10.5 17.0 17.0 30.0 11.1 21.0 18.6 11.0 1994 9.8 2.7 42.0 11.9 18.4 11.0 36.3 1993 9.8 7.4 24.8 10.9 50.0 41.6 14.2 18.8 11.0 39.9 1992 22.9 11.8 71.0 40.0 10.0 21.2 17.9 11.0 42.9 1991 12.2 88.6 40.3 10.0 25.3 11.0 1990 3.0 20.5 22.0 20.0 42.0 16.0 32.2 15.0 40.3 1980­99 1989 24.6 15.0 25.0 94.0 37.8 17.0 35.0 15.0 Countries, 1988 7.1 23.8 27.0 19.0 41.0 36.9 15.0 102.2 1987 23.1 27.0 32.3 37.4 20.0 51.0 60.8 32.0 20.0 39.5 Industrial and 1986 27.0 12.0 23.3 81.8 17.3 17.3 47.5 20.0 51.0 37.0 32.0 20.0 38.1 1985 21.7 35.0 48.3 12.1 51.0 20.0 Developing 1984 22.6 49.0 35.0 for 1983 99.9 48.0 Rates ariffT 1982 28.0 48.0 49.5 1981 1.7 verageA 49.0 in 1980 44.4 29.8 44.0 37.9 28.3 cent) rendsT per The and Faso A.1 Darussalam erdeV African ableT (unweighted; economies Barbuda Rep. Economy Developing Albania Algeria Antigua Argentina Bahamas, Bahrain Bangladesh Barbados Belarus Belize Benin Bolivia Botswana Brazil Brunei Bulgaria Burkina Burundi Cambodia Cameroon Cape Central Chad Chile China 562 Dem. Arab 7.2 6.8 0.0 7.6 8.1 0.0 11.8 11.6 20.5 12.4 32.2 10.9 (continued) 8.0 8.4 6.9 9.0 5.7 0.0 9.3 8.4 0.0 9.5 11.6 15.7 11.3 26.8 20.6 16.4 10.4 13.3 30.0 9.9 7.3 8.0 0.0 8.5 0.0 11.7 17.6 19.2 10.7 10.9 14.5 11.3 35.5 13.6 11.4 14.3 35.0 7.7 9.2 0.1 0.0 11.7 17.6 21.4 10.7 13.4 11.3 12.4 20.4 13.5 12.5 11.4 14.3 15.2 38.7 13.2 5.0 5.5 9.7 0.0 13.3 11.2 20.0 11.9 15.0 17.8 12.3 28.1 10.2 16.3 18.6 15.0 12.0 17.0 10.0 41.0 4.8 0.0 8.5 11.5 34.1 20.6 11.2 22.0 17.8 11.9 28.3 10.1 28.8 17.5 10.8 47.8 4.8 9.3 0.0 8.5 11.5 11.7 23.6 13.2 13.1 17.0 10.8 47.8 19.4 0.0 11.8 15.0 53.0 20.0 0.0 21.1 10.1 42.2 17.0 12.6 79.2 20.3 5.3 0.0 27.0 24.7 25.8 10.4 28.0 17.0 25.0 13.0 81.8 20.6 8.2 0.0 27.3 20.7 30.6 10.3 37.1 33.5 16.0 17.0 16.0 20.0 25.2 0.0 27.3 25.0 11.7 29.6 17.0 6.2 0.0 29.4 22.4 16.4 23.3 12.1 31.9 28.0 42.8 21.1 23.0 27.2 22.8 11.6 15.0 98.8 20.7 6.2 8.9 0.0 33.6 22.4 32.0 21.1 26.0 29.7 37.7 42.8 23.0 20.0 17.4 15.0 31.5 100.0 0.0 21.1 26.4 30.0 24.0 27.0 0.0 61.0 23.8 26.6 30.0 24.0 37.0 0.0 26.3 17.1 30.0 0.0 27.2 43.3 27.7 0.0 23.6 28.7 47.4 29.0 74.3 0.0 30.8 76.4 29.0 20.7 Rep. Rep. The Rica Rep. y d'Ivoire Kong Rep. Rep. Salvador (China) Colombia Congo, Congo, Costa Côte Cuba Cyprus Czech Dominica Dominican Ecuador Egypt, El Estonia Ethiopia Fiji Gabon Gambia, Ghana Grenada Guatemala Guinea Guyana Haiti Honduras Hong Hungar India Indonesia Iran 563 7.5 8.7 8.7 5.2 9.8 7.6 1999 18.0 11.2 10.1 17.7 10.9 21.8 7.6 9.6 5.8 3.9 6.8 7.1 7.6 5.9 1998 11.1 15.7 16.7 19.0 13.3 16.3 23.4 46.5 1997 5.9 4.6 7.3 9.1 7.6 6.9 4.8 10.9 13.3 20.6 16.7 29.1 12.6 22.1 16.9 24.4 1996 6.0 5.0 7.3 8.7 6.1 5.7 9.5 6.0 13.5 13.4 21.5 25.3 16.7 28.9 12.6 15.6 24.4 18.3 41.7 1995 4.5 8.2 16.0 17.4 21.0 17.0 20.3 29.0 13.1 25.7 11.0 10.7 29.0 50.7 1994 5.7 19.3 19.9 13.0 31.4 29.0 13.5 24.4 17.4 51.0 1993 8.3 8.9 19.3 22.0 14.3 22.8 16.1 32.8 56.0 1992 5.5 20.5 33.6 10.1 12.8 13.4 34.4 61.1 1991 8.0 20.3 14.6 34.0 11.4 16.9 13.1 36.4 66.0 1990 2.5 12.2 43.7 13.3 27.0 16.3 33.0 11.1 35.7 64.8 1989 20.0 14.0 37.5 14.9 16.1 17.0 17.0 27.6 13.1 35.6 65.0 1988 4.2 6.1 3.5 18.3 41.7 18.9 15.2 13.0 19.2 11.3 24.0 22.6 33.7 69.0 1987 19.3 17.6 39.2 22.9 16.7 13.6 19.0 39.5 11.3 20.8 15.6 21.0 22.1 23.8 68.9 1986 7.1 6.0 2.9 17.3 17.2 39.2 18.3 25.5 15.8 23.1 41.7 22.6 23.0 66.0 1985 6.9 3.9 17.0 14.2 16.7 25.5 37.9 25.2 27.0 77.0 1984 7.5 14.5 41.7 21.9 19.6 23.6 35.8 23.0 33.0 35.0 77.0 1983 9.4 14.3 23.7 24.0 36.0 22.1 77.0 1982 8.0 16.0 13.8 40.3 23.7 27.0 54.0 32.6 77.6 1981 3.5 1.5 13.2 15.4 10.6 34.9 1980 15.9 13.3 21.9 (continued) A.1 Rep. ableT Economy Israel Jamaica Jordan Kenya Korea, Kuwait Latvia Lebanon Lesotho Libya Lithuania Madagascar Malawi Malaysia Mali Malta Mauritania Mauritius Mexico Mongolia Morocco Mozambique Myanmar Namibia Nepal Nicaragua Niger Nigeria Oman Pakistan 564 8.8 9.0 0.0 6.4 8.5 8.8 13.0 10.1 15.9 13.1 12.6 10.6 16.1 17.1 (continued) 9.2 0.0 6.9 7.2 9.2 9.7 9.2 9.5 11.2 13.2 10.7 19.8 13.9 21.6 20.1 19.5 9.2 0.4 8.1 8.7 12.8 20.5 13.3 13.4 13.1 12.6 20.0 15.1 21.8 13.3 9.3 0.4 8.8 9.7 10.0 20.7 14.3 18.7 10.9 13.0 12.3 10.6 21.0 24.4 13.2 9.3 6.0 0.4 8.0 6.2 20.0 11.6 12.2 21.0 20.0 30.0 11.2 24.5 23.1 15.0 8.0 8.5 0.4 9.7 16.3 21.7 11.5 12.5 12.6 39.5 26.0 24.0 27.5 23.3 8.5 7.3 0.4 17.6 22.6 34.8 12.3 10.0 24.2 27.5 45.6 0.4 9.9 15.4 18.0 24.3 11.7 12.3 12.1 25.0 33.0 0.4 15.9 17.0 26.0 11.7 16.3 42.0 10.5 26.9 37.8 5.0 0.4 9.7 26.0 27.8 15.0 11.0 28.3 43.0 11.0 29.7 39.8 0.4 9.7 42.0 27.6 18.3 12.2 15.0 28.2 40.8 8.0 0.3 10.9 46.0 27.9 10.0 41.0 23.2 12.7 27.3 14.8 12.6 29.8 3.7 0.3 14.2 45.0 27.9 10.9 33.0 12.6 25.8 22.0 27.3 12.9 12.0 19.4 4.2 46.0 27.9 10.9 16.7 15.0 25.8 30.8 17.3 56.6 22.8 32.1 46.0 27.6 31.0 40.0 14.8 26.5 41.2 10.9 42.0 28.8 29.0 30.8 0.3 31.0 29.5 41.3 31.0 21.0 31.4 25.8 14.8 31.0 23.9 1.8 2.5 17.0 34.6 35.0 32.3 1.9 11.2 19.0 41.4 50.6 Fed. and New Arabia Leone Rep. (China) Africa Lanka Kitts Lucia incentV Guinea Nevis Panama Papua Paraguay Peru Philippines Poland Qatar Romania Russian Rwanda Saudi Senegal Sierra Singapore Slovak Slovenia Somalia South Sri St. St. St. Sudan Suriname Swaziland Syria aiwanT anzaniaT Thailand ogoT 565 and Arab 8.2 4.6 5.0 4.6 5.6 4.0 5.2 3.8 3.3 0.0 4.8 1999 12.6 15.1 21.8 9.2 6.8 5.3 4.8 6.0 4.1 5.5 4.5 4.0 0.0 5.2 1998 29.9 12.7 10.0 12.2 12.0 22.2 1997 9.1 5.6 5.8 6.2 5.7 5.4 4.4 0.0 6.6 13.5 13.2 10.1 10.0 11.9 13.6 24.0 1996 4.0 9.5 9.0 6.1 6.4 6.7 4.1 5.9 7.2 5.5 0.0 5.8 12.8 20.0 13.6 25.0 1995 9.1 9.3 6.5 6.8 6.3 5.9 5.9 18.7 30.0 13.4 23.8 15.6 1994 9.6 8.6 7.6 6.3 4.1 27.0 17.1 14.7 11.8 12.7 26.9 17.2 1993 9.8 8.7 7.7 3.7 6.3 8.5 6.1 4.2 6.4 26.7 17.0 15.7 25.9 17.2 1992 6.3 8.7 6.3 18.7 27.5 18.2 16.4 11.0 10.4 1991 6.3 6.3 18.6 27.7 21.5 16.0 12.9 1990 8.8 8.7 6.3 27.4 23.0 19.0 1989 9.1 8.7 3.8 6.9 5.7 4.4 6.3 17.0 27.1 30.6 11.8 10.1 14.2 14.5 1988 8.7 7.1 6.0 6.6 17.3 26.0 22.7 27.5 32.9 1987 8.9 24.0 26.6 19.9 29.1 32.9 16.2 11.8 29.9 1986 4.5 8.7 6.0 24.5 31.4 30.0 40.0 11.8 1985 27.2 24.7 38.0 28.0 11.8 1984 8.5 27.8 22.0 28.0 1983 40.0 10.0 1982 26.4 11.0 1981 1.2 47.0 26.0 9.8 1980 23.8 12.0 (continued) FR Union Samoa A.1 de States Zealand B. way ableT Economy rinidadT obagoT unisiaT urkeyT Emirates Uganda Ukraine United Uruguay enezuela,V R. ietnamV estern W emenY ugoslavia,Y economies Zambia Zimbabwe Industrial Australia Canada European Iceland Japan New Nor Switzerland United 566 ariff 4.0 11.3 longer 1987, a Non-T and 4.4 over 13.1 ariffT Supplement, of available 5.0 13.7 are Countries, Indicators data 5.3 14.9 OECD, Developing 6.3 whichever of 16.1 rates, 1998­2000); 7.2 18.7 Measures (MFN) years, Control 6.8 19.4 radeT (various of 7.9 21.5 most-favored-nation Indicators or Handbook 8.5 24.3 AD, rates, UNCT Development 7.9 applied orld 23.2 or W rates, Bank, 8.2 1990­2000); 23.8 orld W valorem issues, 7.1 (1998). ad 23.4 in (1996); (various Data" goods 24.7 Reincke all Reports y Analysis for and 6.0 26.6 Ingco,, averages Quantitative 27.2 Review--Countr Finger and Policy 8.5 unweighted (1994); 29.7 on radeT "Statistics Riedel and based 30.5 and Bank, are (2000), Desai, 30.0 11.0 rates tariff Dean, Development All CD-ROM 23.1 Base 1994; s, 9.8 available. Data 27.6 Regime Inter-American not Import Integrated indicate of (1997b); y WTO, Blanks Barriers ces: Director verage,A developing economies (129) verage,A industrial economies (23) Note: period. Sour and radeT 567 D E V E L O P M E N T, T R A D E , A N D T H E W T O Table A.2 Average Tariff Rates by Sector and FDI Inflows for All Countries in Recent Years Tariff rate (unweighted, percent) Net FDI inflow (millions of U.S. dollars) All Agri- Manu- Region Economy Year goods culture factures 1997 1998 1999 1 Brunei 1994 2.7 0.0 3.2 5 4 5 1 China 1998 16.8 16.5 16.9 44,236 43,751 40,400 1 Fiji 1996 12.4 11.9 13.4 16 76 30 1 Hong Kong (China) 1998 0.0 0.0 0.0 11,368 14,776 23,068 1 Indonesia 1999 10.9 11.9 10.7 4,677 ­356 ­3,270 1 Korea, Rep. 1999 8.7 12.8 7.8 3,088 5,215 10,340 1 Malaysia 1997 7.1 6.0 7.5 6,513 2,700 3,523 1 Myanmar 1996 5.7 8.9 5.1 387 315 300 1 Papua New Guinea 1999 8.8 24.6 8.0 29 110 170 1 Philippines 1999 10.0 14.2 9.3 1,249 1,752 737 1 Singapore 1998 0.0 0.0 0.1 8,085 5,493 6,984 1 Solomon Islands 1998 22.7 32.8 22.3 34 9 15 1 Taiwan (China) 1999 8.8 16.0 7.1 2,248 222 2,926 1 Thailand 1999 17.1 32.1 14.6 3,732 7,449 6,078 1 Vietnam 1999 15.1 21.5 14.4 2,745 1,972 1,609 2 Bangladesh 1999 22.2 21.4 22.5 141 308 150 2 India 1999 32.2 30.5 32.4 3,577 2,635 2,168 2 Nepal 1999 17.7 12.9 18.9 23 12 132 2 Pakistan 1998 46.5 42.7 46.9 713 507 531 2 Sri Lanka 1997 20.0 23.8 19.1 435 206 202 3 Benin 1996 13.1 13.7 12.8 26 35 31 3 Botswana 1996 11.1 12.3 11.0 100 90 112 3 Burkina Faso 1998 31.1 37.0 29.1 13 10 10 3 Cameroon 1996 18.1 24.3 17.8 45 50 40 3 Central Africa Rep. 1997 7.0 7.6 6.8 6 5 13 3 Chad 1997 15.8 17.0 15.5 15 16 15 3 Congo, Rep. 1997 17.6 18.0 17.5 9 4 5 3 Côte d'Ivoire 1996 19.2 21.2 18.8 450 314 279 3 Gabon 1998 20.6 25.1 19.7 143 211 200 3 Ghana 1995 15.0 20.1 14.1 83 56 115 3 Guinea 1998 16.4 16.6 16.3 17 18 20 3 Kenya 1999 18.0 16.7 18.2 40 42 42 3 Madagascar 1998 6.8 6.4 6.9 14 16 58 3 Malawi 1998 15.7 15.6 15.7 22 70 60 3 Mali 1999 11.2 16.1 10.4 74 36 40 3 Mauritius 1998 19.0 14.9 19.5 55 12 49 3 Mozambique 1997 15.6 16.9 15.3 64 213 384 3 Nigeria 1998 23.4 23.0 24.0 1,539 1,051 1,400 3 Rwanda 1993 34.8 58.0 31.1 3 7 5 3 Senegal 1996 12.3 13.5 12.1 176 71 60 3 South Africa 1999 8.5 8.0 8.6 3,817 561 1,376 3 Tanzania 1999 16.1 17.4 16.2 158 172 183 3 Togo 1997 13.3 13.6 13.3 23 42 35 3 Uganda 1996 13.2 23.7 11.6 175 210 180 3 Zambia 1997 13.6 15.9 13.0 207 198 163 3 Zimbabwe 1998 22.2 27.0 21.7 135 444 59 568 Appendix A: Patterns of Trade and Protection: Selected Country Data 4 Algeria 1998 24.2 21.8 24.9 7 5 6 4 Bahrain 2000 7.7 7.5 7.8 329 181 300 4 Cyprus 1998 8.4 26.1 4.8 68 56 65 4 Egypt 1998 20.5 22.7 20.2 888 1,077 1,500 4 Israel 1998 7.6 19.9 5.7 1,622 1,850 2,256 4 Malta 1999 7.6 6.8 7.9 165 273 811 4 Morocco 1997 22.1 28.9 21.3 1,079 329 847 4 Oman 1997 4.8 4.0 4.9 53 106 70 4 Saudi Arabia 1999 12.6 12.2 12.6 3,044 4,289 4,800 4 Tunisia 1998 29.9 31.0 29.6 366 670 368 4 Turkey 1998 12.7 47.9 5.4 805 940 783 5 Belarus 1998 12.6 11.0 13.3 200 149 225 5 Bulgaria 1998 17.6 26.8 15.3 505 537 770 5 Czech Rep. 1999 6.8 12.3 5.4 1,300 2,720 5,180 5 Estonia 1998 0.0 0.0 0.0 267 581 306 5 Hungary 1999 12.4 32.2 7.4 2,173 2,036 1,944 5 Kazakhstan 1996 9.4 9.9 9.2 1,321 1,152 1,587 5 Latvia 1998 5.8 14.0 2.5 521 357 366 5 Lithuania 1997 4.6 9.1 2.7 355 926 486 5 Moldova 1996 6.7 11.2 4.9 76 81 34 5 Poland 1999 15.9 32.8 10.9 4,908 6,365 7,500 5 Romania 1999 13.1 23.6 10.7 1,215 2,031 961 5 Russian Fed. 1997 12.6 10.9 13.4 6,638 2,761 2,861 5 Slovak Rep. 1999 6.4 12.4 4.6 206 631 322 5 Slovenia 1996 10.6 13.8 9.9 321 165 90 5 Ukraine 1998 10.0 15.7 7.5 624 743 496 6 Argentina 1999 11.0 10.4 11.0 8,755 6,526 23,153 6 Barbados 1999 13.6 20.2 12.0 15 16 15 6 Belize 1998 9.2 21.0 8.2 12 18 3 6 Bolivia 1999 9.7 10.0 8.9 879 957 1,016 6 Brazil 1999 13.6 10.8 13.9 18,743 28,480 31,397 6 Chile 1999 10.0 10.0 10.0 5219 4,638 9,221 6 Colombia 1999 11.6 13.1 11.6 5,639 2,907 1,396 6 Costa Rica 1999 7.2 16.8 5.4 483 559 450 6 Cuba 1997 10.7 9.7 10.9 13 30 15 6 Dominican Rep. 1997 14.5 15.3 14.2 421 700 1,353 6 Ecuador 1999 11.6 15.5 11.0 695 831 636 6 El Salvador 1998 5.7 10.0 4.4 11 872 231 6 Guatemala 1999 7.6 10.7 7.0 85 673 147 6 Guyana 1998 10.4 23.1 9.3 53 47 48 6 Honduras 1999 8.1 12.2 7.5 128 99 230 6 Jamaica 1999 8.7 21.6 6.6 203 369 520 6 Mexico 1999 10.1 11.5 10.0 12,831 10,238 11,233 6 Nicaragua 1999 11.0 16.4 10.3 173 184 300 6 Panama 1998 9.2 11.4 8.5 1,256 1,206 22 6 Paraguay 1999 9.0 10.2 9.0 270 423 306 6 Peru 1998 13.2 14.7 13.0 1,702 1,930 2,068 6 Trinidad and Tobago 1998 9.2 20.0 8.4 1,000 732 633 6 Uruguay 1999 4.6 4.2 4.7 126 164 200 6 Venezuela, R. B. de 1999 12.0 12.5 11.9 5,536 4,435 2,607 (continued) 569 D E V E L O P M E N T, T R A D E , A N D T H E W T O Table A.2 (continued) Tariff rate (unweighted, percent) Net FDI inflow (millions of U.S. dollars) All Agri- Manu- Region Economy Year goods culture factures 1997 1998 1999 7 Australia 1998 5.3 1.2 5.8 7,732 6,345 5,422 7 Canada 1999 4.6 4.6 4.5 11,761 21,705 25,061 7 European Union 1999 5.0 10.0 4.2 128,574 248,615 305,058 7 Iceland 1999 4.0 10.8 2.5 149 148 66 7 Japan 1999 5.2 11.0 3.7 3,200 3,192 12,741 7 New Zealand 1999 3.7 2.0 4.0 2,623 745 ­1,063 7 Norway 1999 3.3 9.7 2.6 3,627 3,599 6,577 7 Switzerland 1998 0.0 0.0 0.0 6,636 7,500 3,413 7 United States 1999 4.8 8.7 4.3 105,488 186,316 275,533 Average/total 1­6 Developing economies (96) 1993­99 13.1 17.0 12.4 194,014 189,155 223,166 7 Industrial economies (23) 1998­99 4.0 6.4 3.5 269,790 478,165 632,808 1 East Asia (15) 1994­99 9.8 13.9 9.4 88,412 83,488 92,915 2 South Asia (5) 1996­99 27.7 26.3 28.0 4,889 3,668 3,183 3 Sub-Saharan Africa (26) 1993­99 16.5 19.2 16.0 7,409 3,954 4,934 4 Middle East and North Africa (11) 1995­99 14.4 20.8 13.2 8,426 9,776 11,806 5 Transition Europe (15) 1996­99 9.6 15.7 7.8 20,630 21,235 23,128 6 Latin America and the Caribbean (24) 1995­99 10.1 13.8 9.5 64,248 67,034 87,200 Sources: WTO, Integrated Data Base CD-ROM (2000), and Trade Policy Review (various issues, 1993­2000); World Bank, World Development Indicators 2000; UNCTAD, World Investment Report 2000. 570 Appendix A: Patterns of Trade and Protection: Selected Country Data Table A.3 Tariff Escalation in Developing and Industrial Countries, 1994­2000 (unweighted average; percent) Agricultural productsa Industrial productsb First Semi- Fully First Semi- Fully Region Economy Year stage processed processed stage processed processed 1 China 1997 19.3 34.3 29.2 7.4 13.3 19.3 1 Fiji 1996 5.0 10.0 22.5 1 Indonesia 1998 4.7 4.4 13.9 3.8 7.9 11.6 1 Korea, Rep. 1999 49.9 93.2 31.8 3.4 7.8 8.0 1 Malaysiab 1997 1.0 7.0 11.9 1 Papua New Guinea 1999 16.2 2.9 10.4 1 Philippines 1998 14.3 20.3 23.2 3.5 7.1 11.1 1 Solomon Islandsb 1998 29.0 15.4 25.6 1 Thailanda 1999 43.5 48.0 38.0 2 Bangladesha 1999 16.1 23.0 29.2 17.6 20.7 24.1 2 India 1997 25.4 29.9 42.8 23.6 35.4 36.4 2 Myanmar 1996 7.4 6.7 10.9 4.3 3.7 6.0 2 Sri Lankab 1994 30.0 40.0 20.0 15.6 22.5 3 Burkina Fasoa, b 1997 31.8 32.6 33.4 28.5 35.7 29.0 3 Cameroona, b 1994 23.9 23.0 25.0 20.0 15.0 21.0 3 Côte d'Ivoire 1994 18.1 21.6 26.2 3 Guineaa, 1998 17.4 18.5 18.1 16.8 16.6 16.1 3 Kenyaa 1999 16.4 24.7 24.7 15.2 17.9 18.6 3 Madagascar 1998 4.9 8.3 7.6 1.1 6.7 7.7 3 Mali 1999 14.5 15.2 18.0 4.7 7.4 12.8 3 Mauritiusb 1994 14.8 17.2 39.5 3 Nigeriab 1999 25.0 24.0 31.0 3 South Africaa, b 1997 12.4 10.9 15.6 4.9 18.6 13.8 3 Tanzaniaa, b 1999 18.5 25.0 23.0 13.3 13.3 18.3 3 Togo 1997 12.6 10.6 15.2 6.2 11.8 15.1 3 Zambiaa, b 1997 18.3 18.2 21.0 14.4 9.5 16.5 4 Bahraina, b 2000 4.8 2.8 11.2 5.0 6.2 9.0 4 Cyprus 1998 24.3 32.3 26.8 0.8 5.6 4.6 4 Egypt, Arab Rep. 1997 33.0 36.9 44.5 24.2 29.5 39.5 4 Israela,b 1999 12.0 7.0 24.5 12.8 4.1 8.4 4 Malta 1999 2.5 6.4 9.3 5.4 7.3 8.5 4 Morocco 1997 48.0 51.6 83.0 22.0 35.6 30.3 4 Tunisiaa,b 1994 35.4 33.7 43.0 32.0 31.3 34.6 4 Turkey 1997 35.1 43.7 64.7 0.5 6.7 5.8 5 Bulgaria 1998 20.1 26.3 32.3 8.4 12.8 17.3 5 Czech Rep. 1998 4.9 16.4 18.6 0.6 4.6 5.6 5 Hungary 1999 20.8 39.4 39.1 2.6 5.8 8.8 5 Latvia 1999 8.0 18.9 17.5 1.1 1.3 3.4 5 Polanda 1999 16.5 22.5 44.9 5.6 9.8 11.2 5 Romania 1999 57.9 110.0 158.4 25.9 16.9 17.1 5 Slovak Rep. 1999 4.6 15.4 17.3 0.6 4.3 5.1 6 Argentina 1998 10.0 13.9 16.0 7.7 12.0 15.3 6 Barbados 1999 22.7 16.9 18.4 11.0 6.9 13.8 6 Bolivia 1998 10.0 10.0 10.0 10.0 10.0 9.3 6 Brazil 2000 9.5 13.2 15.6 8.9 11.9 15.8 (continued) 571 D E V E L O P M E N T, T R A D E , A N D T H E W T O Table A.3 (continued) Agricultural productsa Industrial productsb First Semi- Fully First Semi- Fully Region Economy Year stage processed processed stage processed processed 6 Colombia 1998 12.8 17.7 18.6 6.9 9.6 12.2 6 Costa Rica 1999 10.4 13.0 23.8 2.7 3.7 6.6 6 Dominican Rep.a, b 1995 23.4 21.5 27.3 14.7 13.7 20.7 6 Ecuador 1999 12.5 16.8 18.5 6.7 9.4 12.5 6 El Salvadora, b 1996 12.9 15.0 15.9 7.6 8.4 12.5 6 Guatemala 1999 9.4 11.1 12.1 2.9 6.0 8.1 6 Honduras 1997 12.5 14.0 16.5 5.4 7.5 9.8 6 Jamaicab 1997 0.0 20.0 40.0 15.6 3.0 12.3 6 Mexico 1998 15.1 14.9 30.5 8.2 10.2 14.2 6 Nicaragua 1998 9.6 13.3 12.4 2.7 3.6 5.9 6 Panama 1997 17.2 20.3 22.7 9.1 7.3 14.0 6 Paraguay 1998 9.9 13.4 15.5 7.8 11.0 11.4 6 Peru 1998 13.9 14.6 15.6 12.1 13.1 13.0 6 Trinidad and Tobagob 1997 15.0 2.2 11.1 6 Uruguay 1998 10.1 14.0 15.9 8.2 11.6 12.8 7 Australia 1998 0.3 0.7 2.3 0.7 5.6 6.5 7 Canada 1999 1.7 3.6 7.0 0.7 4.2 5.1 7 European Union 1999 7.3 12.0 13.1 0.6 4.9 4.0 7 Icelanda, b 1999 2.0 9.0 11.5 4.0 0.9 5.7 7 Japan 1999 4.5 14.3 15.5 0.6 4.5 3.5 7 New Zealand 1999 0.5 2.7 2.8 0.3 2.1 5.4 7 Norway 1999 14.8 0.0 3.5 0.0 1.6 3.4 7 Switzerlanda, b 2000 4.6 30.5 41.8 21.6 4.0 8.5 7 United States 1999 7.1 4.5 10.3 0.6 5.0 4.1 Average 1­6 Developing economies (60) 1994­2000 17.9 23.2 27.7 10.7 11.9 15.5 7 Industrial economies (23) 1998­2000 4.8 8.6 12.0 3.2 3.6 5.1 Note: Blanks indicate not available. a. Tariff escalation of agricultural products is based on food processing only. b. Tariff escalation of industrial products is based on all goods. Sources: WTO, Integrated Data Base CD-ROM (2000), and Trade Policy Review (various issues, 1995­2000). 572 Appendix A: Patterns of Trade and Protection: Selected Country Data Table A.4 Pre­ and Post­Uruguay Round Nontariff Barriers for All Goods, Developing and Industrial Countries, 1984­93 (unweighted; percent) Coverage ratio of NTBsa Region code Economy 1984­87 1988­90 1991­93 1 China 10.6 23.2 11.3 1 Hong Kong (China) 3.4 0.5 0.5 1 Indonesia 94.7 9.4 2.7 1 Korea, Rep. 8.8 4.0 2.6 1 Malaysia 3.7 2.8 2.1 1 Papua New Guinea 2.6 1 Philippines 44.9 1 Singapore 14.7 1.0 0.3 1 Taiwan (China) 35.9 1 Thailand 12.4 8.5 5.5 2 Bangladesh 49.4 2 India 80.7 65.4 62.6 2 Nepal 10.7 0.7 2 Pakistan 83.1 25.4 14.5 2 Sri Lanka 13.9 10.1 3.8 3 Angola 0.7 3 Benin 17.0 3 Burkina Faso 80.6 3 Burundi 17.1 0.3 3 Cameroon 20.7 3 Central Africa Rep . 5.1 3 Congo, Rep. 4.6 3 Congo, Dem. Rep. 49.6 100.0 3 Côte d'Ivoire 6.6 3 Ethiopia 22.5 3 Ghana 48.4 3 Guinea 38.2 3 Kenya 67.3 37.8 3 Madagascar 56.7 1.7 3 Malawi 96.1 91.3 3 Mauritius 35.2 3 Mozambique 56.9 3 Nigeria 17.0 8.7 8.8 3 Senegal 7.2 3 Sierra Leone 100.0 3 Somalia 6.3 3 Sudan 10.0 3 Tanzania 62.2 79.7 3 Uganda 13.9 3 Zambia 0.0 3 Zimbabwe 2.5 93.6 4 Algeria 67.8 93.0 9.5 4 Bahrain 1.5 4 Egypt, Arab Rep. 32.9 45.2 (continued) 573 D E V E L O P M E N T, T R A D E , A N D T H E W T O Table A.4 (continued) Coverage ratio of NTBsa Region code Economy 1984­87 1988­90 1991­93 4 Iran 99.3 4 Jordan 12.9 4 Kuwait 3.8 3.5 4 Libya 10.3 4 Morocco 27.6 4 Oman 3.6 4 Qatar 1.3 4 Saudi Arabia 5.4 3.9 4 Syria 36.6 4 Tunisia 76.2 63.7 32.7 4 United Arab Emirates 1.0 4 Yemen 28.7 5 Cyprus 32.2 5 Romania 0.0 5 Turkey 97.2 96.4 5 Yugoslavia, FR 27.7 29.2 6 Argentina 31.9 16.1 0.2 6 Bahamas, The 0.1 6 Bolivia 25.0 2.0 6 Brazil 35.3 3.2 1.5 6 Chile 10.1 10.6 0.1 6 Colombia 73.2 73.8 1.7 6 Costa Rica 0.8 6 Ecuador 59.3 63.6 6 El Salvador 19.2 6 Guatemala 7.4 6 Guyana 16.0 6 Haiti 30.8 6 Jamaica 6.6 6 Mexico 12.7 6.3 3.9 6 Nicaragua 27.8 6 Paraguay 9.9 1.8 6 Peru 53.4 6 Trinidad and Tobago 23.4 6 Uruguay 14.1 6 Venezuela, R. B. de 44.1 11.9 2.4 7 Australia 3.4 0.7 7 Canada 11.1 11.0 7 European Union 26.6 23.7 7 Iceland 3.9 7 Japan 13.1 12.2 7 New Zealand 14.1 0.4 7 Norway 26.6 23.7 7 Switzerland 12.9 13.5 7 United States 25.5 22.9 574 Appendix A: Patterns of Trade and Protection: Selected Country Data Average 1­6 Developing economies (80) 29.8 32.7 8.1 7 Industrial economies (23) 16.7 12.4 1 East Asia (10) 23.2 7.1 3.6 2 South Asia (5) 47.6 33.6 20.4 3 Sub-Saharan Africa (26) 32.7 47.1 8.8 4 Middle East and North Africa (15) 27.3 41.9 21.1 5 Transition Europe (4) 39.3 62.8 6 Latin America and the Caribbean (20) 25.1 23.4 1.7 Note: Blanks indicate not available. a. Includes additional quantitative restrictions in the form of all types of licenses and import authorizations, quotas, import prohibitions, advance import deposits, foreign exchange restrictions, fixed customs valuations, and state trading monopolies. Calculated as percentage of products within a category affected by an NTB applied to a tariff line. Sources: UNCTAD, Directory of Import Regimes, 1994, and Handbook of Trade Control Measures of Developing Countries, Supplement, 1987; OECD, Indicators of Tariff and Non-Tariff Barriers (1997b). 575 98 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 1995­ min..raV pricing 94 0.0 1.0 0.0 0.0 0.0 0.0 3.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 1989­ 98 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 1995­ Import monitoring 94 0.0 0.0 0.0 0.0 0.0 2.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 1989­ 16.0 98 0.0 0.0 0.0 7.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 1995­ 25.0 12.0 21.0 quotas ariffT 94 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 1989­ 26.0 98 0.0 0.0 0.0 0.0 2.0 0.0 1.0 0.0 0.0 0.0 5.0 0.0 0.0 3.0 0.0 0.0 0.0 1995­ Quotas 94 0.0 3.0 3.0 2.0 1.0 0.0 2.0 0.0 0.0 1.0 0.0 0.0 0.0 0.0 5.0 0.0 0.0 0.0 1989­ 98 0.0 0.0 0.0 0.0 1.0 6.0 1.0 0.0 1.0 0.0 7.0 9.0 0.0 0.0 1.0 0.0 1995­ 14.0 1989­98 Prohibition 94 0.0 5.0 0.0 4.0 4.0 0.0 0.0 3.0 0.0 3.0 0.0 0.0 0.0 0.0 1989­ 43.0 17.0 14.0 53.0 98 5.0 2.0 0.0 1.0 0.0 9.0 2.0 5.0 3.0 0.0 1.0 Countries, 1995­ 31.0 20.0 11.0 94.0 23.0 31.0 Non-auto. licensing 94 2.0 7.0 1.0 0.0 8.0 0.0 0.0 1989­ 53.0 32.0 55.0 36.0 34.0 99.0 87.0 10.0 36.0 23.0 14.0 Developing in 98 5.2 2.1 2.1 1.0 8.3 3.1 1.0 1995­ 31.3 25.0 19.6 17.5 93.8 22.7 30.9 16.7 11.5 21.6 NTMs Products Core 94 2.1 1.0 8.2 3.1 All 1989­ 53.6 50.0 56.3 11.5 36.5 54.2 99.0 17.7 86.6 14.4 10.3 36.5 22.7 57.3 for (China) Barriers Kong Rep. d'Ivoire Africa Lanka Economy auritius igeria yprus Nontariff Fiji Hong Indonesia Korea, Malaysia Philippines Singapore Thailand Bangladesh India Pakistan Sri Benin Cameroon Côte Ghana Kenya Senegal South Uganda Zambia Zimbabwe Egypt A.5 cent) 1 1 1 1 1 1 1 1 2 2 2 2 3 3 3 3 3 3M 3N 3 3 3 3 3 4C 4 ableT code (per Region 576 0.0 0.0 1.0 1.0 4.0 6.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 administered 2.0 0.0 0.0 0.0 0.0 5.0 0.0 2.0 6.0 24.0 31.0 and quotas, 0.0 0.0 0.0 0.0 0.0 0.0 0.0 1.0 0.0 0.0 0.0 4.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 prohibitions, 100.0 licensing, 1.0 0.0 0.0 0.0 0.0 6.0 0.0 0.0 7.0 0.0 0.0 0.0 11.0 include 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 1.0 NTMs Core 0.0 8.0 1.0 1.0 0.0 0.0 0.0 0.0 1.0 0.0 0.0 0.0 0.0 categories. ct 0.0 0.0 0.0 2.0 0.0 0.0 0.0 3.0 2.0 0.0 1.0 produ 0.0 0.0 0.0 1.0 1.0 0.0 1.0 1.0 1.0 0.0 0.0 3.0 two-digit 11.0 HS all of 0.0 0.0 0.0 0.0 0.0 7.0 1.0 7.0 0.0 0.0 0.0 cent per in ratio 0.0 1.0 0.0 6.0 6.0 5.0 5.0 6.0 0.0 0.0 2.0 13.0 11.0 frequency 5.0 3.1 0.0 0.0 0.0 0.0 51.0 54.0 10.0 55.0 28.0 as calculated 2.1 5.2 6.2 6.2 5.2 0.0 0.0 13.4 19.8 21.6 10.3 13.4 17.7 are (NTBs) 5.2 3.1 0.0 5.2 6.3 58.3 54.2 16.5 55.2 27.8 32.3 barriers Nontariff Rep. available. Rica (1999c). not Salvador Morocco unisia urkey enezuela Argentina Bolivia Brazil Chile Colombia Costa Dominican El Mexico Paraguay Peru Uruguay indicate Michalopoulos 4 4T 4T 5 5 5 5 5 5 5 5 5 5 5 5 5V Blanks ce: Note: pricing. Sour 577 D E V E L O P M E N T, T R A D E , A N D T H E W T O Table A.6 Tariff Peaks and Preference Margins, Canada, 1999 (unweighted; percent) Number of 6­digit MFN Tariff peak in HS 2-digit product lines tariff LDC GSP Mexico Chile CAR 01 Live animals 4 198.8 0 0 0 0.25 0.25 02 Meat and edible meat offal 10 109.9 0.6 0.17 0.17 0.6 0.6 04 Dairy products; birds' eggs; honey 26 197.5 0.31 0.03 0.05 0.5 0.54 06 Live tree & other plant; bulbs, cut flowers 1 15.2 0.47 0.47 0.51 0.6 1 08 Edible fruit and nuts; melons 1 16.6 0 0 0 0 0 10 Cereals 3 70.2 0 0 1 0.69 0.53 11 Prod mill indust; malt; starches 7 85.3 0.14 0.02 0.14 0.07 0.14 12 Oil seed, oleagi fruits; misc grain 1 18.3 0 0 0 0 0 13 Lac; gums, resins and other veg 1 74 0 0 0 0 0 15 Animal/veg fats and oils and prod 8 28 0.8 0.57 1 1 1 16 Prep of meat, fish or mollusks 5 68.7 0.59 0.14 0.22 0.5 0.66 17 Sugars and sugar confectionery 1 16.6 0 0 0 0 0 18 Cocoa and cocoa preparations 2 85.5 1 0.18 1 0.73 1 19 Prep of cereal, flour, starch/ milk prod 3 54.5 0.96 0.26 0.87 0.96 1 20 Prep of vegetable, fruit, nuts prod 2 19.4 0 0 0.77 0.48 1 21 Miscellaneous edible preparations 4 48.9 0.97 0.37 0.77 0.83 1 22 Beverages, spirits and vinegar 3 26.7 0.67 0.27 1 1 1 23 Residues and waste from food indust 2 30.3 0.29 0 0.4 0.5 0.5 24 Tobacco and manufactured 1 17.6 0 0 1 1 1 33 Essential oils & resinoids, perf 1 18 1 0.74 1 0.57 1 35 Albuminoidal subs; modified starches 2 18 0 0 0 0.5 0.5 39 Plastics and articles thereof 5 16.7 1 0.36 0.7 0.72 1 40 Rubber and articles thereof 7 16.8 1 0.37 0.85 1 1 42 Articles of leather; saddlery, travel pr 3 16.6 1 0.42 0.72 1 1 43 Furskins and artificial fur thereof 2 19.2 1 0.34 0.61 1 1 51 Wool, fine/coarse animal hair nes 7 16.5 0.87 0.41 1 1 0 52 Cotton 66 17.3 0.15 0.06 0.62 0.62 0 53 Other vegetable textile fibers and yarns 2 16 1 0.13 1 1 0 54 Man-made filaments 34 19 0.01 0 1 1 0 55 Man-made staple fibres 68 19 0.1 0.03 1 1 0 578 Appendix A: Patterns of Trade and Protection: Selected Country Data 56 Wadding, felt and nonwoven, yarns etc. 22 16.9 0.39 0.16 0.77 0.77 0 57 Carpets and other textile floor coverings 14 18.9 1 0.47 0.77 0.77 0 58 Special woven fab; tufted tex fab, etc. 31 17.9 0.34 0.11 0.65 0.65 0 59 Impregnated, coated, cover/ laminated 13 17.8 0.9 0.33 0.73 0.73 0 60 Knitted or crocheted fabrics 16 18 0.38 0.12 0.65 0.65 0 61 Art of apparel and clothing access 114 23.3 0.07 0.02 0.62 0.62 0 62 Art of apparel and clothing access 116 22.4 0.09 0.04 0.66 0.66 0 63 Other made-up textile articles 52 22.1 0.14 0.06 0.68 0.68 0.03 64 Footwear, gaiters and articles 23 20.8 0.12 0.07 0.7 0.67 0 65 Headgear and parts thereof 4 18.7 0.61 0.12 1 1 0 67 Prepr feathers & down; art flower nes 4 21.3 1 1 1 1 1 68 Art of stone, plaster, cement, asbestos 1 21.3 0 0 1 1 1 70 Glass and glassware 7 16.6 0.46 0.22 0.73 1 1 85 Electrical mech equip parts, sound pr 3 16.7 0 0 0 0 0 86 Railw/tramw locom, rolling stock, etc. 7 15 1 0.55 0.92 1 1 89 Ships, boats and floating structures 10 22.5 1 0.04 0.74 1 1 91 Clocks and watches and parts nes 6 18.6 1 0.5 1 1 1 94 Furniture; bedding, mattress, cushion 2 21 0 0 0.77 1 1 95 Toys, games and sports requisites nes 1 16.4 0.13 0.06 0.84 1 1 96 Miscellaneous manufactured articles 4 17.5 0.68 0.34 0.76 1 1 Note: nes, not elsewhere specified; CAR, Caribbean Community; GSP, Generalized System of Preferences; LDC, least-developed countries; MFN, most-favored-nation. Tariff peaks are defined as MFN duties that exceed 15 percent. Sources: OECD and WTO tariff files. 579 D E V E L O P M E N T, T R A D E , A N D T H E W T O Table A.7 Tariff Peaks and Preference Margins, European Union, 1999 (unweighted; percent) Number of Tariff peak in HS 2-digit product 6-digit lines MFN LDC GSP ACP 01 Live animals 7 38.2 0.06 0 0.3 02 Meat and edible meat offal 41 71 0.08 0 0.1 03 Fish & crustacean, mollusk nes 17 18.7 1 0.1 1 04 Dairy prod; birds' eggs; honey 25 59.1 0.12 0.01 0.06 06 Live tree & other plant; bulb, cut flowers 2 16.9 1 0.18 1 07 Edible vegetables and roots and tubers 12 25.4 0.79 0.15 0.66 08 Edible fruit and nuts; melons 8 20.2 0.66 0.12 0.64 09 Coffee, tea, mat and spices 2 16 0.5 0.69 1 10 Cereals 14 75.6 0.06 0 0.06 11 Prod mill indust; malt; starches 31 38.2 0.17 0.02 0.2 12 Oil seed, oleagi fruits; misc grain 1 74.4 0.15 0 0.16 13 Lac; gums, resins and other veg 1 17.8 1 0.3 1 15 Animal/veg fats and oils and prod 8 56 0.6 0.19 0.51 16 Prep of meat, fish or mollusks 22 23.5 0.68 0.2 0.67 17 Sugars and sugar confectionery 9 37.6 0.14 0.03 0.21 18 Cocoa and cocoa preparations 1 24 0.25 0.1 0.25 19 Prep of cereal, flour, starch/ milk prod 13 34.1 0.37 0.11 0.39 20 Prep of vegetable, fruit, nuts prod 42 26.1 0.88 0.15 0.88 21 Miscellaneous edible preparations 8 19.2 0.95 0.28 0.78 22 Beverages, spirits and vinegar 12 35.7 0.71 0.33 0.77 23 Residues and waste from food indust 6 71.4 0.06 0.03 0.11 24 Tobacco and manufactured 8 56.2 1 0.39 1 29 Organic chemicals 3 33.9 1 0.33 0.53 35 Albuminoidal subs; modified starches 2 24.9 1 1 1 38 Miscellaneous chemical products 2 45.9 0.5 0 0.33 56 Wadding, felt and nonwoven, yarns, etc. 2 21.1 1 0.15 1 64 Footwear, gaiters and articles 13 18.2 1 0.3 1 87 Vehicles o/t railw/tramw roll stock, pts 5 16.3 1 0.44 1 Note: nes, not elsewhere specified; ACP, African, Caribbean, and Pacific (Lomé Convention) countries; GSP, Generalized System of Prefer- ences' LDC, least-developed countries; MFN, most-favored-nation. Sources: OECD and WTO tariff files. 580 Appendix A: Patterns of Trade and Protection: Selected Country Data Table A.8 Tariff Peaks and Preference Margins, Japan, 1999 (unweighted; percent) Number of MFN Tariff peak in HS 2-digit product 6-digit lines tariff LDC GSP 02 Meat and edible meat offal 9 39.31 0.13 0.13 03 Fish & crustacean, mollusk nes 4 15 0 0 04 Dairy prod; birds' eggs; honey 25 28.99 0.05 0.05 07 Edible vegetables and roots and tubers 1 15.8 0 0 08 Edible fruit and nuts; melons 11 19.81 0.15 0.09 09 Coffee, tea, mat and spices 5 17.81 0.49 0.11 10 Cereals 1 63.38 0 0 11 Prod mill indust; malt; starches 28 23.24 0.11 0.06 12 Oil seed, oleagi fruits; misc grain 1 19.1 0 0 15 Animal/veg fats and oils and prod 2 26.99 0 0 16 Prep of meat, fish or mollusks 4 20.69 0.05 0.05 17 Sugars and sugar confectionery 11 71.25 0.05 0.05 18 Cocoa and cocoa preparations 6 22.77 0.49 0.18 19 Prep of cereal, flour, starch/milk prod 13 21.91 0.15 0.04 20 Prep of vegetable, fruit, nuts prod 32 22.69 0.23 0.06 21 Miscellaneous edible preparations 7 22.35 0.19 0.11 22 Beverages, spirits and vinegar 16 38.65 0.39 0.16 24 Tobacco and manufactured 2 18.63 0 0 29 Organic chemicals 2 20 1 1 35 Albuminoidal subs; modified starches 2 23.31 1 1 38 Miscellaneous chemical products 1 80.83 1 1 41 Raw hides and skins (other than fur) 10 26.08 1 0.57 42 Articles of leather; saddlery, travel pr 5 15.52 0.2 0.1 43 Furskins and artificial fur thereof 8 16.25 0.47 0.47 53 Other vegetable textile fibres and yarns 4 16 1 1 58 Special woven fab; tufted tex fab, etc 4 17.9 1 0.5 60 Knitted or crocheted fabrics 3 15.7 1 0.5 64 Footwear, gaiters and articles 16 36.24 0.81 0.4 Note: nes, not elsewhere specified; GSP, Generalized System of Preferences; LDC, least-developed countries; MFN, most-favored-nation. Sources: OECD and WTO tariff files. 581 D E V E L O P M E N T, T R A D E , A N D T H E W T O Table A.9 Tariff Peaks and Preference Margins, United States, 1999 (unweighted; percent) Number of Tariff peak in HS 2-digit 6-digit MFN product description lines tariff LDC GSP Mexico ATP CAR 02 Meat and edible meat offal 2 19.2 0 0 1 0 0 04 Dairy prod; birds' eggs; honey 2 20.9 0.38 0 0.25 0.38 0.38 07 Edible vegetables and roots nes 10 20.56 0.88 0.18 0.9 1 1 08 Edible fruit and nuts; melons 5 16.66 0.8 0.18 0.74 0.8 0.8 11 Prod mill indust; malt; starches 1 16.3 1 0 1 1 1 12 Oil seed, oleagi fruits; misc, etc. 3 77.95 0 0.33 0.67 0 0 15 Animal/veg fats and oils and prod 4 19.92 0.5 0.25 0.5 0.5 0.5 19 Prep of cereal, flour, milk prod 2 16.79 0.84 0.5 0.63 0.84 0.84 20 Prep of vegetable, fruit, nuts 11 28.67 0.55 0.11 0.56 0.55 0.55 21 Miscellaneous edible prep 1 19.8 0.74 0 0.43 0.74 0.74 24 Tobacco and manufactured 7 73.48 0.14 0.09 0.96 0.14 0.14 28 Inorgn chem; compds of prec 1 15.1 1 1 1 1 1 29 Organic chemicals 4 16.75 1 1 1 1 1 30 Pharmaceutical products 1 30 0 0 0 0 0 42 Articles of leather; saddlery 1 20 0 0 0.6 0.13 0.13 51 Wool, fine/coarse animal hair 4 20.45 0 0 1 0 0 52 Cotton 10 18.34 0 0 1 0 0 54 Man-made filaments 27 16.37 0 0 1 0 0 55 Man-made staple fibres 56 16.27 0 0 0.97 0 0 56 Wadding, felt and nonwoven, yarn 1 15.2 0 0 1 0 0 58 Special woven fab; tufted fabrics 15 18.47 0 0 1 0.07 0.2 60 Knitted or crocheted fabrics 5 18.58 0 0 1 0 0 61 Art of apparel and & clothing access 58 19.5 0 0 0.85 0.02 0.05 62 Art of apparel and clothing access 41 18.85 0 0 0.83 0.02 0.14 64 Footwear, gaiters and parts, etc. 12 27.77 0.01 0.01 0.57 0.01 0.01 69 Ceramic products 4 17.63 0.09 0.09 0.27 1 1 70 Glass and glassware 4 16.16 1 0.08 0.42 1 1 82 Tool, implement, cutlery, spoons 1 15.23 1 0.65 0.65 1 1 86 Railw/tramw locom, rolling stock 7 17.2 1 1 1 1 1 87 Vehicles o/t railw/tramw roll stock 5 25 1 0 1 1 1 96 Miscellaneous manufactures 2 20.66 0.78 0.57 1 0.78 0.78 Note: nes, not elsewhere specified; ATP, Andean Trade Preferences Act; CAR, GSP, Generalized System of Preferences; LDC, least-developed countries; MFN, most-favored-nation. Sources: OECD and WTO tariff files. 582 Appendix A: Patterns of Trade and Protection: Selected Country Data Table A.10 Comparison of MFN Applied Tariffs of Labor-Intensive Products, Selected Countries, 1997­99 (unweighted; percent) Labor-intensive product Region Industrial Textiles and Protection code Economy Year product clothing Footwear Average ratio Developing economies 1 China 1997 16.5 27.7 25.0 27.6 1.7 1 Hong Kong (China) 1999 0.0 0.0 0.0 0.0 1.0 1 Indonesia 1998 9.7 14.9 16.8 14.9 1.5 1 Korea, Rep. 1999 7.6 10.4 11.5 10.4 1.4 1 Myanmar 1997 5.1 11.2 5.3 11.0 2.2 1 Philippines 1998 9.2 17.8 22.2 18.0 2.0 1 Singapore 1998 0.0 0.0 0.0 0.0 1.0 2 India 1997 35.2 44.1 45.0 44.1 1.3 3 Egypt 1997 35.0 52.4 61.8 52.7 1.5 3 Madagascar 1998 6.9 17.3 10.0 17.0 2.5 3 Mali 1999 10.4 19.7 21.1 19.7 1.9 3 Morocco 1997 32.1 47.1 50.0 47.2 1.5 4 Bulgaria 1998 15.2 23.4 25.8 23.6 1.6 4 Cyprus 1998 4.8 9.6 10.3 9.6 2.0 4 Czech Rep. 1998 5.0 7.5 9.3 7.6 1.5 4 Estonia 1998 0.0 0.0 0.0 0.0 1.0 4 Hungary 1999 7.4 9.3 9.5 9.3 1.3 4 Latvia 1999 2.5 7.4 9.7 7.5 3.0 4 Malta 1999 7.9 10.1 10.0 10.1 1.3 4 Romania 1999 17.5 24.6 25.5 24.7 1.4 4 Slovak Rep. 1999 4.6 7.3 7.9 7.4 1.6 4 Turkey 1997 5.8 9.9 23.6 10.5 1.8 5 Argentina 1998 13.6 20.9 26.6 21.1 1.6 5 Bolivia 1998 9.6 10.0 10.0 10.0 1.0 5 Brazil 1998 14.8 20.7 27.0 20.9 1.4 5 Chile 1999 10.0 10.0 10.0 10.0 1.0 5 Colombia 1998 10.9 18.5 19.2 18.6 1.7 5 Costa Rica 1999 5.4 12.6 15.4 12.7 2.4 5 Ecuador 1999 11.0 18.6 19.2 18.6 1.7 5 El Salvador 1997 7.1 18.7 18.5 18.7 2.6 5 Guatemala 1999 7.0 20.9 23.3 20.9 3.0 5 Mexico 1998 12.4 22.2 33.6 22.8 1.8 5 Nicaragua 1998 4.9 11.2 13.1 11.3 2.3 5 Paraguay 1998 11.1 19.0 21.0 19.0 1.7 5 Peru 1998 13.0 17.8 20.0 17.9 1.4 5 Uruguay 1998 12.0 20.9 22.6 21.0 1.8 Industrial economies 6 Australia 1998 5.8 16.2 13.1 16.0 2.8 6 Canada 1999 4.5 11.4 14.8 11.6 2.6 6 European Union 1999 4.2 9.4 10.0 9.4 2.2 6 Japan 1999 3.7 8.6 19.1 9.1 2.5 6 New Zealand 1999 4.0 8.3 11.6 8.6 2.2 (continued) 583 D E V E L O P M E N T, T R A D E , A N D T H E W T O Table A.10 (continued) Labor-intensive product Region Industrial Textiles and Protection code Economy Year product clothing Footwear Average ratio 6 Norway 1999 2.6 9.0 5.0 8.8 3.4 6 United States 1999 4.3 10.1 14.1 10.3 2.4 Memorandum: averages 1 East Asia (7) 1997­99 6.9 11.7 11.5 11.7 1.7 2 South Asia (1) 1997 35.2 44.1 45.0 44.1 1.3 3 Middle East and Africa (4) 199799 21.1 34.1 35.7 34.2 1.6 4 Other Europe (10) 1997­99 7.1 10.9 13.2 11.0 1.6 5 Latin America (14) 1997­99 10.2 17.3 20.0 17.4 1.7 1­5 Developing economies (36) 1997­99 10.6 17.0 18.9 17.1 1.6 6 Industrial economies (21) 1998­99 4.2 10.4 12.5 10.5 2.5 Source: WTO, Integrated Data Base CD-ROM (2000). 584 Appendix F R A N C I S N G B Trade Indicators growth rate (G) over the period can be calculated as: and Indices Gi = (Xt / Xt )(1/n­1)*100 2 1 where Xt and Xt are the trade 1 2 values of product i in the begin- ning period and the end period, respectively, and n is number of years. Export Product Dynamics T Although some products may his appendix describes and defines not constitute a large share of exports in a country, the most commonly used indicators, there are several reasons to identify dynamic (fast- indices, and ratios that are used by trade analysts to growing) products in exports. If above-average assess trade patterns and characteristics, and changes growth in these products continues for an extended in them. All of these can be calculated using stan- period, these items may eventually become an dard spreadsheets and the CD-ROM that is included important source of a country's export earnings. In in this Handbook. (See Appendix C for a description addition, if the dynamic products have specific pro- of the database.) More detailed (disaggregated) data duction characteristics, this could also convey can be drawn from the United Nations COM- important information on export opportunities in TRADE statistical database or downloaded from the relation to other similar goods. Finally, there is an World Trade Database (WTD) maintained by Statis- obvious interest in identifying dynamic products to tics Canada, at . Since many focus future multilateral or bilateral negotiations on low-income countries fail to report their own trade the removal of trade barriers on such products in data, partner-country trade statistics may have to be export markets. The most straightforward method used to construct trade series. of identifying dynamic products is to sort products on the basis of their growth rate over a given period. Relative Growth Rates of Merchandise Exports and Imports Trade Intensity Index This indicator is used to compare rates of growth of The trade intensity index (T) is used to determine exports and imports of broad classes of goods in whether the value of trade between two countries is one country with those for world trade or the trade greater or smaller than would be expected on the of its competitors, including the major products in basis of their importance in world trade. It is exports and imports. The annual compound defined as the share of one country's exports going 585 D E V E L O P M E N T, T R A D E , A N D T H E W T O to a partner divided by the share of world exports to information about potential trade prospects with the partner. It is calculated as: new partners. Countries with similar RCA profiles, such as those in Africa, are unlikely to have high Tij = (xij/Xit)/(xwj/Xwt) bilateral trade intensities unless intraindustry trade is involved. RCA measures, if estimated at high lev- where xij and xwj are the values of country i's exports els of product disaggregation, can focus attention and of world exports to country j and where Xit and on other nontraditional products that might be suc- Xwt are country i's total exports and total world cessfully exported. The RCA index of country i for exports, respectively. An index of more (less) than product j is often measured by the product's share unity indicates a bilateral trade flow that is larger in the country's exports in relation to its share in (smaller) than expected, given the partner country's world trade: importance in world trade. RCAij = (xij /Xit) / (xwj / Xwt) Intraindustry Trade where xij and xwj are the values of country i's exports Some analyses of factors influencing the success or of product j and of world exports of product j and failure of efforts to promote industrialization and where Xit and Xwt refer to the country's total exports growth conclude that a growing level of intraindus- and world total exports. A value of less than unity try trade (IIT) plays an important positive role. implies that the country has a revealed comparative Intraindustry exchange produces extra gains from disadvantage in the product. Similarly, if the index international trade over and above those associated exceeds unity, the country is said to have a revealed with comparative advantage because it allows a comparative advantage in the product. country to take advantage of larger markets. By engaging IIT, a country can simultaneously reduce Export Specialization Index the number of similar products it produces while increasing the variety of goods available to domestic The export specialization (ES) index is a slightly consumers. The IIT index ranges between zero and modified RCA index in which the denominator is one, with larger values indicating a greater level of usually measured by specific markets or partners. It trade between firms in the same industry. Higher provides product information on revealed special- IIT ratios suggest that net gains from specialization ization in the export sector of a country and is cal- in different products are being exploited and that culated as the ratio of the share of a product in a the participating country is increasing its integra- country's total exports to the share of this product tion into the world economy. IIT is calculated as: in imports to specific markets or partners rather than its share in world exports: IITjk = 1 ­ [sumi | Xijk ­ Mijk| / (Xijk + Mijk)] ES = (xij / Xit) / (mkj / Mkt) where Xijk and Mijk represent exports and imports of products from industry i in country j to and from where xij and Xit are export values of country i in country k. The computation is generally confined to product j and total exports of country i, respective- manufactured goods defined at the Standard Indus- ly, and where mkj and Mkt are the import values of trial Trade Classification (SITC) three-digit level. product j in market k and total imports in market k. The ES is similar to the RCA in that a value of the index less than unity indicates a comparative disad- Revealed Comparative Advantage Index vantage and a value above unity represents special- Measures of revealed comparative advantage (RCA) ization in this market. have been used to help assess a country's export potential. The RCA indicates whether a country is Export Diversification (or Concentration) in the process of extending the products in which it Index has a trade potential, as opposed to situations in which the number of products that can be competi- Export diversification is held to be important for tively exported is static. It can also provide useful developing countries because many developing 586 Appendix B: Trade Indicators and Indices countries are often highly dependent on relatively information on distinctive export patterns from few primary commodities for their export earnings. country to country. It is defined as: Unstable prices for these commodities may subject a developing country exporter to serious terms of XS (j, k) = sum [min (Xij, Xik) * 100] trade shocks. Since the covariation in individual commodity prices is less than perfect, diversifica- where Xij and Xik are industry i's export shares in tion into new primary export products is generally country j's and country k's exports, which usually viewed as a positive development. The strongest include a group of countries or competitors. The positive effects are normally associated with diversi- index varies between zero and 100, with zero indi- fication into manufactured goods, and its benefits cating complete dissimilarity and 100 representing include higher and more stable export earnings, job identical export composition. This measure is sub- creation and learning effects, and the development ject to aggregation bias (as the data are more finely of new skills and infrastructure that would facilitate disaggregated, the index will tend to fall) and hence the development of even newer export products. embodies a certain arbitrariness due to product The export diversification index (DX) for a country choice (see Noland 1997 for details). is defined as: Trade Complementarity Index DXj = (sum |hij ­ hi|) / 2 The trade complementarity index can provide use- where hij is the share of commodity i in the total ful information on prospects for intraregional trade exports of country j and hi is the share of the com- in that it shows how well the structures of a coun- modity in world exports. The related measure used try's imports and exports match. It also has the by UNCTAD is the concentration index, or attraction that its values for countries considering Hirschman index (H), which is calculated using the the formation of a regional trade agreement can be shares of all three-digit products in a country's compared with others that have formed or tried to exports: form similar arrangements (see Michaely 1996 for details). The index of trade complementarity (TC) Hj = sqrt [ sum (xi / Xt)2] between countries k and j is defined as: where xi is country j's exports of product i (at the TCij = 100 ­ sum (|mik ­ xij| / 2) three-digit SITC classification) and Xt is country j's total exports. The index has been normalized to where xij is the share of good i in the global exports account for the number of actual three-digit prod- of country j and mik is the share of good i in all ucts that could be exported. Thus, the maximum imports of country k. The index is zero when no value of the index is 239 (the number of individual goods are exported by one country or imported by three-digit products in SITC revision 2), and its the other and 100 when the export and import minimum (theoretical) value is zero, for a country shares exactly match. with no exports. The lower is this index, the less concentrated are a country's exports. Changes in Global Demand for Major Exports Export Similarity Index The index of global demand changes is a constant Many countries have an unusual pattern of export market share analysis of export performance in a specialization in relation to the rest of the world. country due to the relative favorable or unfavorable Often, some product exports, typically manufac- changes in global demand prospects. It indicates tures, have grown more rapidly than the average of how rapidly a country's recent exports would grow world exports. It is not clear, however, to what relative to world trade if the country just main- extent these results reflect a common tendency tained its current market for these products. This among countries and to what extent the results are approach isolates the influence of change in global driven by the performance of individual countries. demand for specific goods from any changes in the The export similarity index (XS) provides useful country's market shares or from diversification into 587 D E V E L O P M E N T, T R A D E , A N D T H E W T O new product lines. Specifically, the projected impact goods it exports. In some cases improvements in of global demand changes (GD) on the exports of a market shares can offset relatively sluggish demand product is derived as: for a product, whereas the erosion of shares can make a bad situation worse. The influence of mar- GDj = sum Si (Xit ­ Xi ) ket share changes (CD) on exports can be estimated 0 0 using a procedure analogous to that used for quan- where Si is country j's global market share for tification of demand effects. This competitive 0 product i in initial period 0 and Xit and Xi repre- effects measure for a product is expressed as: 0 sent global exports of product i in periods 0 and t. The right-hand side of the equation is summed over CDj = (Sit ­ Si ) * Mgt 0 all traditional products to produce an aggregate demand change index for the country. where Si and Sit represent a country's global market 0 share for product i in base year 0 and end year t and Changes in Global Market Share for Major Mgt is global imports of the traditional product in year t. The index indicates the dollar value of export Exports gains or losses associated with a country's market Aside from the influence of global demand changes, share changes. Summing over all traditional products there are other factors that can influence changes in will indicate whether the country's competitive posi- the level and growth of a country's exports. One of tion improved or worsened and what was the magni- these is change in a country's market shares for the tude of the associated change in the value of exports. 588 Appendix A L E S S A N D R O N I C I TA M A R C E L O O L A R R E A G A C A Dataset on Nations. Tariff data are from the WTO's Trade Policy Review series and the Internet version of Trade and the Integrated Data Base (IDB), and from UNCTAD's Trade Production, Analysis and Information Sys- tem (TRAINS).5 The original 1976­99 source for input-output tables is the Global Trade Analysis Pro- ject (GTAP).6 The Trade and Production Database uses an OECD concor- dance that maps trade data com- T piled on the basis of Standard he Trade and Production Database International Trade Classifica- CD-ROM that is included with this tion (SITC) codes into the ISIC categories. A con- Handbook merges trade, production, and tariff data cordance from the Harmonized Commodity available from different sources into a common Description and Coding System (HS) to the ISIC, classification, the International Standard Industrial developed by Jerzy Rozanski of the World Bank, was Classification, rev. 2, abbreviated here as ISIC (2).1 used to filter tariff data into the ISIC classification. Data availability varies, but the database potentially All filters are provided in the CD-ROM under the covers 67 developing and industrial countries over directory "Concordances." the period 1976­99. The database complements those provided in Feenstra (1996, 2000) and Feen- Description of the Database on Trade and stra, Lipsey, and Bowen (1997).2 Production The United Nations Industrial Development Organization (UNIDO) is the source for produc- The Trade and Production Database is constructed tion-related data, which were obtained from the using the ISIC classification and includes trade, pro- CD-ROM version of UNIDO's Industrial Statistics duction, and tariff data. Depending on the country, Databases at the three- and four-digit levels of the the database covers a time span from the late 1970s ISIC classification.3 The main source of trade data is to the late 1990s. When available, data on tariffs and the United Nations Statistical Department, which an input-output table are provided for each coun- collects data from individual countries and reports try. The database is divided into two independent the data in the Commodity Trade Statistics (COM- databases. The first covers 67 countries and reports TRADE).4 The World Bank's World Integrated data at the three-digit ISIC level for a total of 28 Trade Solution (WITS) software was used to "mir- manufacturing sectors. The second covers a subset ror" trade, using partner data, when countries did of 24 countries and reports the data at the more dis- not report their trade statistics to the United aggregated four-digit ISIC level, covering 81 manu- 589 D E V E L O P M E N T, T R A D E , A N D T H E W T O facturing sectors. Each data set has undergone partner countries) are also provided.7 The World extensive quality control and has been examined for Bank's WITS system was used to mirror missing anomalies. trade data. To make the database manageable, the trade flows are aggregated according to World Bank regions and by income level. The database also Production Data reports data on trade flows for particularly interest- The production data are collected by UNIDO and ing markets such as the EU, Japan, and the United OECD through the joint annual collection program States, as well as world totals, producing a total of 34 for general industrial statistics and are published in regional groups.8 the UNIDO annual commercial publication, the The trade data are quite complete; there are very International Yearbook of Industrial Statistics. few country periods for which data are missing. UNIDO provides internationally consistent data by When the data are not available, the missing obser- collecting annual data directly from all non-OECD vations are reported as dots or blanks. All trade val- member countries through its country question- ues are reported in thousands of U.S. dollars. naire. The OECD collects data for its member states and provides the information to UNIDO. The data Tariff Data are usually obtained from industrial census statis- tics and then compiled into ISIC categories. The The tariff data utilized in the database originate industrial data cover only the manufacturing sector from two sources: the WTO and UNCTAD. The and are published at two different levels of detail. WTO tariff data are from published Trade Policy The three-digit level of aggregation covers 28 man- Review (TPR) reports and from the Internet version ufacturing sectors; the four-digit level covers 81 of the IDB. The published TPR data, consisting of manufacturing sectors. For each sector, the data on tariff averages in ISIC (2) nomenclature, were coded production report yearly values in thousands of manually by Bank staff into the database. The raw U.S. dollars for total output, value added, gross fixed data were converted to ISIC (2) using a concordance capital formation, and average wages. Values for the developed in the WTO Trade Policy Review Divi- other variables--number of enterprises, total num- sion. The tariff averages may include ad valorem ber of employees, and number of female employ- equivalents; in cases where no ad valorem equiva- ees--are expressed in units. The data published by lent was available, the ad valorem part of compound UNIDO are not complete across all years or indus- duties was used, but only for those duties for which tries; missing observations are reported as a dot or a the ad valorem part is always lower than or equal to blank. the total value of the duty. In addition, only out-of- quota duties have been retained in the calculations. The IDB data are expressed in HS nomenclature Trade Data at the national tariff line level (that is, more detailed The trade data are collected and organized by the than the six-digit level). The tariff data may contain United Nations Statistical Department and are ad valorem equivalents if these were provided by the reported in the COMTRADE database. For the pur- reporting country, and, depending on how the data pose of the Trade and Production Database, the are submitted by the reporting country, in-quota data are first downloaded in the SITC rev. 2 classifi- duties may have been included in the averages. Con- cation and are then transformed into the ISIC clas- version of the IDB information from the HS to the sification. This process utilizes the concordance ISIC and the calculation of tariff averages were filters developed by the OECD and yields two undertaken by the World Bank. The methodology slightly different concordance tables, one for employed and the soundness and accuracy of the exports and one for imports. These tables do not results are obviously not the responsibility of the follow a one-to-one correspondence, but matching WTO. The ISIC concordance and the methodology is achieved through a method that involves a series for calculating tariff averages used by the Bank for of carefully estimated weights. The Trade and Pro- the IDB data are not strictly comparable to the duction Database is balanced and reports values for methodology behind the tariff averages and the imports and exports. Data on mirrored exports ISIC conversion published by the WTO Trade Policy (exports calculated using import data reported by Review Division. 590 Appendix C: A Dataset on Trade and Production, 1976­99 Tariff data are also available through the TRAINS manufacturing sector that is required to produce one database maintained by UNCTAD. TRAINS is a unit of output in each manufacturing sector. Using comprehensive computerized information system this GTAP data, the Trade and Production Database at the HS-based tariff line level that covers tariff, provides an intermediate import-share table that paratariff, and nontariff measures, as well as import demonstrates the import share of intermediates uti- flows by origin, for more than 100 countries. In the lized by each sector in each country. These tables are best cases the TRAINS data begin in the late 1980s. available only in EXCEL format. UNCTAD reports the tariff data utilizing the six- digit HS classification. The conversion from the HS Problems and Special Considerations classification into ISIC was achieved using a one-to- one concordance table. TRAINS data are far from The data in the Trade and Production Database complete. Although there are only a few countries have been grouped and organized to facilitate the for which no tariff data are available, the time-series use of the database for a large number of purposes. are quite sparse. The database is not designed to produce quick The Trade and Production Database may include answers but rather to assist researchers in the tariff data for a particular country from UNCTAD lengthy and cumbersome exercise of collecting and and the WTO. Discrepancies may occur due to dif- organizing data. In order to give the researcher the ferences in the concordances and methodologies maximum degree of flexibility, the data have not used by the WTO Trade Policy Review Division, the been changed beyond the adaptations described World Bank in converting the raw data in the WTO above. Nevertheless, a few points need to be empha- IDB files, and UNCTAD. For example, UNCTAD sized. Monetary data are not deflated and are calculates the simple averages using as the denomi- expressed in thousands of U.S. dollars. If the data nator only the actual number of dutiable lines, but were not supplied in U.S. dollars, the common prac- the WTO Trade Policy Review Division includes all tice was followed of using the yearly average lines. When filtering the IDB tariffs, we followed the exchange rate to convert the domestic currency to TPR approach. The tariff data reported in the Trade the dollar.10 Caution should also be used when ana- and Production Database are most-favored-nation lyzing data gathered from Germany (GER).11 To simple averages at the three- or four-digit level of produce a consistent time series, data on Germany the ISIC. before and immediately after unification are con- structed as the sum of the Federal Republic of Ger- many (DFA) and the Democratic Republic of Input-Output Tables Germany (DDR). Especially for the years immedi- Input and output tables are based on the Global ately after unification, it is possible that some data Trade Analysis Project (GTAP) database, version 4. on the DDR were not reported, thus producing a The GTAP database utilizes data from the early 1990s sudden shift in the time-series. The production data in constructing its input and output tables. Only one from UNIDO are subject to differences in national table is provided for each country in the database. classifications, and assumptions are needed to con- The GTAP aggregates some countries in regions; vert from the national (country-specific) industrial those countries therefore have the same input and classification into the ISIC classification. For exam- output tables.9 Those interested in how the GTAP ple, similar industries may be allocated to slightly constructs each particular input/output table may different ISIC sectors in different countries, or data refer to the GTAP database or consult the GTAP from industries of the same size may be too small to Website, . The tables reported in the be reported in some countries, while they may be Trade and Production Database are aggregated at the fully reported in others. These kinds of problem are three-digit level of the ISIC classification. To facilitate generally most pronounced at the more disaggre- the use of the tables within the database, the data gated level.12 reflecting input and output have been broken down A common issue in trade data is the presence of into two tables. The first table reports the share of the label "not classified" for a trade partner. This is output from each manufacturing sector that is sold the case when the country does not know, or does to other industrial sectors, and the second table not wish to disclose, the origin or the destination of reports the value of intermediate inputs from each a trade flow. The Trade and Production Database 591 D E V E L O P M E N T, T R A D E , A N D T H E W T O deals with this issue by reporting data on "not clas- Notes sified" countries as separate observations instead of 1 The Trade and Production Database is also available at allocating them to any region. Nevertheless, the . It is provided on an "as "not classified" value may be negative due to con- is" basis. cordance aggregation between the SITC and the 2 Feenstra (1996) covers U.S. imports from 1972 to 1974. Feen- ISIC classifications. In this case, the negative value stra, Lipsey, and Bowen (1997) covers trade, production, and is split across all regions according to weights cal- tariff data from 1970 to 1992, but production data are only culated on the basis of existing documented trade. available for OECD countries. It does include nontariff barriers This problem affects only a minimal part of the (coverage ratios), which are excluded from the Trade and Pro- data and is equivalent to assuming that exports or duction Database. Feenstra (2000) covers only trade, but at a much higher level of product disaggregation than the one fol- imports not classified by country of destination or lowed here. origin are distributed to each region using as weight the documented trade flows.13 In the even 3 For more information on UNIDO's industrial databases, see . rarer case that, after the transformation from SITC to ISIC, the documented trade flows with the 4 For more information on the COMTRADE database and other regions turns out to be negative, the value of "not products of the United Nations Statistical office, see . classified" is diminished by the amount of the sum of those negative values, and the negative values are 5 For more information on WTO products, see set to zero. These operations are performed only on . For more information on UNCTAD's the data aggregated by region and income level. To TRAINS database, see . give maximum flexibility to researchers, the reallo- cation of the negative values is not performed in 6 For more information on the GTAP dataset, see the ASCII file reporting country-by-country trade . flows. 7 Generally, import data are of better quality than export data, Another recurring issue with trade data, as dis- for fiscal reasons. Mirrored data, however, need to be used cussed in detail by Feenstra (1996) and by Hanson with caution, as noted by Yeats (1995). and Feenstra (2001), is the existence of entrepôts-- 8 Because of rounding errors and aggregation issues, there may that is, countries where transits of trade flows take be very slight differences (usually less than 0.01 percent) place but that are not the origin or the final destina- between the sums of the different regions and totals. Users should be aware of these possible discrepancies. tion of the flows. In many cases the country of ori- gin (O) reports the entrepôt (T) as the destination 9 If the input and output tables are not country-specific, the of the shipment. Meanwhile, the entrepôt country name of the region is also reported in each table. does not report the import, and the final importer 10 The researcher should keep this practice in mind and treat (F) reports the original exporter (O) as the origin. with caution cases where there has been a large and sudden This creates a surplus (between O and T) and a change in the exchange rate. deficit (between O and F). In the example above, 11 In other databases the country code identifying Germany may country (F) reports an import from (O) that is not be DEU instead of the one used here, GER. reported (as an export to F) by country (O), creat- 12 Note that the four-digit codes starting with 312 are commonly ing a discrepancy. The researcher should keep this in collapsed into the three-digit category 311. mind when analyzing entrepôts such as Hong Kong 13 Negative values were encountered in less than 0.01 percent of (China), Macau (China), Singapore, and the the observations. Netherlands. For this reason, trade data also include values of mirrored exports.14 In many cases there 14 Mirrored data are available from 1980 to 1998. Many coun- tries failed to report trade data before 1980 and for 1999, pro- are huge discrepancies that are attributable to rea- ducing incomplete results for those years. sons such as transport costs, different product clas- sifications, entrepôts, and poor accounting 15 In some cases, however, mirrored exports may be considered methods. It is advisable to use mirrored exports more precise than exports because trade flows are usually bet- ter recorded at entrance (imports). Therefore, mirrored exports only where there are serious doubts about the capa- contain useful and utilizable information, on a bilateral basis, bility of the reporting country in managing the col- where the partner countries have good customs administra- lection of records on trade flows.15 tions. 592 Glossary of Trade-Related Terms ACP African, Caribbean, and Pacific countries, a oped and implemented by UNCTAD that covers group of mostly former European colonies. most foreign trade procedures and handles mani- Actionable subsidy A type of subsidy that is not fests and customs declarations, accounting proce- prohibited under WTO rules but against which a dures, transit and suspense procedures. member may respond by imposing a countervailing Binding See Tariff concession. duty. Border tax adjustment Fiscal measure compensat- Administered protection See Contingent protec- ing, in whole or in part, for the different treatment tion. either between imports and similar domestic prod- Ad valorem An ad valorem duty (tariff, charge, and ucts or between exports and similar products sold so on) is based on the value of the dutiable item and on the domestic market. For example, refunds of expressed in percentage terms: for example, a duty domestic indirect taxes on goods destined for of 20 percent on the value of automobiles. export; or charges on imports similar to the taxes Advisory Centre on WTO Law Entity based in levied on like domestic products. Also see Duty Geneva that provides legal counseling on WTO law drawback. and dispute settlement to developing and transition Cairns Group Coalition of developing and industri- countries that are WTO members on a subsidized al country exporters of agricultural commodities basis, depending on the income level of the request- formed in the Uruguay Round to negotiate stronger ing government; 72 WTO members are eligible to multilateral disciplines on agricultural trade poli- request assistance. cies. AGOA (African Growth and Opportunities Act) Capacity building In a trade context, activities sup- U.S. legislation providing duty-free access for a large ported by the donor community aimed at strength- number of products for 35 African economies. ening the ability of stakeholders in developing Aggregate measure of support Measure of the countries to develop national trade policy, under- total support given to an activity as a result of poli- take analysis, and identify their interests in interna- cies such as production subsidies and market price tional trade negotiations. support policies. Used in the WTO Agreement on Cartel Arrangement between firms to control a Agriculture. market--for example, to fix prices or limit competi- Antidumping Trade policy used by importing gov- tion between members of the cartel. ernments to counteract dumping, for example by Ceiling binding Often used to describe a situation imposing duties or negotiating price increases. where there is a large difference between the tariff Appellate Body WTO body that hears appeals that is actually applied and the level at which the against the findings of dispute settlement panels. tariff is bound in GATT (the "ceiling"). ASYCUDA Automated System for Customs Data. A C.I.F. Cost, insurance, and freight. The cost of a computerized customs management system devel- good delivered to the importing country's port. The Glossary draws on material in Finger and Olechowski (1987) and the glossary maintained by Professor Alan Deardorff on his home page: http://www-personal.umich.edu/~alandear/glossary/. Chapter 55 contains a listing of acronyms of regional trade agreements. Many of the terms are discussed further in the accompanying CD-ROM on applied trade policy for developing countries. 593 D E V E L O P M E N T, T R A D E , A N D T H E W T O CITES Convention on International Trade in June 2000 in Cotonou, Benin. Replaces the Lomé Endangered Species of Wild Fauna and Flora. Convention. Its main objective is poverty reduction, Establishes rules for trading such species, up to and "to be achieved through political dialogue, develop- including a complete ban on all trade. ment aid and closer economic and trade coopera- Codex Alimentarius Commission The "food tion." code"--an international set of standards, codes of Counter trade Form of barter committing the practice, and guidelines and recommendations exporter to offset the value of his exports, in whole relating to food quality and safety, including codes or in part, by imports from his trading partner. Also governing hygienic processing practices, recom- see Offset requirement. mendations relating to compliance with standards, Countervailing duty Duty levied on imports of limits for pesticide residues, and guidelines for con- goods that have benefited from production or taminants, food additives, and veterinary drugs. export subsidies. The duty is intended to offset the The Codex Alimentarius Commission is the body effect of the subsidy. responsible for compiling the standards. Credit (for autonomous liberalization) Mecha- Compensatory adjustment Measure taken, after nism through which developing countries are withdrawing a (tariff or other) concession, to com- granted recognition in WTO talks for unilateral lib- pensate for such withdrawal (GATT Art. XXVIII). eralization of the trade regime that has occurred in Competition policy Legislation and regulations the period before negotiations commence. Past designed to protect and stimulate competition in efforts by developing countries to establish such a markets by outlawing anticompetitive business mechanism were not successful. practices such as cartels, market sharing, or price Customs duty Charge levied on imports and listed fixing. in an importing country's tariff schedules. Duties Computable general equilibrium (CGE) models may be specific or ad valorem or a combination of Mathematical characterizations of the economy, the two (ad valorem with a specific minimum or the used to predict the impact of policy changes, taking greater of the two). into account both direct effects as well as indirect Customs union A group of countries forming a sin- effects that work through labor and other markets. gle customs territory in which (1) tariffs and other Concertina approach Method of reducing tariffs barriers are eliminated on substantially all the trade by lowering the highest rates first, then the next between the constituent countries for products highest, and so forth. originating in these countries, and (2) there is a Content, domestic or local Rules establishing a common external trade policy (common external minimum proportion (by value or volume) of a tariff) that applies to nonmembers. product that must be domestically or locally pro- Customs valuation Establishment, according to duced in order to obtain a benefit (for example, a tar- defined criteria, of the value of goods for the pur- iff concession). pose of levying ad valorem customs duties on their Contestability A market is contestable if new sup- importation. pliers can enter it easily. The threat of such entry is a Decoupling Action to ensure that subsidies to pro- discipline on the incumbent suppliers and can pre- ducers (usually farmers) are unrelated to produc- vent prices from rising far above costs, because any tion so as to provide no incentive to increase excess profits will be rapidly followed by entry. production; in contrast, simple subsidies per unit of Contingent protection Trade barriers that are output tend to increase production. imposed if certain circumstances (contingencies) Deep integration Intergovernmental cooperation are met. Examples include antidumping or counter- in designing and applying domestic policies such as vailing duties (to offset subsidies) and safeguards. taxes, health and safety regulations, and environ- Also called administered protection. mental standards. May involve either harmoniza- Copyright Instrument to protect the rights of tion of policies or mutual recognition; generally authors of original works (print, audio, video, film, occurs in the context of regional integration agree- software) from unauthorized copying and use. Gen- ments. erally for the life of the author, plus 50 years. Deficiency payment Direct monetary payment by Cotonou Agreement Partnership agreement government to producers to compensate for the dif- between the EU and the ACP countries signed in ference between the market price of a good and a 594 Glossary higher guaranteed price for that good in the case of, without according such treatment to other con- say, low international commodity prices. tracting parties." See also Generalized system of Degressivity Mechanism to ensure that the applica- preferences. tion of a measure gradually becomes less severe over Escape clause Clause in a legal text allowing tempo- time. For example, a tariff set at 50 percent that is rary derogation from its provisions under certain reduced by 10 percentage points each year and specified emergency conditions. See also Safeguard becomes zero in year 5. Action (GATT Art. XIX.) Differential and more favorable treatment See Europe Agreement Free trade agreement between Special and differential treatment and Enabling the EU and various Central and Eastern European clause. countries. Dispute Settlement Body WTO body that is Everything but Arms A 2001 EU initiative to grant responsible for dealing with disputes between WTO least developed countries duty- and quota-free members. Consists of all WTO members meeting access for their exports. together to consider the reports of dispute settle- Exchange control Restrictions imposed by a gov- ment panels and the Appellate Body. ernment or central bank over the holding, sale, or Domestic content See Content. purchase of foreign exchange. Typically used when Dumping A form of price discrimination by which the exchange rate is fixed and the central bank is the export price of the product exported from one unable or unwilling to enforce the rate by exchange- country to another is less than the comparable price, market intervention. in the ordinary course of trade--that is, including Exhaustion Policy stance of a country regarding transport and related costs--for the like product parallel imports of goods protected under intellec- when destined for consumption in the exporting tual property rights. Under national exhaustion, country (GATT Art. VI). Also defined as sales below rights end upon the first sale of the good within a the estimated cost of production. The margin of nation, and right holders may prevent unauthorized dumping is the difference between the two prices. imports of the goods concerned. Under interna- Duty drawback scheme A duty drawback scheme tional exhaustion, rights end upon the first sale any- (often administratively demanding) is a form of where in the world, after which parallel imports are border tax adjustment whereby the duties or taxes permitted. levied on imported goods are refunded, in whole or Export-processing zone (EPZ) A designated area or in part, when the goods are re-exported. The idea is region in which firms can import duty-free as long to reduce the burden on exporters while maintain- as the imports are used as inputs into the produc- ing tariffs for revenue or protective purposes. See tion of exports. Traditional EPZs are fenced-in also Temporary admission. industrial estates specializing in manufacturing for Economic needs test Measure requiring a demon- exports. Modern ones have flexible rules that may stration that an import (of goods but more usually permit domestic sales upon payment of duties natural service providers) cannot be satisfied by when leaving the zone. EPZs generally also provide local producers or service providers. a liberal regulatory environment for the firms Effective rate of protection A measure of the pro- involved as well as infrastructure services. tection afforded by an import restriction calculated Export promotion A strategy for economic devel- as a percentage of the value added in the product opment that emphasizes support for exports concerned. Takes into account the protection on through removal of anti-export biases created by output and the cost-raising effects of protection on policy. May be associated with policies such as duty inputs. drawbacks, export subsidies, marketing support, or Emergency action See Safeguard action. matching grants for exporters. Enabling clause 1971 GATT Decision on "Differen- Externality Occurs when the action of one agent tial and More Favorable Treatment, Reciprocity and (person, firm, government) affects directly other Fuller Participation of Developing Countries". One agents, making them better or worse off. Beneficial of the so-called Framework agreements, it enables effects are called positive externalities; harmful ones WTO members, notwithstanding the nondiscrimi- negative externalities. nation requirements, to "accord differential and Fast track A procedure under which the U.S. Con- more favorable treatment to developing countries, gress agrees to consider implementing legislation for 595 D E V E L O P M E N T, T R A D E , A N D T H E W T O international trade agreements on an "up or down" ket countries to promote international financial sta- basis, that is, gives up its right to propose amend- bility. The Managing Director of the IMF, the Presi- ments. Now called Trade promotion authority. dent of the World Bank, and the Chairpersons of Foreign trade zone An area within a country where the International Monetary and Financial Commit- imported goods can be stored or processed without tee and Development Committee of the IMF and being subject to import duty. Also called a free zone, World Bank participate in G-20 deliberations. free port, or bonded warehouse. See also Export- G-24 Established in 1971, an intergovernmental processing zone. group of 24 developing countries that has the objec- Formula approach Method of negotiating down tive of concerting the position of the developing tariffs or other barriers to trade by applying a gener- countries on monetary and development finance al rule (formula). For example, a rule specifying that issues. The only formal developing country group- all tariffs are to be cut to a certain fraction of their ing within the IMF and World Bank. Meets twice a initial level, or that an agreement should cover a cer- year, preceding the Spring and Fall meetings of the tain proportion of economic activity (sectors). two institutions. Framework Agreements The GATT "Agreements G-77 A coalition of developing countries within the Relating to the Framework for the Conduct of United Nations, established in 1964 at the end of the International Trade," resulting from the Tokyo first session of UNCTAD, intended to articulate and Round: (1) Differential and More Favorable Treat- promote the collective economic interests of its mem- ment, Reciprocity and Fuller Participation of Devel- bers and enhance their negotiating capacity. Original- oping Countries (the "Enabling Clause"); (2) ly with 77 members, it now (in 2002) has 133. Declaration on Trade Measures Taken for Balance- GATT General Agreement on Tariffs and Trade. of-Payments Purposes; (3) Safeguard Action for Major Articles: Development purposes; and (4) Understanding I General MFN requirement. Regarding Notification, Consultation, Dispute Set- II Tariff schedules (bindings). tlement and Surveillance. III National treatment. Free on board (f.o.b.) The price of a traded good V Freedom of transit of goods. including its value and the costs associated with VI Allows antidumping and counter- loading it on a ship or an aircraft, but excluding vailing duties. Superseded by the international transportation (freight) costs, insur- GATT 1994 Agreement on ance, and payments for other services involved in Antidumping, and the Agreement moving the good to the point of final consumption. on Subsidies and Countervailing Free-trade area A group of countries in which the Measures. tariffs and other barriers are eliminated on substan- VII Requires that valuation of goods for tially all trade between them. Each member main- customs purposes be based on actual tains its own external trade policy against value. Superseded by the GATT 1994 nonmembers. Also called free trade agreement or Agreement on the Implementation free trade arrangement. Contrasts with custom of Article VII. union. VIII Requires that fees connected with G-7 A group of seven major industrial countries import and export formalities be whose heads of state have met annually since 1976 cost-based. in summit meetings to discuss economic and politi- IX Reaffirms MFN for labeling require- cal issues. The seven are Canada, France, Germany, ments and calls for cooperation to Italy, Japan, the United Kingdom, and the United prevent abuse of trade names. States. X Obligation to publish trade laws and G-8 The G-7 plus Russia, which have met as a full regulations; complemented by the economic and political summit since 1998. WTO's Trade Policy Review Mecha- G-20 International forum of finance ministers and nism and numerous notification central bank governors representing 19 countries requirements in specific WTO agree- plus the EU. Created in 1999 by the G-7 with the ments. aim of promoting discussion, study, and review of XI Requires the general elimination of policy issues among industrial and emerging mar- quantitative restrictions. 596 Glossary XII Permits trade restrictions if neces- II MFN obligation. Option to invoke sary to safeguard the balance of pay- exemptions on a one-time basis. ments. III Notification and publication. Oblig- XIII Requires that quotas be administered ation to create an enquiry point. in a nondiscriminatory manner. IV Increasing participation of develop- XVI Established GATT 1947 rules on ing countries. High income coun- subsidies. Complemented by the tries to take measures to facilitate WTO Agreement on Subsidies and trade of developing nations. Countervailing Measures. V Economic integration. Allows for XVII Requires that state trading enterpris- free trade and similar agreements. es follow MFN. VI Allows for domestic regulation. XVIII Allows developing countries to Requirements concerning the design restrict trade to promote infant and implementation of service sec- industries and to protect the bal- tor regulation, including in particu- ance-of-payments (imposing weaker lar qualification requirements. conditionality than Article XII). VII Recognition of qualifications, stan- XIX Allows for emergency action to dards and certification of suppliers. restrict imports of particular prod- VIII Monopolies and exclusive suppliers. ucts if these cause serious injury to Requires that such entities abide by the domestic industry. Comple- MFN and specific commitments mented by the WTO Agreement on (Articles XVI and XVII) and do not Safeguards. abuse their dominant position. XX General exceptions provision-- IX Business practices. Recognition that allows trade restrictions if necessary business practices may restrict trade. to attain non-economic objectives Call for consultations between (health, safety). members on request. XXI Allows trade to be restricted if neces- XIV General exceptions. Allows measures sary for national security reasons. to achieve non-economic objectives. XXII Requires consultations between par- XVI Market access. Defines a set of poli- ties involved in trade disputes. cies that may only be used to restrict XXIII GATT's main dispute settlement pro- market access for a scheduled sector vision, providing for violation and if they are listed in a member's spe- non-violation complaints. Comple- cific commitments. mented by the WTO Understanding XVII National treatment. Applies in a sec- on Rules and Procedures Governing tor if a commitment to that effect is the Settlement of Disputes. made and no limitations or excep- XXIV Sets out the conditions under which tions are listed in a member's sched- the formation of free trade areas or ule. customs unions is permitted. XIX Calls for successive negotiations to XXVIII Allows for renegotiation of tariff expand coverage of specific commit- concessions. ments (Articles XVI and XVII). XXVIII bis Calls for periodic rounds of negotia- Generalized System of Preferences (GSP) The tions to reduce tariffs. GSP is a system through which industrialized high- XXXIII Allows for accession income countries grant preferential access to their Part IV Calls for more favorable and differ- markets to developing countries. Also called Gener- ential treatment of developing coun- alized System of Trade Preferences. tries. Entered into force in June 1966. Geographical indication Measure intended to pro- GATS General Agreement on Trade in Services. tect the reputation for quality of goods originating Major Articles : in a particular geographic location by limiting the I Definition. Trade in services covers use of distinctive place names or regional appella- all four modes of supply. tions to goods actually produced in those locations. 597 D E V E L O P M E N T, T R A D E , A N D T H E W T O Government procurement Purchasing, leasing, financed trust fund managed by six agencies--IMF, rental, or hire purchasing by government entities or International Trade Centre (Geneva), UNCTAD, agencies. UNDP, World Bank, and WTO--to work with Graduation Concept linking the rights and obliga- LDCs to undertake diagnostic studies aimed at tions of a developing country to its level of develop- assisting countries to identify key constraints to bet- ment. Referred to in the WTO Trade Policy Review ter integration into the world economy and to pro- Mechanism. Generally used in the context of GSP vide follow-up trade-related technical and financial and similar types of preferential treatment of low- assistance. income countries as a mechanism or set of criteria Intra-industry trade Trade in which a country both to determine when countries cease to be eligible for exports and imports goods that are classified to be preferences. in the same industry. Grandfather clause A clause exempting signatories Labeling Requirement, either mandatory or volun- from certain treaty obligations for legislation or reg- tary, to specify whether a product satisfies certain ulations that were adopted before accession to the conditions relating to the process by which it was treaty and that are inconsistent with the treaty. produced or its characteristics. Gray-area measure Measure whose conformity Least developed country (LDC) A country that with contractual obligations is unclear: for example, satisfies a number of criteria established by the voluntary export restraints under pre-WTO rules of United Nations that together imply a very low level the GATT. of economic development. As of 2002, the UN had Green room Used to describe discussions in the classified 49 countries in the LDC group. Used in WTO among a subset of countries, generally the the WTO Subsidies Agreement, where LDCs are major OECD members and a small number of granted differential treatment. developing countries. Licensing (of imports or exports) Practice GTAP The Global Trade Analysis Project, based at requiring approval to be granted by the relevant Purdue University in the United States. It provides government authority, or by a body designated data and models for computable general equilibri- by such authority, as a prior condition to um modeling. importing or exporting. Harmonized System (HS) "Harmonized Com- Automatic licensing Where approval is freely modity Description and Coding System". Nomen- granted--for example, licensing for keeping sta- clature developed by the World Customs tistical records. Organization for customs tariffs and international Nonautomatic licensing Where approval is not trade statistics. freely granted. This may be used as a restriction HIPC Heavily Indebted Poor Countries Initiative. An itself, or it may be used to administer a quota. The agreement among official creditors to help the most license may be subject to certain conditions: for heavily indebted countries to obtain debt relief. example, a requirement to export; the use to which Impairment Damage to, or weakening of, benefits the imported good is to be put; the purchase of a accruing under contractual rights and obligations. specified quantity of the domestically produced (GATT Art. XXIII). like product; or the availability on the domestic Import substitution Theory of and approach to market of the domestically produced like product. development that focuses on providing domestic Discretionary licensing Nonautomatic licensing substitutes for all imported manufactures via trade (see above). protection and various types of industrial policies. Linking scheme An import-licensing requirement Infant industry Infant industry arguments suggest that forces an importer to purchase specified that new (non-traditional) industries must be pro- amounts of the same type of product from domes- tected from import competition while they are tic producers before they can apply for import establishing themselves. This is a so-called second- licenses. An example is a two-tier quota allocation best argument in that it does not address the funda- system for licenses in which obtaining a license to mental market failures that cause industries to fail buy or sell on a market is linked to the amount to develop (such as financial market imperfections). bought or sold in a second market. Integrated Framework for Trade-related Techni- Local (or domestic) content requirements See cal Assistance (IF) Joint activity and donor- Content. 598 Glossary Lomé Convention This agreement was between the Most favored nation (MFN) principle MFN is the EU and the ACP countries on trade concessions "normal," non-discriminatory tariff charged on (GSP treatment), development aid, and general imports of a good. In commercial diplomacy, cooperation. Replaced by the Cotonou Agreement exporters seek MFN treatment that is, the promise in 2000. that they will be treated as well as the most favored MAI Multilateral Agreement on Investment. Effort exporter. Called normal trade relations in the Unit- by the OECD in the late 1990s to establish a set of ed States. disciplines on investment-related matters. Negotia- Multifiber Arrangement (MFA) "Arrangement tions failed and were suspended in 1998. Regarding International Trade in Textiles." Negoti- Market access Refers to the conditions under which ated as a temporary exception to the GATT in 1973. imports compete with domestically produced sub- Regulates trade in certain textile products between stitutes. These are determined by the extent to signatories by means of negotiated bilateral quotas. which foreign goods are confronted with tariffs, dis- Superceded by the WTO Agreement on Textiles and criminatory taxes, and other regulations. Clothing in 1995, which specifies that all quotas are Matching grant Subsidy that is conditional on a to be abolished by 2005. co-payment or contribution by an industry or Mutual recognition The acceptance by one country enterprise. of another country's certification that a product has Maximum (minimum) price system (for imports) satisfied a product standard. Often based on formal Price(s) decreed by the authorities of the importing agreements between countries if the standards are country and above (below) which price(s) imports mandatory. may not enter the domestic market. Actual import National treatment Principle that foreign goods, prices below the decreed minimums trigger a pro- services, and persons (investors), once they have tective action, such as the imposition of additional entered a country and satisfied any formalities that duties or of a quantitative restriction. Different are required, are treated in exactly the same way as terms are used in different countries and different national goods, services, or persons. In particular, sectors: basic import price, minimum import price, they face the same internal taxes and no additional reference price, and trigger price. restrictions. Markup In economics, a measure of the difference Necessity test Procedure to determine whether a between the unit price of a good and its marginal policy restricting trade is necessary to achieve the cost of production. In WTO terms sometimes used objective that the measure is intended to attain. to indicate the extent to which an applied tariff Negative list In an international agreement, a list of exceeds the bound rate. those items, entities, products, and so on to which Mercantilism An economic philosophy of the 16th the agreement will not apply, the commitment and 17th centuries that international commerce being to apply the agreement to everything else. should primarily serve to increase a country's finan- Contrasts with positive list. cial wealth, especially of gold and foreign currency. Nominal rate of protection The proportion by To that end, exports are viewed as desirable and which the (tariff-inclusive) internal price of an imports as undesirable unless they lead to even import exceeds the border or world price. See also greater exports. In a WTO context, the term is often Effective rate of protection. used to describe the quid pro quo nature of bar- Noneconomic objective Describes situations gaining over trade policies. where a policy objective is other than the efficient Mixing regulation Describes two kinds of prac- allocation of resources. In the trade policy setting, it tices: (1) regulation specifying the proportion of refers to the view that a restriction on imports may domestically produced content in products offered serve a purpose that goes beyond the restriction of for sale on the domestic market; (2) regulation trade itself. In general desired changes in output, specifying, for any imports of a given product, the consumption, and so forth can be achieved at lower quantity of a domestically produced like product economic cost through other types of policies. that must be purchased by the importer. Nontariff barrier (NTB) A catchall phrase describ- Mode of supply Term used in the GATS context to ing barriers to international trade other than the identify how a service is provided by a supplier to a tariffs--for example, quotas, licensing, or voluntary buyer. export restraints. 599 D E V E L O P M E N T, T R A D E , A N D T H E W T O Nontariff measure Any government action with a Patent A right granted to its owner to exclude all potential effect on the value, volume, or direction of others from making, selling, importing or using the trade. Also see Nontariff barrier. product or process described in the patent for a Nonviolation Procedure under WTO disputed set- fixed period of time, generally 20 years. To be tlement provisions under which a WTO member patentable, inventions have to be novel, non-obvi- argues that actions by another member, even ous, and be useful or have industrial applicability. though allowed under WTO rules, nullify or impair Phytosanitary regulation Pertaining to the health benefits expected under the agreement. of plants. See SPS measure. Normal value Price charged by an exporting firm in Plurilateral agreement In WTO, an agreement to its home market. Used to compare with the price which membership is voluntary, dealing with an charged by the firm on an export market to deter- issue that is not covered by the WTO. In 2002 there mine if there is dumping. (GATT Art. VI). See also were two plurilateral agreements--on civil aircraft Dumping. policies and government procurement. Nullification Negation of benefits accruing under Positive list In an international agreement, a list of the WTO as a result of actions taken by a member. those items, entities, products, and so on to which See also Dispute Settlement Body and Panel. the agreement will apply, with no commitment to Offset requirement Requirement, stipulated by the apply the agreement to anything else. authorities of the importing country, that exporters Poverty Reduction Strategy Paper (PRSP) Docu- to that country compensate for their exports by, say, ment describing a country's macroeconomic, struc- purchasing products of the importing country or tural and social policies, and programs to promote investing in the importing country. Also see growth and reduce poverty, as well as associated Counter trade. external financing needs. PRSPs are prepared by Orderly marketing arrangement See Voluntary governments through a participatory process export restraint. involving civil society and development partners, Origin rule Criterion for establishing the country of including the World Bank and the IMF, and provide origin of a product. Often based on whether pro- the basis for concessional lending and debt relief duction (processing) leads to a change in tariff under the enhanced HIPC Initiative. heading (classification) or on the level of value PPM Production and processing method. Used in added in the country where the good was last instances where trade policy action by a country is processed. motivated by a desire to ensure that imports have Panel In WTO, a group of three independent been produced in a way that satisfies a national or experts nominated by the WTO secretariat from a international production or process norm. Often roster approved by members that is responsible for these norms will be environmental in nature. determining the validity of allegations brought by Precautionary principle Policy under which meas- one WTO member against another claiming nullifi- ures are motivated by the possibility that use of cer- cation or impairment of rights or obligations (that tain technologies (for example, biotechnology, is, violation of WTO rules and disciplines). genetically modified organisms, and pesticides) Parallel imports Trade that is made possible when a could be harmful to human or animal health and good that is protected under intellectual property safety or the environment, although there is no cer- provisions (patents, copyrights) is sold in different tainty to that effect. countries for different prices. A parallel import Predatory pricing Action by a firm to lower prices comprises arbitrage activity and occurs when so much that rival firms are driven out of business, traders import the good from a lower-price market after which the firm raises prices again to exploit the into a higher-price country. resulting monopoly power. Para tariff Charges on imports that act as a tariff but Preference Preferential treatment. In GATT terms, are not included in a country's tariff schedule. Exam- this represents a derogation, in the sense of treat- ples include a statistical tax, stamp fees, and so forth. ment that is more favorable than MFN. See also Partial equilibrium analysis The study of one mar- Generalized System of Preferences and Special and ket in isolation, assuming that anything that hap- Differential Treatment. pens in it does not materially affect any other Preshipment inspection Mechanism under which market. goods are inspected and certified in the country of 600 Glossary origin by specialized inspection agencies or firms. without regard to the country or countries of origin Often used by importing governments to combat (destination) of the product. over- or underinvoicing of imports by having the Bilateral quota Quota applied to imports from value of consignments determined by independent (exports to) a specific country. entities (firms). Quota by country Quota that not only specifies the PRSP See Poverty Reduction Strategy Paper. total volume, or value, of the product to be import- Price discrimination The practice of charging dif- ed (exported), but also allocates the trade between ferent customers different prices for the same good the various countries of origin (destination). in order to exploit their different degrees of enthusi- Quota rent The economic rent received by the holder asm for it--for example, lower off-peak fares of a right to import under a quota. Equals the domes- exploit workers' need to travel in the rush hour, tic price of the imported good, net of any tariff, minus while allowing less urgent personal travel to take the world price, times the quantity of imports. place at other times. When this occurs internation- Real exchange rate The nominal exchange rate ally and the lower price is charged for export, it is adjusted for inflation. Unlike most other real vari- called dumping. ables, this adjustment requires accounting for price Price undertaking Commitment by an exporter to levels in two currencies. The real exchange rate is: R either raise prices or reduce sales in a market as a = EP*/P where E is the nominal domestic-currency way of settling an antidumping suit brought by price of foreign currency, P is the domestic price import-competing domestic firms. Generally has an level, and P* is the foreign price level. Equivalent to effect analogous to a quota. the real price of foreign goods; that is, the quantity Principal-supplier rule Rule, in bilateral negotiating of domestic goods needed to purchase a unit of for- procedures, according to which an import conces- eign goods. Also defined as the relative price of trad- sion on a specific product is to be negotiated only ed goods in terms of non-traded goods. with the country that is actually or potentially the Reference price See Maximum (minimum) price major supplier of that product. Note that the WTO system. MFN rule requires that the concession be extended Remedy Legal term to describe a measure recom- to all other members. mended by a WTO dispute settlement panel that Prisoner's dilemma A situation where agents with aims to bring the policies of a member found to perfect information that act rationally (that is, pur- have violated WTO rules or disciplines into compli- sue their "selfish" best interests) are confronted with ance with its obligations. a set of payoffs (or rewards) in which not cooperat- Rent-seeking Refers to activities that use resources ing is the dominant strategy, even though coopera- to obtain incomes through transfers but which do tion would in principle increase their joint payoffs. not increase national income. Such activities result Producer subsidy equivalent A measure of the in an extra cost to society (the loss of income from aggregate value of the gross transfers from con- the diversion of resources away from productive sumers and taxpayers to farmers due to policy towards rent-seeking activities) beyond the distor- measures. Also called producer support estimate. tionary costs associated with measures that give rise Protocol of accession Legal document recording to the rents. the conditions and obligations under which a coun- Request-offer procedure Negotiating procedure try accedes to an international agreement or organ- based on the tabling, by each party, of a list of con- ization. cessions requested of other parties, followed by an Quad Refers to the participants in the Quadrilateral offer list of the concessions that could be granted if meetings--that is, Canada, the European Union, its request were met. Japan, and the United States. Restrictive business practice Measures by business Quantitative restriction or quota Measure restrict- enterprises to limit access to markets and restrain ing the quantity of a good imported (or exported). competition (such as the formation of a cartel). Quantitative restrictions include quotas, nonauto- Retaliation Imposition of a trade barrier in matic licensing, mixing regulations, voluntary response to another country's increasing its level of export restraints, and prohibitions or embargoes. trade restrictions. Global Quota Quota specifying the total volume, or Revealed comparative advantage (RCA) The ratio value, of the product to be imported (exported) of a country's exports of a good to the world's 601 D E V E L O P M E N T, T R A D E , A N D T H E W T O exports of that good divided by that country's share WTO rules or granting them preferential treatment of exports of manufactures in the world exports of in the application of WTO rules. manufactures. The index for country i good j is Special drawing right International payment facil- RCAij = 100(Xij /Xwj)/(Xit /Xwt) where Xab is exports ity administered by the IMF. Also used as an inter- by country a (w = world) of good b (t = total for all national unit of accounting, defined in terms of the goods). A value of the index above (below) one, is five most important national currencies in interna- interpreted as a revealed comparative advantage tional trade. (comparative disadvantage) for the good. Special safeguard In the WTO Agreement on Agri- Rollback The phasing out of measures inconsistent culture, a protectionist measure that can be trig- with the provisions of an agreement. gered automatically by a decline in prices or an Round In the WTO context, a multilateral trade increase in imports. negotiation. There have been eight rounds: Geneva Specific commitment Under the GATS, technical (1947), Annecy (1949), Torquay (1950­51), Geneva term describing the commitments made by WTO (1955­56), Dillon (1960­61), Kennedy (1963­67), members on national treatment and market access Tokyo (1973­79), and Uruguay (1986­94). A ninth for service sectors. multilateral negotiation was launched in Doha, Specific tariff A specific duty (tariff, import tax) Qatar, at the end of 2001. expressed in terms of a fixed amount per unit of the Rule of origin See Origin rule. dutiable item. For example, $1,000 on each import- Safeguard action Emergency protection to safe- ed vehicle or $50 on each ton of wheat. guard domestic producers of a specific good from Specificity A policy measure that applies to one or a an unforeseen surge in imports (GATT Art. XIX), to subset of enterprises or industries as opposed to all protect a country's external financial position and industries. balance ­of payments (GATT Art. XII, XVIII:B), or SPS See Sanitary and phytosanitary measure. to protect an infant industry in a developing coun- Standard Rule, regulation, or procedure specifying try (GATT Art. XVIII:A or C). See also Escape characteristics that must be met by a product (such clause. as dimensions, quality, performance, or safety). Sanitary and phytosanitary (SPS) measure A When these put foreign producers at a disadvan- technical requirement specifying criteria to ensure tage, they may constitute a nontariff barrier. See also food safety and animal and plant health. Many Technical barrier to trade. international SPS standards are set by the State trading Trade by a government agency or FAO/WHO. See also Codex Alimentarius Commis- enterprise or by an enterprise to which the govern- sion. ment has granted exclusive or special privileges in Second-best argument (for protection) Any argu- respect of international trade. State trading does not ment for protection that can be countered by point- necessarily involve a monopoly or quantitative ing to a less costly policy that would achieve the restriction of trade and does not require state own- same desired result. Also refers to rationales for pro- ership (GATT Art. XVII). tection to partially correct a distortion in the econo- Standstill A commitment not to take any new trade my when the first-best policy for that purpose is not restrictive or distorting measure. available. For example, if domestic production gen- Strategic trade policy The use of trade policies to erates a positive externality and a production sub- alter the outcome of international competition in a sidy to internalize it is not available, then a tariff country's favor, usually by allowing its firms to cap- may be second-best optimal. ture a larger share of industry profits. Selectivity Application of a rule, regulation, or Structural adjustment Process of reallocating trade action on a discriminatory basis to certain resources and changing the structure of production countries. and employment of a national economy to reflect Shallow integration Reduction or elimination of changing economic policies or trading conditions. border barriers to trade. Contrasts with deep inte- Subsidy Assistance granted by government to the gration. production, manufacture, or export of specific Special and differential treatment The principle goods, and taking the form either of direct pay- in WTO that developing countries be accorded spe- ments, such as grants or loans, or of measures hav- cial privileges, either exempting them from some ing equivalent effect, such as guarantees, 602 Glossary operational or support services or facilities, and fis- Terms of trade The price of a country's exports rel- cal incentives. ative to the price of its imports. Sunset clause Provision in a legal instrument limit- Total factor productivity (TFP) A measure of the ing the duration of validity of a particular measure output of an industry or economy relative to its or policy. inputs. The term and its abbreviation often refer to Tariff See Customs duty. the growth of this measure. Tariff binding In GATT context, commitment by Trade capacity The supply-side ability (capacity) of countries not to raise particular tariff items above a a country to benefit from the opportunities offered specific or bound level. Also referred to as ceiling by the world market and MFN or preferential access bindings. The so-called schedule of tariff conces- to markets. sions of each WTO member is annexed to its proto- Trade creation Occurs when liberalization results col of accession. See also Ceiling binding. in imports displacing less efficient local production Tariff equivalent Measure of the protective effect of and/or expanding consumption that was previously an NTB--the tariff that would have the exact same thwarted by artificially high prices due to protec- effect on imports as the NTB. tion. Tariff escalation Occurs if the tariff increases as a Trade diversion Occurs when a trade reform dis- good becomes more processed. Escalation discour- criminates between different trading partners and a ages imports of more processed varieties of the less efficient (higher cost) source displaces a more good (discouraging foreign processing activity) and efficient (lower cost) one. Can arise whenever some offers domestic processors positive levels of effective preferred suppliers are freed from barriers but oth- protection. For example, low duties on tomatoes, ers are not. higher duties on tomato paste, and yet higher duties Trade integration Process of reducing barriers to on tomato ketchup. trade and increasing participation in the interna- Tariff peaks Tariffs that are particularly high, often tional economy through trade. Also used to describe defined as rates that exceed 15 percent or the aver- efforts to integrate trade policy and strengthening age nominal tariff by a factor of more than three. of trade-related institutions into a country's overall Tariff rate quota (TRQ) Measure under which a development strategy. good is subject to a MFN tariff, but a certain quanti- Trade Policy Review Mechanism WTO mechanism ty (the "quota") is admitted at a lower, sometimes for periodic review of the trade policies and prac- zero, tariff. TRQs are mainly applied to agricultural tices of members. trade and can be seasonal. Trade promotion authority See Fast track. Tariffication Procedure of converting NTBs into Trade-related Investment Measure Policy used by their tariff equivalents. In the Uruguay Round, all governments to influence the operations of foreign industrial countries' agricultural NTBs were tariffed investors by establishing specific performance stan- and bound. dards relating to trade. Examples are export per- Technical barrier to trade Trade-restrictive effect formance requirements and local content rules arising from the application of technical regulations (mandating that investors use a certain proportion or standards such as testing requirements, labeling of domestic inputs in their production). requirements, packaging requirements, marketing Trade-related Technical Assistance Services standards, certification requirements, origin-mark- financed and/or provided by donors and develop- ing requirements, health and safety regulations, and ment agencies to strengthen trade-related institu- sanitary and phytosanitary regulations. tions and build trade capacity in developing Technical regulation A mandatory requirement or countries. See also Integrated Framework. standard specifying the characteristics that an import- Trademark Distinctive mark or name to identify a ed product must meet.Usually aimed to protect public product, service, or company. health or safety. See Technical barrier to trade. Transaction value Used for customs valuation pur- Temporary admission Customs regime under poses--the price of a good actually paid or payable. which firms may import intermediates duty free if Transparency Clarity, openness, predictability, and used in export production, and are required to doc- comprehensibility (used in regard to individual ument ex post that imports have been used for this trade-related regulations and operation of institu- purpose. See also Duty drawback. tions). 603 D E V E L O P M E N T, T R A D E , A N D T H E W T O Trigger price See Maximum (minimum) price sys- former agrees to limit exports of a specified good to tem. avoid dislocation of the industry in, and possible TRIMs See Trade-related Investment Measure. imposition of mandatory restrictions by, the TRIPs Trade-related intellectual property rights. In importing country. The restraint agreement may be WTO, used as an acronym for the Agreement on concluded at either industry or government level. In Trade-Related Aspects of Intellectual Property the latter case, it is sometimes referred to as an Rights. "orderly marketing arrangement." TRTA See Trade-related Technical Assistance. Voluntary restraint agreement See Voluntary UPOV International Union for the Protection of New export restraint. Varieties of Plants. Seeks to ensure that signatories Waiver Authorized deviation from a previously acknowledge the achievements of breeders of new undertaken and legally binding obligation. In WTO, plant varieties by making available to them an exclu- a waiver can be sought through invocation of Art. sive property right, on the basis of a set of uniform IX WTO. Conditions under which waivers are and clearly defined principles. Based in Geneva. granted are generally negotiated and limited in Value added The value of output minus the value time. of all inputs used in production. Equals, by defini- Welfare Welfare is the "enjoyment" that consumers tion, the contribution of, and payments to, primary are inferred to gain from their consumption. While factors of production (labor, capital, and land). welfare cannot be measured directly, economists Variable levy Import charge that brings the import often use a measure of real income or purchasing price of a good into line with or above a decreed power as a way of measuring welfare in money internal price to protect domestic production from terms. import competition. Given the decreed internal WITS World Integrated Trade Solution--database price, the levy will vary inversely with the world and software package developed by UNCTAD and price. the World Bank to allow analysis of market access Voluntary export restraint Informal agreement conditions and the impact of own and partner between an exporter and an importer whereby the country liberalization. 604 Bibliography Abbott, F. M. 1998a. "The Enduring Enigma of TRIPS: A Challenge Aitken, Brian, and Ann E. Harrison. 1999. "Do Domestic Firms for the World Economic System." Journal of International Eco- Benefit from Direct Foreign Investment?" American Economic nomic Law 1 (4): 497­521. Review 89 (3): 605­18. ------. 1998b. "First Report (Final) to the Committee on Interna- Aitken, Brian, Ann E. Harrison, and Robert E. Lipsey. 1996. "Wages tional Trade Law of the International Law Association on the and Foreign Ownership: A Comparative Study of Mexico, Subject of Parallel Importation." Journal of International Eco- Venezuela, and the United States." Journal of International Eco- nomic Law 1: 607­36. nomics 40: 345­71. ------. 2000. "The TRIPS-Legality of Measures Taken to Address Akerlof, G. A. 1970. "The Market for Lemons: Qualitative Uncer- Public Health Crises: Responding to USTR-Industry Positions tainty and the Market Mechanism." Quarterly Journal of Eco- That Undermine the WTO." Florida State University, Tallahas- nomics 84: 488­500. see. Processed. Alexander, Ian, and Antonio Estache. 1999. "The Role of Regulatory Abrenica, J. 2000. "The Philippine Automotive Industry." Prepared Reform and Growth: Lessons from Latin America." Prepared for for the PECC-ASEAN Auto Project, School of Advanced Man- TIPS Annual Forum. Johannesburg, September 19­22. Processed. agement Technology, Manila. Ali, M. Shaukat. 2000. "Trade and Industrial Policy in Pakistan: Acemoglu, Daron, Simon Johnson, and James A. Robinson. 2000. Post­Uruguay Round Challenges." Prepared for World Bank "The Colonial Origins of Comparative Development: An Empir- WTO-2000 project. World Bank, Washington, D.C. Processed. ical Investigation." NBER Working Paper 7771. National Anderson, K. 1997. "Environmental Standards and International Bureau of Economic Research, Cambridge, Mass. Trade." In Michael Bruno and Boris Pleskovic, eds., Annual Adams, Charles. 1993. For Good and Evil: The Impact of Taxes on World Bank Conference on Development Economics 1996, the Course of Civilization. Lanham, Md.: Madison Books. 317­38. Washington, D.C.: World Bank. Adams, Philip D. 1998. "Long-Run Effects of APEC Trade Liberali- Anderson, Kym, and Richard Blackhurst, eds. 1992. The Greening sation: An Applied General Equilibrium Analysis." World Econo- of World Trade Issues. London: Harvester Wheatsheaf. my 21: 931­52. Anderson, Kym, Bernard Hoekman, and Anna Strutt. 2001. "Agri- ADB (Asian Development Bank). 1999. Asian Development Outlook culture and the WTO: Next Steps." Review of International Eco- 1999. Manila. nomics 9 (2, May): 192­214. Ades, Alberto, and Rafael Di Tella. 1997. "National Champions Anderson, Robert, and Nancy T. Gallini. 1998. Competition Policy and Corruption: Some Unpleasant Interventionist Arithmetic." and Intellectual Property Rights in the Knowledge-Based Economy. Economic Journal 107 (July): 1023­42. Calgary, Alberta: University of Calgary Press. ------.1999. "Rents, Competition and Corruption." American Eco- Anjaria, S. J. 1987. "Balance of Payments and Related Issues in the nomic Review 89 (4, September): 982­93. Uruguay Round of Trade Negotiations." World Bank Economic Adlung, Rudolf, A. Carzeniga, and A. Mattoo. 1999. "The Pattern Review 1 (4): 669­88. of Restrictions in GATS Schedules." World Trade Organization, Arora, Ashish. 1996. "Contracting for Tacit Knowledge: The Provi- Geneva. Processed. sion of Technical Services in Technology Licensing Contracts." AfDB (African Development Bank). 2000. African Development Journal of Development Economics 50: 233­56. Report 2000. New York: Oxford University Press. Arrow, K. J. 1962. "Economic Welfare and the Allocation of Aggarwal, Mita. 1995. "International Trade, Labor Standards, and Resources for Invention." In R. R. Nelson, ed., The Rate and Labor Market Conditions: An Evaluation of the Linkages." U.S. Direction of Inventive Activity, 609­25. Princeton, N.J.: Prince- International Trade Commission Office of Economics Working ton University Press. Paper 95-060C. Washington, D.C. Askari, H., and J. T. Cummings. 1977. "Estimating Agricultural Agosin, M. 2001. "Export Performance in Chile: Lessons for Supply Response with the Nerlove Model: A Survey." Interna- Africa." In Gerald K. Helleiner, ed., Non-Traditional Exports and tional Economic Review 18 (2): 257­92. Development in Sub-Saharan Africa: Experience and Issues. Aswichahyono, A., T. Anas, and J. Rizal. 2000. "The Indonesian WIDER Studies in Development Economics and Policy. Automotive Industry." Prepared for the PECC-ASEAN Auto Pro- Houndsmill, Basingstoke, Hampshire, U.K.: Palgrave. ject, School of Advanced Management Technology, Manila. 605 D E V E L O P M E N T, T R A D E , A N D T H E W T O Atinc, Tamar Manuelyan. 1997. "Sharing Rising Incomes: Dispari- Bardhan, Pranab. 1997. "Corruption and Development: A Review ties in China." World Bank, Washington, D.C. of the Issues." Journal of Economic Literature 35: 1320­46. Australia, Department of Foreign Affairs and Trade. 1999. Global Bardsley, P., and P. Cashin. 1990. "Underwriting Assistance to the Trade Reform: Maintaining Momentum. Canberra. Australian Wheat Industry: An Application of Option Pricing Australia, Industries Assistance Commission. 1981. Passenger Theory." Australian Journal of Agricultural Economics 34 (3): Motor Vehicles and Components: Post-1984 Assistance Arrange- 212­22. ments. Canberra. Barfield, Claude E. 1996. "Regionalism and US Trade Policy." In ------.1984. Local Content Schemes: A Technical Analysis. Working Jagdish Bhagwati and Arvind Panagariya, The Economics of Pref- Paper. July. Canberra. erential Trade Agreements. Washington, D.C.: AEI Press. Australia, Industry Commission. 1997. The Automotive Industry. Barro, Robert J., and Xavier Sala-i-Martin. 1995. Economic Growth. Vols. 1 and 2. Report 58. May 26. Canberra. New York, McGraw-Hill. Azam, Jean-Paul, and Shantayanan Devarajan. 1997. "The CFA Barth, James R., Gerard Caprio, Jr., and Ross Levine. 2001a. "Bank Franc Zone in Africa: A Symposium." Journal of African Regulation and Supervision: What Works Best?" Presented at Economies 6 (1, March): 1­2. the 13th Annual World Bank Conference on Development Eco- Baeumer, Ludwig. 1989. "Considerations concerning a Definition nomics, Washington, D.C., May 1­2. of Geographical Indications." Presented to the WIPO Sympo- Barth, James R., Gerard Caprio, Jr., and Ross Levine. 2001b. "The sium on the International Protection of Geographical Indica- Regulation and Supervision of Banks around the World: A New tions, Santenay, France, September 9­10. Database." Policy Research Working Paper Series 2588. World ------. 1997. "Protection of Geographical Indications under Bank, Development Research Group, Washington, D.C. WIPO Treaties and Questions concerning the Relationship Belderbos, R., and P. Holmes. 1995. "The Economics of Matsushita between Those Treaties and the TRIPS Agreement." Presented Revisited." Anti-Trust Bulletin 40 (winter): 825­57. to the WIPO Symposium on the International Protection of Ben-David, Dan. 1993. "Equalizing Exchange: Trade Liberalization Geographical Indications in the Worldwide Context, Eger, and Income Convergence." Quarterly Journal of Economics 108 Hungary, October 24­25. (3): 653­79. Bagwell, Kyle, and Robert W. Staiger. 1999. "An Economic Theory Benjamin, Nancy, and Xinshen Diao. 2000. "Liberalizing Services of GATT." American Economic Review 48 (1): 216­48. Trade in APEC: A General Equilibrium Analysis with Imperfect Bajaj, Harbans L. 1995. "Managing Contracts for Power Projects." Competition." Pacific Economic Review 5 (1). Presented at the International Conference on Contract Man- Bergman L., C. Doyle, J. Gual, L. Hultkrantz, D. Neven, L. Roller, agement in the Construction Industry, New Delhi. and L. Waverman. 1998. Europe's Network Industries: Conflicting Balassa, Bela. 1989. "Tariff Policy and Taxation in Developing Priorities--Telecommunications. London: Centre for Economic Countries." Policy Research Working Paper Series 281. World Policy Research. Bank, Washington, D.C. Besen, Stanley M., and Leo J. Raskind. 1991. "An Introduction to Balasubramanyam, V. N. 1991. "Putting TRIMs to Good Use." the Law and Economics of Intellectual Property." Journal of Eco- World Development 19: 1215­24. nomic Perspectives 5 (1): 3­27. Balasubramanyam, V. N., M. Salisu, and D. Sapsford. 1996. "For- Bhagwati, Jagdish N. 1968. The Theory and Practice of Commercial eign Direct Investment and Growth in EP and IS Countries." Policy. International Finance Section, Department of Econom- Economic Journal 106: 92­105. ics. Princeton, N.J.: Princeton University. Baldwin, Richard E. 1994. Towards an Integrated Europe. London: ------. 1971. "The Generalized Theory of Distortions and Wel- Centre for Economic Policy Research. fare." In Jagdish N. Bhagwati, Ronald W. Jones, Robert A. ------. 2001. "Regulatory Protectionism, Developing Nations, and Mundell, and Jaroslav Vanek, eds., Trade, Balance of Payments a Two-Tier World Trade System." In Keith E. Maskus and John S. and Growth: Papers in International Economics in Honor of Wilson, eds., Quantifying the Trade Effect of Technical Barriers: Charles P. Kindleberger. Amsterdam: North-Holland. Can It Be Done? Ann Arbor: University of Michigan Press. ------. 1987. "Directly Unproductive Profit-Seeking (DUP) Activi- Baldwin, Richard E., and Anthony Venables. 1997. "International ties." In John Eatwell, Murray Milgate, and Peter Newman, Economic Integration." In Gene M. Grossman and Kenneth eds., The New Palgrave: A Dictionary of Economics, 845­47. Rogoff, eds., Handbook of International Economics, vol. 3. Ams- New York: Stockton Press. terdam: North-Holland. ------. 1988. Protectionism. Cambridge, Mass.: MIT Press. Baldwin, Robert. E. 1969. "The Case against Infant Industry Pro- ------. 1998. "Powerful Reasons for the MAI to Be Dropped Even tection." Journal of Political Economy 77: 295­305. from the WTO Agenda." Financial Times (October 22). ------. 1992. "Assessing the Fair Trade and Safeguards Laws in Bhagwati, Jagdish N., and Robert E. Hudec, eds. 1996. Fair Trade Terms of Modern Trade and Political Economy Analysis." World and Harmonization: Prerequisites for Free Trade? 2 vols. Cam- Economy 15 (2, March): 185­202. bridge, Mass.: MIT Press. Baldwin, Robert, and J. David Richardson. 1972. "Government Bhagwati, Jagdish, and Arvind Panagariya. 1986. The Economics of Purchasing Policies, Other NTBs, and the International Mone- Preferential Trade Agreements. Washington, D.C.: AEI Press. tary Crisis." In H. English and K. Hay, eds., Obstacles to Trade in Bhagwati, J., and T. N. Srinivasan. 1969. "Optimal Intervention to the Pacific Area. Ottawa: Carleton School of International Achieve Non-Economic Objectives.", Review of Economic Stud- Affairs. ies 36: 27­38. Barajas, Adolfo, Roberto Steiner, and Natalia Salazar. 1999. "For- ------. 1999. "Outward-Orientation and Development: Are Revi- eign Investment in Colombia's Financial Sector." IMF Working sionists Right?" Yale University Economic Growth Center Dis- Paper WP/99/150. Washington, D.C. cussion Paper 806. New Haven, Conn. 606 Bibliography Bienaymé, Marie-Hèléne. 1989. "The Possible Content of a New Borensztein, E., J. de Gregorio, and J. W. Lee. 1998. "How Does Treaty on the Protection of Geographical Indications at the Foreign Direct Investment Affect Economic Growth?" Journal of Multilateral Level (Part I)." Presented to the WIPO Symposium International Economics 45: 115­35. on the International Protection of Geographical Indications, Borrell, Brent. 1997. "Policy-Making in the EU: The Bananarama Santenay, France, September 9­10. Story, the WTO, and Policy Transparency." Australian Journal of Biggs, Tyler. 1999. "A Micro-Econometric Evaluation of the Mauri- Agricultural and Resource Economics 41 (June): 263­76. tius Technology Diffusion Scheme." RPED Discussion Paper. Bosworth, Malcolm. 2000. "Comment." In Pierre Sauvé and World Bank, Regional Program on Enterprise Development, Robert M. Stern, eds., GATS 2000: New Directions in Services Washington, D.C. Trade Liberalization, 211­17. Washington, D.C.: Brookings Blackhurst, Richard. 1991. "Trade Policy Is Competition Policy." Institution Press and Harvard University. Paper prepared for the OECD, Paris. Botsch, Andreas. 1995. "Labor Standards and the Multilateral Trade ------. 1998a. "Capacity Building in Africa for Enhanced Partici- and Investment System." In Rajah Rasiah and Norbert von Hoff- pation in the WTO." Prepared for the dissemination confer- man, Social and Environmental Clauses and Free Trade. Bonn, Ger- ence of the AERC collaborative research project on Africa and many: Friedrich Ebert Stiftung.Bovard, James. 1991. The Fair the world trading system. African Economic Research Consor- Trade Fraud: How Congress Pillages the Consumer and Decimates tium, Nairobi. Processed. American Competitiveness. New York: St. Martin's Press. ------. 1998b. "The Capacity of the WTO to Fulfill Its Mandate." Braga, Carlos Prima. 1996. "Trade-Related Intellectual Property In Anne O. Krueger, ed., The WTO as an International Organi- Issues: The Uruguay Round Agreement and Its Economic Impli- zation. Chicago: University of Chicago Press. cations." In Will Martin and L. Alan Winters, eds., The Uruguay ------. 2001. "Reforming WTO Decision-Making: Lessons from Round and the Developing Economies. Cambridge, U.K.: Cam- Singapore and Seattle." In Klaus Deutsch and Bernhard Speyer, bridge University Press. eds., World Trade Organization Millennium Round: Freer Trade in Braga, Carlos A. Primo, Carsten Fink, and Claudia Paz Sepulveda. the Twenty-first Century. London: Routledge. 2000. Intellectual Property Rights and Economic Development. Blackhurst, R., J. Francois, and E. Enders. 1996. "Market Access World Bank Discussion Paper 412. Washington, D.C. and Developing Countries." In Will Martin and L. Alan Winters, Brander, J. A. 1995. "Strategic Trade Policy." In Gene M. Gross- eds., The Uruguay Round and the Developing Economies. Cam- man and Kenneth Rogoff, eds., Handbook of International Eco- bridge, U.K.: Cambridge University Press. nomics, vol. 3. Amsterdam: North-Holland. Blackhurst, Richard, William Lyakurwa, and Ademola Oyejide. Brander, J., and B. Spencer. 1985. "Export Subsidies and Internation- 2000. "Options for Improving Africa's Participation in the al Market Rivalry." Journal of International Economics 18: 83­100. WTO." World Economy 23 (4): 491­510. Bronckers, Marco C. E. J., and Pierre Larouche. 1997. "Telecom- Blakeney, Michael. 1998. "Communal Intellectual Property Rights munications Services and the World Trade Organisation." Jour- of Indigenous People in Cultural Expressions." Journal of World nal of World Trade 31 (3): 5­48. Intellectual Property 1 (6): 985­1002. Brookes, Martin, and Saki Wahhaj. 2000. "The Shocking Effect of ------. 2000. "The Protection of Traditional Knowledge under B2B." Goldman Sachs Global Economics Paper 37. Global Eco- Intellectual Property Law." European Intellectual Property Review nomics Week (February 3). 22 (6, June): 251­61. Brown, Drusilla K., and Robert M. Stern. 2001. "Measurement and Blejer, Mario, and Adriene Cheasty. 1990. "Fiscal Implications of Modeling of the Economic Effects of Trade and Investment Trade Liberalization." In Vito Tanzi, ed., Fiscal Policy in Open Barriers in Services." Review of International Economics 9 (2, Developing Economies, 66­81. Washington, D.C.: IMF. May): 282­86. Blomström, Magnus. 1986. "Foreign Investment and Productive Brown, Drusilla K., Alan V. Deardorff, and Robert M. Stern. 1996. Efficiency: The Case of Mexico." Journal of International Eco- "Modeling Multilateral Liberalization in Services." Asia-Pacific nomics 15: 97­110. Economic Review 2: 21­34. ------. 1989. Foreign Investment and Spillovers. London: Rout- Brown, Drusilla K., Alan V. Deardorff, Alan K. Fox, and Robert M. ledge. Stern. 1996. "The Liberalization of Services Trade: Potential Blomström, Magnus, and Hakan Persson. 1983. "Foreign Invest- Impacts in the Aftermath of the Uruguay Round." In Will Mar- ment and Spillover Efficiency in an Underdeveloped Economy: tin and L. Alan Winters, eds., The Uruguay Round and the Devel- Evidence from the Mexican Manufacturing Industry." World oping Economies. Cambridge, U.K.: Cambridge University Press. Development 11 (6): 493­501. Bruton, H. J. 1998. "A Reconsideration of Import Substitution." Blomström, Magnus, and Fredrik Sjöholm. 1999. "Technology Journal of Economic Literature 36: 903­36. Transfer and Spillovers: Does Local Participation with Multina- Buffie, Edward. 2001. Trade Policy in Developing Countries. Cam- tionals Matter?" European Economic Review 43 (4­6): 915­23. bridge, U.K.: Cambridge University Press. Bora, B., and S. Guisinger. 1997. "Impact of Investment Liberaliza- Buigues, P., A. Jacquemin, and A. Sapir, eds. 1995. European Poli- tion in APEC." Flinders University, Adelaide, Australia. Processed. cies on Competition, Trade and Industry. Cheltenham, U.K.: Bora, B., and I. N. Neufeld. 2000. "Tariffs and the East Asian Edward Elgar. Crises." UNCTAD, Trade Analysis Branch, Geneva. Processed. Burtless, Gary, Robert Z. Lawrence, Robert E. Litan, and Robert J. Bora, B., and R. Pomfret. 1995. "Manufacturing Policies." In R. Shapiro. 1998. Globaphobia: Confronting Fears about Open Pomfret, ed., Australia's Trade Policies. Melbourne: Oxford Uni- Trade. Washington, D.C.: Brookings Institution Press. versity Press. Cadot, Olivier, Jaime de Melo, and Marcelo Olarreaga. 2000. "Can Bora, Bijit, Peter J. Lloyd, and Mari Pangestu. 2000. "Industrial Pol- Duty Drawbacks Have a Protectionist Bias?" World Bank, Wash- icy and the WTO." World Economy 23 (4, April): 543­59. ington, D.C. Processed. 607 D E V E L O P M E N T, T R A D E , A N D T H E W T O Canadian Chamber of Commerce. 1995. "North American Cus- Clarke, George, Robert Cull, Laura D'Amato, and Andrea Molinari. toms Survey." August. Ottawa. 1999. "The Effect of Foreign Entry on Argentina's Domestic Cardona, T. D., S. G. Ilangantileke, and A. Noomhorm. 2000. "Afla- Banking Sector." Policy Research Working Paper Series 2158. toxin Research on Grain in Asia: Its Problems and Possible Solu- World Bank, Washington, D.C. tions." in Mycotoxin Prevention and Control in Foodgrains, edited Clément, Jean A. P. 1994. "Striving for Stability: CFA Franc by R. L. Semple, A. S. Frio, P. A. Hicks and J. V. Lozare. Bankok, Realignment." Finance and Development 31 (2, June): 10­13. Thailand: UNDP/FAO Regional Network Inter-Country Coopera- Clément, Jean A. P., with Johannes Mueller, Stéphane Cossé, and tion on Preharvest Technology and Quality Control of Food- Jean Le Dem. 1996. Aftermath of the CFA France Devaluation. grains (REGNET) and the ASEAN Grain Postharvest Programme. IMF Occasional Paper 138. Washington, D.C. Casella, A. 1996. "Free Trade and Evolving Standards." In Jagdish N. Clements, K. W., and L. A. Sjaastad. 1984. How Protection Taxes Bhagwati and Robert E. Hudec, eds., Fair Trade and Harmoniza- Exports. Thames Essay 39. London: Trade Policy Research Cen- tion: Prerequisites for Free Trade? Cambridge, Mass.: MIT Press. tre. Cassel, D. 1996. "Corporate Initiatives: A Second Human Rights Coe, David T., Elhanan Helpman, and Alexander W. Hoffmaister. Revolution?" Fordham International Law Journal 19: 1963­84. 1997. "North-South R&D Spillovers." Economic Journal 107: Caves, Richard E. 1974. "Multinational Firms, Competition, and 134­49. Productivity in Host-Country Industries." Economica 41: Collins, John. 2000. "World Music World-Wide Sale." Presented to 176­93. the workshop "Developing the Music Industry in Africa," ------. 1996. Multinational Enterprise and Economic Analysis. Cam- World Bank, Washington, D.C., June 20. bridge, U.K.: Cambridge University Press. Corbo, Vittorio, and Jaime de Melo. 1987. "Lessons from the Chadha, Rajesh, Bernard Hoekman, Will Martin, Ademola Oyejide, Southern Cone Policy Reforms." World Bank Research Observer Mari Pangestu, Diana Tussie, and Jamel Zarrouk. 2000. "Devel- 2 (2, July): 111­42. oping Countries and the Next Round of WTO Negotiations." Corbo, Vittorio, and Stanley Fischer. 1994. "Lessons from the World Economy 23 (4, April): 431­36. Chilean Stabilization and Recovery." In Barry P. Bosworth, Chang, Philip, Guy Karsenty, Aaditya Mattoo, and Jurgen Richter- Rudiger Dornbusch, and Raúl Labán, eds., The Chilean Econo- ing. 1999. "GATS, the Modes of Supply and Statistics on Trade my: Policy Lessons and Challenges, ch. 1. Washington, D.C.: in Services." Journal of World Trade 33 (3, June): 93­115. Brookings Institution Press. Chang, Won, and L. Alan Winters. 2001. "Preferential Trading Corbo, Vittorio, Jaime de Melo, and James Tybout. 1986. "What Arrangements and Excluded Countries: Ex-post Estimates of Went Wrong with the Recent Reforms in the Southern Cone?" the Effects on Prices." World Economy 24 (6,June): 797­807. Economic Development and Cultural Change 34 (3, April): Chao, C. C., and E. S. H. Yu. 1993. "Content Protection, Urban 607­40. Unemployment and Welfare." Canadian Journal of Economics Corden, W. M. 1971. The Theory of Protection. Oxford, U.K.: 26: 481­92. Clarendon Press. Charnovitz, Steve. 1996. "Participation of Nongovernmental Orga- ------. 1974. Trade Policy and Economic Welfare. Oxford, U.K.: nizations in the World Trade Organization." University of Pennsyl- Clarendon Press. (2d ed., 1997.) vania Journal of International Economic Law 17 (Spring): 331­57. ------. 1984. "The Normative Theory of International Trade." In ------. 1998. "The Moral Exception in Trade Policy." Virginia Jour- P. B. Kenen and R. W. Jones, eds., Handbook of International nal of International Law 38 (4, Summer): 689­745. Economics, vol. 1, ch. 2. Amsterdam: North-Holland. Chartrand, H. 1996. "Intellectual Property Rights in the Post-Mod- ------. 1987. Protection and Liberalization: A Review of Analytical ern World." Journal of Arts Management, Law and Society 25: Issues. IMF Occasional Paper 54. Washington, D.C. 306­19. ------. 1997. Trade Policy and Economic Welfare. Oxford, U.K.: Chatterjee, Pratap, and Matthias Finger. 1994. The Earth Brokers: Clarendon Press. Power, Politics, and World Development. London and New York: Cornish, W. R. 1999. Intellectual Property: Patents, Copyright, Trade Routledge. Marks and Allied Rights. 4th ed. London: Sweet & Maxwell. Choi, Inbom. 2000. "Long and Winding Road to the Government Correa, Carlos. 2001. "Implementing the TRIPS Agreement: Procurement Agreement: Korea's Accession Experience." Paper Options for Latin American Countries." LATN Working Papers. presented at the World Bank-PECC Trade Policy Forum Semi- Latin American Trade Network, Buenos Aires. nar on "East Asia and Options for WTO 2000 Negotiations," Correa, Carlos M., and Abdulqawi A. Yusuf, eds. 1998. Intellectual held in Manila July 19­20.Chudnovsky, D., and A. López. Property and International Trade: The TRIPS Agreement. London: 2000. "Policy Competition for Foreign Direct Investment: The Kluwer Law International. Global and Regional Dimensions." LATN Working Paper 5. Cottani, Joaquin A., Domingo F. Cavallo, and M. Shahbaz Khan. February. Latin American Trade Network, Buenos Aires. 1990. "Real Exchange Rate Behavior and Economic Perfor- Cicerone, Michael J. 2001. "Overview: International Merger Con- mance in LDCs." Economic Development and Cultural Change trol." Antitrust 15 (2): 15­23. 39 (1, October): 61­76. Claessens, Stijn, and Thomas Glaessner. 1999. "Internationaliza- Cottier, Thomas. 1998. "The Protection of Genetic Resources and tion of Financial Services in Asia." Paper presented at the con- Traditional Knowledge: Towards More Specific Rights and ference "Investment Liberalization and Financial Reform in the Obligations in World Trade Law." Journal of International Eco- Asia-Pacific Region," Sydney, Australia, August. nomic Law 1 (4, December): 555­84. Clarette, Ramon. 1989. "The Economic Effects of Trade Liberaliza- Cowhey, Peter, and Mikhail M. Klimenko. 2001. "The WTO Agree- tion on Philippine Agriculture." Philippine Review of Economics ment and Telecommunications Policy Reform." Policy Research and Business 26 (2): 208­35. Working Paper Series 2601. World Bank, Washington, D.C. 608 Bibliography Crawford, Jo-Ann, and Sam Laird. 2000. "Regional Trade Agree- Dee, P., A. Hardin, and M. Schuele. 1998. "Early Voluntary Sec- ments and the WTO." CREDIT Research Paper 00/03. Centre toral Liberalisation." Productivity Commission Staff Research for Research in Economic Development and International Paper. Ausinfo, Canberra. Trade, University of Nottingham, U.K. DeFreitas, D., C. Kenny, and R. Schware. 2001. "Caribbean Coop- Croft, Adrian. 1997. "EU Body Seeks New Russian, China Dump- eration: Rise of the Regional Regulator." Journal of Policy, Regu- ing Policy." Reuters (December 16). lation and Strategy for Telecommunications Information and Croome, John. 1999. Reshaping the Trading System. Deventer: Media 3: 195­201. Kluwer. Delich, L. Valentina. 2000. "Latinoámerica y el sistema de solución Cunningham, B. 1996. "Tanzania: Strategy and Action Plan to de diferencias de la Organización Mundial del Comercio." In Reform Customs Administration." June. Tanzania Revenue Marcelo Olarreaga and Ricardo Rocha, eds., La nueva agenda Authority, Dar es Salaam. del comercio en la OMC. Bogotá: Centro Editorial Universidad Curzon, Gerard. 1965. Multilateral Trade Diplomacy. London: del Rosario. Michael Joseph. Deltas, George, and Simon J. Evenett. 1997. "Quantitative Esti- CUTS (Consumer Unity & Trust Society). 1999. "Conditions Nec- mates of the Effects of Preference Policies." In Bernard M. essary for the Liberalisation of Trade and Investment to Reduce Hoekman and Petros C. Mavroidis, eds., Law and Policy in Pub- Poverty." Final report to the Department for International lic Purchasing: The WTO Agreement on Government Procurement. Development (DFID). August. London. Ann Arbor: University of Michigan Press. ------. 2001. "Contours of a National Competition Policy: A de Melo, Jaime, Marcelo Olarreaga, and Wendy Takacs. 2000. Development Perspective." Available at "Pricing Policy under Double Market Power: Madagascar and . the International Vanilla Market." Review of Development Eco- Dages, B. Gerard, Linda Goldberg, and Daniel Kinney. 2000. nomics 4 (1, February): 1­20. "Foreign and Domestic Bank Participation in Emerging Mar- de Melo, Jaime, Arvind Panagariya, and Dani Rodrik. 1992. kets: Lessons from Mexico and Argentina." Economic Policy "Regional Integration: An Analytical and Empirical Overview." Review 6 (3, September): 17­36. Federal Reserve Bank of New In Jaime de Melo and Arvind Panagariya, eds., New Dimensions York. in Regional Integration. Cambridge, U.K.: Cambridge University Dahl, H., S. Devarajan, and S. van Wijnbergen. 1989. "Revenue- Press. Neutral Tariff Reform: Theory and an Application to Denizer, Cevdet. 1999. "Foreign Entry in Turkey's Banking Sector, Cameroon." World Bank, Washington, D.C. Processed. 1980­1997." Prepared for World Trade Organization­World Dam, Kenneth. 1970. The GATT: Law and the International Econom- Bank Conference on Liberalization and Internationalization of ic Organization. Chicago: University of Chicago Press. Financial Services, Geneva, May. Das, Nibedita. 2000. "Technology, Efficiency and Sustainability of Devarajan, Shantayanan. 1997. "Real Exchange Rate Misalign- Competition in the Indian Telecommunications Sector." Journal ment in the CFA Zone." Journal of African Economies 6 (1, of Information Economics and Policy 12: 133­54. March): 35­53. Dean, Judith M., Seema Desai, and James Riedel. 1994. Trade Poli- Devarajan, Shantayanan, and Lawrence E. Hinkle. 1994. "The CFA cy Reform in Developing Countries since 1985: A Review of the Franc Parity Change: An Opportunity to Restore Growth and Evidence. World Bank Discussion Paper 267. Washington, D.C. Reduce Poverty." Africa Spectrum 29 (2): 131­51. Deardorff, Alan. 1994. "Third-Country Effects of a Discriminatory Devereux, Michael, and Rachel Griffiths. 1998. "Taxes and the Tariff." World Economy 17 (January): 75­86. Location of Production: Evidence from a Panel of US States." Deardorff, Alan V., and Robert M. Stern. 1985. "The Structure of Journal of Public Economics 68: 335­67. Tariff Protection: Effects of Foreign Tariffs and Existing NTBs." Devlin, Robert, Antoni Estevadeordal, and Luis Jorge Garay. 1997. Review of Economics and Statistics 62 (4, November): 539­48. "Normas en acuerdos preferenciales de comercio en las Améri- ------. 1986. The Michigan Model of World Production and Trade. cas." Inter-American Development Bank, Division of Integra- Cambridge, Mass.: MIT Press. tion, Trade, and Hemispheric Issues, Washington, D.C. ------. 1998. Measurement of Nontariff Barriers. Ann Arbor: Uni- Processed. versity of Michigan Press. Dietz, Adolf. 2000. "Term of Protection in Copyright Law and Pay- Debroy, Bibek, and Garry Pursell. 1997. "Government Procure- ing Public Domain: A New German Initiative." European Intel- ment Policies in India." In Bernard M. Hoekman and Petros C. lectual Property Review 22 (11): 506­11. Mavroidis, eds., Law and Policy in Public Purchasing: The WTO Diller, Janelle M., and David A. Levy. 1997. "Child Labor, Trade, Agreement on Government Procurement. Ann Arbor: University and Investment: Toward the Harmonization of International of Michigan Press. Law." American Journal of International Law 91 (4, October): Decker, P. T., and W. Corson. 1995. "International Trade and 663­66. Worker Displacement: Evaluation of the Trade Adjustment Dinwoodie, Graeme B. 1997. "Reconceptualizing the Inherent Assistance Program." Industrial and Labor Relations Review 48 Distinctiveness of Product Design Trade Dress." North Carolina (4): 758­74. Law Review 75: 471­606. Dee, Philippa, and Kevin Hanslow. 2001. "Multilateral Liberaliza- Dixit, A. 1984. "International Trade under Oligopolistic Competi- tion of Services Trade." In Robert M. Stern, ed., Services in the tion." Economic Journal.94, Supplement, Conference Papers: International Economy. Ann Arbor: University of Michigan Press. 1­16. Dee, Philippa, Chris Geisler, and Greg Watts. 1996. "The Impact Dixit, A. K., and G. M. Grossman. 1982. "Trade and Protection of APEC's Free Trade Commitment." Staff Information Paper. with Multistage Production." Review of Economic Studies 43: Industry Commission, Canberra. 583­94. 609 D E V E L O P M E N T, T R A D E , A N D T H E W T O Djankov, Simeon, and Bernard Hoekman. 2000. "Foreign Invest- Ehrenberg, D. 1995. "The Labor Link: Applying the International ment and Productivity Growth in Czech Enterprises." World Trading System to Enforce Violations of Forced and Child Bank Economic Review 14 (1): 49­64. Labor." Yale Journal of International Law 20: 361­417. Dollar, David. 1992. "Outward-Oriented Developing Economies Ehrlich, Paul R., and Edward O. Wilson. 1991. "Biodiversity Stud- Really Do Grow More Rapidly: Evidence from 95 LDCs, ies: Science and Policy." Science 253 (August 16): 758­62. 1976­1985." Economic Development and Cultural Change 40: Eichengreen, Barry. 2001. "Capital Account Liberalization: What 523­44. Do the Cross-Country Studies Tell Us?" World Bank Economic Dollar David, and Aart Kraay. 2001. "Growth Is Good for the Poor." Review 15 (3): 341­65. Policy Research Working Paper Series 2587. World Bank, Devel- Elbadawi, Ibrahim, and Nader Majd. 1996. "Adjustment and Eco- opment Research Department, Washington, D.C. Processed. nomic Performance under a Fixed Exchange Rate: A Compara- Domberger, Simon, Christine Hall, and Eric Ah Lik Lee. 1995. "The tive Analysis of the CFA Zone." World Development 24: 939­51. Determinants of Price and Quality in Competitively Tendered Elbehri, A., M. Ingco, T. Hertel, and K. Pearson. 1999. "Agriculture Contracts." Economic Journal 105: 1454­70. and WTO 2000: Quantitative Assessment of Multilateral Liber- Donovan, D., and Y. H. Mai. 1996. "APEC Trade Liberalisation: The alization of Agricultural Policies." World Bank, Washington, Impact of Increased Capital Mobility." Australian Commodities D.C. Available at . Dornbusch, Rudiger, and Sebastian Edwards. 1994. "Exchange Esty, Daniel C. 1996. "Environmental Regulation and Competitive- Rate Policy and Trade Strategy." In Barry P. Bosworth, Rudiger ness: Theory and Practice." In Simon S. C. Tay and Daniel C. Dornbusch, and Raúl Labán, The Chilean Economy: Policy Esty, eds., Asian Dragons and Green Trade: Environment, Eco- Lessons and Challenges, ch. 2. Washington, D.C.: Brookings nomics, and International Law. Singapore: Times Academic Institution Press. Press. Dougherty, Sean M. 1997. "The Role of Foreign Technology in European Commission. 2001. "EU Approves `Everything But Arms' Improving Chinese Productivity." MIT Science and Technology Trade Access for Least Developed Countries." Press Release, Initiative, Beijing. Processed. February 26. Brussels. Drahos, Peter. 2000. "Indigenous Knowledge, Intellectual Property Evans, Gareth J. 1993. Cooperating for Peace: The Global Agenda and Biopiracy: Is a Global Bio-Collecting Society the Answer?" and Beyond. Sydney, Australia: Allen and Unwin. European Intellectual Property Review 22 (6, June): 245­50. Evans, Philip, and Thomas Wurster. 2000. Blown to Bits: How the ------. 2001. "Negotiating Intellectual Property Rights: New Economics of Information Transforms Strategy. Boston, Between Coercion and Dialogue." Presented to the Ninth Mass.: Harvard Business School Press. Annual Conference on International Intellectual Property Law Evenett, Simon J., and Bernard M. Hoekman. 2000. "Government and Policy, Fordham University School of Law, New York, Procurement of Services and Multilateral Disciplines." In Pierre April 19­20. Sauvé and Robert M. Stern, eds., GATS 2000: New Directions in Drake, William, and Kalypso Nicolaides. 2000. "The Information Services Trade Liberalization, 143­64. Washington, D.C.: Brook- Revolution and Services Trade Liberalization after 2000." In ings Institution Press and Harvard University. Pierre Sauvé and Robert M. Stern, eds., GATS 2000: New Direc- Evenett, Simon J., and Valerie Y. Suslow. 2000. "Preconditions on tions in Services Trade Liberalization. Washington, D.C.: Brook- Private Restraints on Market Access and International Cartels." ings Institution Press and Harvard University. Journal of International Economic Law 3 (4): 593­631. Drechsler, L. 1990. "A Note on the Concept of Services." Review of Evenett, Simon J., Margaret C. Levenstein, and Valerie Y. Suslow. Income and Wealth 36 (3): 309­16. 2001. "International Cartel Enforcement: Lessons from the Dutz, Mark. 2001. "Pre-Shipment Inspection." PREM Note 53. 1990s." World Economy 24 (9, September): 1221­45. World Bank, Washington, D.C. Evenson, Robert E., and Larry E. Westphal. 1997. "Technological Dyer, Geoff. 2001. "Patentes: Un ministro desafiante." Clarin (Feb- Change and Technology Strategy." In Hollis Chenery and T. N. ruary 12). Buenos Aires. Srinivasan, eds., Handbook of Development Economics, vol. 3. Eaton, J., and G. Grossman. 1986. "Optimal Trade and Industrial Amsterdam: North-Holland. Policy under Oligopoly." Quarterly Journal of Economics 101: Faini, Riccardo, and Jaime de Melo. 1990. "Adjustment, Invest- 383­406. ment and the Real Exchange Rate in Developing Countries." Ebrill, Liam, Janet Stotsky, and Reint Gropp. 1999. Revenue Impli- Economic Policy (October): 492­519. cations of Trade Liberalization. IMF Occasional Paper 180. FAO (Food and Agriculture Organization). 1995. The State of the Washington, D.C. World's Forests. Rome. Edwards, Sebastian. 1989. "Exchange Rate Misalignment in Develop- Feenstra, Robert C. 1996. "U.S. Imports, 1972­1994: Data and ing Countries." World Bank Research Observer 4 (1, January): 3­21. Concordances." NBER Working Paper 5515. National Bureau ------. 1993. "Openness, Trade Liberalization and Growth in of Economic Research, Cambridge, Mass. Developing Countries." Journal of Economic Literature 31 (3, ------. 2000. World Trade Flows, 1980­1997. March. University of September): 1358­93. California, Davis. Processed. ------. 1998. "Openness, Productivity and Growth: What Do We Feenstra, Robert C., Robert E. Lipsey, and Harry P. Bowen. 1997. Really Know?" Economic Journal 108: 383­98. World Trade Flows, 1970­1992 with Production and Tariff Data. Edwards, Sebastian, and Daniel Lederman. 1998. "The Political NBER Working Paper 5910. National Bureau of Economic Economy of Unilateral Trade Liberalization: The Case of Chile." Research, Cambridge, Mass. NBER Working Paper 6510. National Bureau of Economic Feketekuty, G. 2000. "Assessing and Improving the Architecture of Research, Cambridge, Mass. GATS." In Pierre Sauvé and Robert M. Stern, eds., GATS 2000: 610 Bibliography New Directions in Services Trade Liberalization, ch. 4. Washing- ------. 1998. "GATT Experience with Safeguards: Making Eco- ton, D.C.: Brookings Institution Press and Harvard University. nomic and Political Sense of the Possibilities That the GATT Fernandez, Raquel, and Jonathan Portes. 1998. "Returns to Allows to Restrict Imports." Policy Research Working Paper Regionalism: An Analysis of the Nontraditional Gains from Series 2000. World Bank, Development Research Group, Wash- Regional Trade Agreements." World Bank Economic Review 12: ington, D.C. 197­220. ------. 1999. Statement made at the Workshop on Developing Ficsor, Mihály. 1997. "Attempts to Provide International Protection Countries and the New Round of Multilateral Trade Negotia- for Folklore by Intellectual Property Rights." Presented to the tions, Harvard University, November 5­6. UNESCO-WIPO World Forum on the Protection of Folklore, ------. 2001. "Implementing the Uruguay Round Agreements: Phuket, Thailand, April 8­10. Problems for Developing Countries." World Economy 24 (9, Fikentscher, W., and U. Immenga, eds. 1995. Draft International September): 1097­1108. Antitrust Code. Baden-Baden, Germany: Nomos. Finger, J. M., and Dean DeRosa. 1980. "The Compensatory Findlay, Christopher, and Deunden Nikomborirak, 1999. "Air Finance Facility and Export Stabilization." Journal of World Transport." Paper prepared for East Asian Conference on Trade Law 14 (January­February): 14­22. Options for the WTO 2000 Negotiations, Manila, July 19­20. Finger, J. M., and Bernard Hoekman. 1999. "Developing Countries Available at and a New Trade Round: Lessons from Recent Research." http://www1.worldbank.org/wbiep/trade/wto2000_region_w World Bank, Washington, D.C. Processed. ork.html#EASTASIA. Finger, J. M., and M. Kreinin. 1976. "A New International Eco- Findlay, R. 1978. "Relative Backwardness, Direct Foreign Invest- nomic Order: A Critical Survey of the Issues." Journal of World ment and the Transfer of Technology: A Simple Dynamic Trade Law 16 (6, November): 493­512. Model." Quarterly Journal of Economics 92: 1­16. Finger, J. Michael, and Tracy Murray. 1993. "Antidumping and Finger, J. M. 1969. "Substitution and the Effective Rate of Protec- Countervailing Duty Enforcement in the United States." In J. tion." Journal of Political Economy 77 (6, December): 972­75. Michael Finger, ed., Antidumping, How It Works and Who Gets ------. 1974. "GATT Tariff Concessions and the Exports of Devel- Hurt. Ann Arbor: University of Michigan Press. oping Countries: United States Concessions at the Dillon Finger, J. Michael, and Julio Nogués. 2002. "The Unbalanced Round." Economic Journal 335: 566­75. Uruguay Round Outcome: The New Areas in Future WTO ------. 1975. "Tariff Provisions for Offshore Assembly and the Negotiations." World Economy 25 (3, March): 321­40. Export Earnings of Developing Countries." Economic Journal Finger, J. Michael, and Andrzej Olechowski, eds. 1987. The (June): 365­71. Uruguay Round: A Handbook for the Multilateral Trade Negotia- ------. 1976a. "Effects of the Kennedy Round Tariff Concessions tions. Washington, D.C.: World Bank. on the Exports of Developing Countries." Economic Journal 86 Finger, J. Michael, and Ludger Schuknecht. 2001. "Market Access (341): 87­95. Advances and Retreats: The Uruguay Round and Beyond." In ------. 1976b. "Trade and Domestic Effects of the Offshore Bernard Hoekman and Will Martin, eds., Developing Countries Assembly Provision in the United States Tariff." American Eco- and the WTO: A Pro-Active Agenda. Oxford, U.K.: Basil Blackwell. nomic Review 66 (4, September): 598­611. Finger, J. Michael, and Philip Schuler. 1999. "Implementation of ------.1979. "Trade Liberalization: A Public Choice Perspective." Uruguay Round Commitments: The Development Challenge." In Ryan C. Amacher, Gottfried Haberler, and Thomas D. Wil- Policy Research Working Paper Series 2215. World Bank, Devel- lett, eds., Challenges to a Liberal International Economic Order. opment Research Group, Washington, D.C. Washington, D.C.: American Enterprise Institute for Public Poli- ------. 2000. "Implementation of Uruguay Round Commitments: cy Research. The Development Challenge." World Economy 23 (4, April): ------. "Policy Research." Journal of Political Economy 89: 511­25. 1270­71. Finger, J. M., and L. A. Winters. 1998. "What Can the WTO Do for ------. 1982. "Incorporating the Gains from Trade into Policy." Developing Countries?" In Anne O. Krueger, ed., The WTO as World Economy 5 (December): 367­77. an International Organization. Chicago: University of Chicago ------. 1986. "Ideas Count, Words Inform." In R. H. Snape, ed., Press. Issues in World Trade Policy. London: Macmillan. Finger, J. M., and A. J. Yeats. 1976. "Effective Protection by Trans- ------. 1991a. "Development Economics and the GATT." In Jaime portation Costs and Tariffs: A Comparison of Magnitudes." de Melo and André Sapir, eds., Trade Theory and Economic Quarterly Journal of Economics (February) 90 (1, February): Reform--North, South, and East: Essays in Honor of Béla Balassa. 169­76. Cambridge, Mass.: Basil Blackwell. Finger, J. M., H. Keith Hall, and Douglas R. Nelson. 1982. "The ------. 1991b. "The GATT as International Discipline over Trade Political Economy of Administered Protection." American Eco- Restrictions: A Public Choice Approach." In Roland Vaubel and nomic Review 72 (3, June): 598­611. Thomas D. Willett, eds., The Political Economy of International Finger, J. Michael, Merlinda D. Ingco, and Ulrich Reincke. 1996. Organizations: A Public Choice Approach. Boulder, Colo.: West- "The Uruguay Round: Statistics on Tariff Concessions Given view. and Received." World Bank, Washington, D.C. ------. 1991c. "That Old GATT Magic No More Casts Its Spell Finger, J. M., Ulrich Reincke, and Adriana Castro. 2002. "Market (How the Uruguay Round Failed)." Journal of World Trade 25 Access Bargaining in the Uruguay Round: Rigid or Relaxed Rec- (2): 19­22. iprocity?" In Jagdish N. Bhagwati, ed., On Going Alone: The ------, ed. 1993. Antidumping: How It Works and Who Gets Hurt. Case for Relaxed Reciprocity in Freeing Trade. Cambridge, Mass.: Ann Arbor: University of Michigan Press. MIT Press. 611 D E V E L O P M E N T, T R A D E , A N D T H E W T O Fink, Carsten. 1999. "Entering the Jungle of Intellectual Property Gallup, John Luke, Steven Radelet, and Andrew Warner. 1998. Rights Exhaustion and Parallel Imports." In Owen Lippert, ed., "Economic Growth and the Income of the Poor." CAER II Dis- Competitive Strategies for the Protection of Intellectual Property. cussion Paper 36. Consulting Assistance on Economic Reform Vancouver, B.C.: Fraser Institute. (CAER II) Project, Harvard Institute for International Develop- ------. 2000. "How Stronger Patent Protection in India Might ment, Cambridge, Mass. Processed. Affect the Behavior of Transnational Pharmaceutical Indus- Ganslandt, Mattias, Keith E. Maskus, and Eina V. Wong, 2001. tries." Policy Research Working Paper Series 2352. World Bank, "Developing and Distributing Essential Medicines to Poor Washington, D.C. Countries: The DEFEND Proposal." World Economy 24: 779­95. Fink, Carsten, Aaditya Mattoo, and Randeep Rathindran. 2001. Garay, L. J., and Antoni Estevadeordal. 1994. "Apuntes sobre las "Liberalizing Basic Telecommunications: The Asian Experi- normas de origen como instrumento de política comercial." ence." Policy Research Working Paper Series 2718. World October. Inter-American Development Bank, Division of Inte- Bank, Washington, D.C. gration, Trade, and Hemispheric Issues. Washington, D.C. Foroutan, Faezeh. 1997. "Preconditions and Sustainability of Processed. Reforms." In John Nash and Faezeh Foroutan, eds., Trade Policy ------. 1996. "Protección, desgravación preferencial y normas de and Exchange Rate Reform in Sub-Saharan Africa, ch. 5. Devel- origen." Integración & Comercio 0 (January­April): 2­29. opment Issues 6. Canberra: National Centre for Development Garay, L. J., and Luis Felipe Quintero. 1997. "Caracterización, Studies. estructura y racionalidad de las normas de origen del G3 y de Fox, Eleanor M. 1997. "Toward World Antitrust and Market Access." la ALADI: Su relevancia en el caso de Colombia." Inter-Ameri- American Journal of International Law 91 (January): 1­25. can Development Bank, Division of Integration, Trade, and ------. 1999. "Competition Law and the Millennium Round." Hemispheric Issues. Washington, D.C. Processed. Journal of International Economic Law 2 (4, December): 665­79. Garcia Bercero, Ignacio, and Stefan D. Amarasinha. 2001. "Mov- Francois, Joseph 1997. "External Bindings and the Credibility of ing the Trade and Competition Debate Forward." Journal of Reform." In Ahmed Galal and Bernard Hoekman, eds., Regional International Economic Law 4 (3): 481. Partners in Global Markets: Limits and Possibilities of the Euro- GATT. 1958. Antidumping and Countervailing Duties. Geneva. Med Agreements. New York: Cambridge University Press. ------. 1993a. International Trade and the Trading General, ------. 1999. "Estimates of Barriers to Trade in Services." Erasmus 1992­93. Geneva. University, Rotterdam, The Netherlands. Processed. ------. 1993b. "Modalities for the Establishment of Specific Bind- ------. 2001. "Trade Policy Transparency and Investor Confi- ing Commitments under the Reform Program." dence: Some Implications for an Effective Trade Policy Review MTN.GNG/MA/W/24. ATT Goods Negotiating Group. Decem- Mechanism." Review of International Economics 9 (2, May): ber 20. 303­16. ------. 1994a. Analytical Index: Guide to GATT Law and Practice. Francois, J., and W. Martin. 2001. "Commercial Policy Variability, 6th ed. Geneva. Bindings, and Market Access." Tinbergen Institute. Processed. ------. 1994b. The Results of the Uruguay Round of Multilateral Francois, J., H. Glisman, and D. Spinanger. 2000. "The Cost of EU Trade Negotiations: The Legal Texts. Geneva: GATT Secretariat. Trade Protection in Textiles and Clothing." Working Paper 997. Gauthier, Gilles, and others. 2000. "Déjà Vu or New Beginning for Kiel Institute of World Economics, Germany. Safeguards and Subsidies Rules in Services Trade?" In Pierre Francois, Joseph F., Douglas Nelson, and N. David Palmeter. 1996. Sauvé and Robert M. Stern, eds., GATS 2000: New Directions in "Public Procurement: A Post­Uruguay Round Perspective." Services Trade Liberalization, 165­83. Washington, D.C.: Brook- Discussion Paper 1412. Centre for Economic Policy Research, ings Institution Press and Harvard University. London. Gervais, Daniel. 1998. The TRIPS Agreement: Drafting History and Francois, Joseph, Hugh Arce, Kenneth Reinert, and Joseph Flynn. Analysis. London: Sweet & Maxwell. 1996. "Commercial Policy and the Domestic Carrying Trade." ------. 2001. "Traditional Knowledge: A Challenge to the Interna- Canadian Journal of Economics 19: 181­98. tional Intellectual Property System." Presented to the Ninth Frankel, Jeffrey, and David Romer. 1999. "Does Trade Cause Annual Conference on International Intellectual Property Law and Growth?" American Economic Review 89 (3, June): 379­99. Policy, Fordham University School of Law, New York, April 19­20. Fraser, R. W. 1988. "A Method for Evaluating Supply Response to Geuze, Matthijs, and Hannu Wager. 1999. "WTO Dispute Settle- Price Underwriting." Australian Journal of Agricultural Economics ment Practice Relating to the TRIPS Agreement." Journal of 32 (1): 22­36. International Economic Law 2 (2): 347­84. Frazer, T., and P. Holmes. 1995. "Self-Restraint: Cars Complaints Ghei, Nita, and Lawrence E. Hinkle. 1999. "A Note on Nominal and the Commission." European Public Law 1 (1): 85­95. Devaluations, Inflation and the Real Exchange Rate." In Freund, J., and R. E. Walpole. 1980. Mathematical Statistics. 3d ed. Lawrence Hinkle and Peter Montiel, eds., Exchange Rate Mis- Englewood Cliffs, N.J.: Prentice-Hall. alignment: Concepts and Measurement for Developing Countries. Fried, Jonathan T. 2000. "The Importance of Integrating Trade, A World Bank Research Publication. New York: Oxford Univer- Finance and Development Perspectives. Notes for Remarks at sity Press. the Integrated Framework Seminar on the Policy Relevance of Ghei, Nita, and Stephen B. Kamin. 1999. "The Use of the Parallel Mainstreaming Trade into Country Development Strategies: Market Rate as a Guide to Setting the Official Exchange Rate." Perspectives of Least-Developed Countries." January 29. WTO, In Lawrence Hinkle and Peter Montiel, eds., Exchange Rate Mis- Geneva. Processed. alignment: Concepts and Measurement for Developing Countries. Gabunia, Georgi. 1998. "Reasonable Protectionism." Expert, no. A World Bank Research Publication. New York: Oxford Univer- 33 (September). sity Press. 612 Bibliography Ghei, Nita, and Lant Pritchett. 1999. "The Three Pessimisms: Real Guruswamy and Jeffrey A. McNeely, eds., Protection of Global Exchange Rates and Trade Flows in Developing Countries." In Biodiversity: Converging Strategies. Durham, N.C.: Duke Univer- Lawrence Hinkle and Peter Montiel, eds., Exchange Rate Mis- sity Press. alignment: Concepts and Measurement for Developing Countries. Guruswamy, Lakshman D., and Jeffrey A. McNeely, eds., 1998. A World Bank Research Publication. New York: Oxford Univer- Protection of Global Biodiversity: Converging Strategies. Durham, sity Press. N.C.: Duke University Press. Ghura, Dhaneshwar, and Thomas J. Grennes. 1993. "The Real Haddad, Mona, and Ann Harrison. 1993. "Are There Positive Exchange Rate and Macroeconomic Performance in Sub-Saha- Spillovers from Direct Foreign Investment? Evidence from ran Africa." Journal of Development Economics 42 (1, October): Panel Data for Morocco." Journal of Development Economics 42: 155­74. 51­74. Gibbs, M., and M. Mashayekhi. 1998. "The Uruguay Round Hallaert, Jean-Jacques. 2000. "Un bilan a mi-parcours du SPG Negotiations on Investment: Lessons for the Future." UNC- Europeen: impact du volet industriel sur les pays en devel- TAD, Division on Trade in Goods and Services and Commodi- oppement d'Asie." Science Po (GEM), Paris. Processed. ties, Geneva. Processed. Hanson, Gordon H., and Robert C. Feenstra. 2001. "Intermedi- Gibney, Mark, and David Emerick. 1996. "The Extraterritorial aries in Entrepôt Trade: Hong Kong Re-exports of Chinese Application of U.S. Law and the Protection of Human Rights: Goods." NBER Working Paper 8088. National Bureau of Eco- Holding Multinational Corporations to Domestic and Interna- nomic Research, Cambridge, Mass. tional Standards." Temple International and Comparative Law Harberger, A. C. 1988. Reflections on Uniform Taxation. Presented Journal 10: 123­45. to the 44th Congress of the International Institute of Public Glass, Amy J., and Kamal Saggi. 1998. "Multinational Firms and Finance, Istanbul. Technology Transfer." Policy Research Working Paper Series Hardin, Alexis, and Leanne Holmes. 1997. Services Trade and For- 2067. World Bank, Washington, D.C. eign Direct Investment. Staff Research Paper, Industry Commis- Glassman, C. A. 2000. "Customer Benefits from Current Informa- sion. Canberra: Australian Government Publishing Services. tion Sharing by Financial Services Companies." Study for the Harris, R. G., and N. Schmitt. 1999. "Do Two Wrongs Make a Financial Services Roundtable, December. Right? Export Incentives and Bias in Trade Policy." In J. Piggott Global Currency Report. 1999. Vol. 44, no. 1 (January­February­ and A. D. Wodland, eds., International Trade Policy and the Pacif- March). Brooklyn, N.Y.: Currency Data & Intelligence, Inc. ic Rim. Houndsmill, Basingstoke, Hampshire, U.K.: Macmillan. Globerman, Steven. 1979. "Foreign Direct Investment and Harrison, Ann. 1994. "Productivity, Imperfect Competition, and `Spillover' Efficiency Benefits in Canadian Manufacturing Indus- Trade Reform: Theory and Evidence." Journal of International tries." Canadian Journal of Economics 12: 42­56. Economics 36: 53­73. Gould, David M., and William C. Gruben. 1996. "The Role of Harrison, Donald. 1994. "Rules of Origin under the North Ameri- Intellectual Property Rights in Economic Growth." Journal of can Free Trade Agreement." In Judith H. Bello, Alan F. Holmes, Development Economics 48: 323­50. and Patrick M. Norton, eds., North American Free Trade Agree- Graham, Edward, and E. Wada. Forthcoming. "Foreign Direct ment: A New Frontier in International Trade and Investment in Investment in China: Effects on Growth and Economic Perfor- the Americas, ch. 3. Chicago: American Bar Association. mance." In Peter Drusdale, ed., Achieving High Growth: Experi- Harrison, Glenn W., Thomas F. Rutherford, and David G. Tarr. ence of Transitional Economies in East Asia. Oxford, U.K.: Oxford 1993. "Trade Reform in the Partially Liberalized Economy of University Press. Turkey." World Bank Economic Review 7 (May): 191­217. Greenaway, D. 1992. "Trade-Related Investment Measures and ------. 1996. "Increased Competition and Completion of the Development Strategies." Kyklos 45: 139­59. Market in the European Union: Static and Steady State Grossman, Gene. 1981. "The Theory of Domestic Content Protec- Effects." Journal of Economic Integration 11 (3): 332­65. tion and Content Preference." Quarterly Journal of Economics ------. 1997. "Trade Policy Options for Chile: A Quantitative Eval- 96 (4, November): 583­603. uation." Policy Research Working Paper Series 1783. World ------. 1990. "Promoting New Industrial Activities." OECD Eco- Bank, Washington, D.C. nomic Studies (14): 87­125. Harrold, Peter, Malathi Jayawickrama, and Deepak Bhattasali. 1996. Grossman, Gene M., and Elhanan Helpman. 1991. Innovation and Practical Lessons for Africa from East Asia in Industrial and Trade Growth in the World Economy. Cambridge, Mass.: MIT Press. Policies. World Bank Discussion Paper 310. Washington, D.C. ------. 1994. "Protection for Sale." American Economic Review 84 Hathaway, D., and M. Ingco. 1996. "Agricultural Liberalization (4): 873­50. under the Uruguay Round." In Will Martin and L. Alan Winters, ------. 1995. "Technology and Trade." In Gene M. Grossman eds., The Uruguay Round and the Developing Economies. New and Kenneth Rogoff, eds., Handbook of International Economics, York: Cambridge University Press. vol. 3. Amsterdam: North-Holland. Hatta, Tatsuo. 1977. "A Theory of Piecemeal Policy Recommenda- Grossman, Gene, and Carl Shapiro. 1988a. "Counterfeit-Product tions." Review of Economic Studies 44: 1­21. Trade." American Economic Review 78 (1): 59­75. ------. 1986. "Welfare Effects of Changing Commodity Tax Rates ------.1988b. "Foreign Counterfeiting of Status Goods." Quarter- toward Uniformity." Journal of Public Economics 29 (1, Febru- ly Journal of Economics 103 (1): 79­100. ary): 99­112. Guisinger, Steven, and associates. 1985. Investment Incentives and Hausman, J. A., and J. MacKie-Mason. 1988. "Price Discrimination Performance Requirements. New York: Praeger. and Patent Policy." RAND Journal of Economics 19 (2): 253­65. Gupta, Anil K. 1998. "Rewarding Local Communities for Conserv- Heal, Geoffrey M. 1998. "Markets and Biodiversity." In Lakshman ing Biodiversity: The Case of the Honey Bee." In Lakshman D. D. Guruswamy and Jeffrey A. McNeely, eds., Protection of Glob- 613 D E V E L O P M E N T, T R A D E , A N D T H E W T O al Biodiversity: Converging Strategies. Durham, N.C.: Duke Uni- ------. 1995. Trade Laws and Institutions: Good Practices and the versity Press. World Trade Organization. World Bank Discussion Paper 282. Heeks, R. 1998. "The Uneven Profile of Indian Software Exports." Washington, D.C. Working Paper 3. Institute for Development Policy and Man- ------. 1996. "Assessing the General Agreement on Trade in Ser- agement. University of Manchester, U.K. vices." In Will Martin and L. Alan Winters, eds., The Uruguay Helleiner, Gerald K., ed. 2001. Non-Traditional Exports and Devel- Round and the Developing Economies. Cambridge, U.K.: Cam- opment in Sub-Saharan Africa: Experience and Issues. Helsinki: bridge University Press. World Institute for Development Economics Research. ------. 1997. "Competition Policy and the Global Trading Sys- Henderson, David. 1999. The Multilateral Agreement in Investment: tem." World Economy 20: 383­406. A Story and Its Lessons. Wellington: New Zealand Roundtable. ------. 1998a. "Preferential Trade Agreements." In Robert Z. Available at < www.nzbr.org.nz >. Lawrence, ed., Brookings Trade Forum: 1998, 299­320. Wash- Henson, Spencer, Rupert Loader, Alan Swinbank, Maury Bredahl, ington, D.C.: Brookings Institution Press. and Nicole Lux. 2000. "Impact of Sanitary and Phytosanitary ------. 1998b. "Using International Institutions to Improve Public Measures on Developing Countries." Centre for Food Econom- Performance." World Bank Research Observer 13 (2): 249­69. ics Research, Department of Agricultural and Food Economics, ------.2000. "The Next Round of Services Negotiations: Identify- University of Reading, U.K. ing Priorities and Options." Federal Reserve Bank of St. Louis Herin, Jan. 1986. "Rules of Origin and Differences between Tariff Review 82: 31­47. Levels in EFTA and in the EC." AELE Occasional Paper 13. ------. 2002. "Strengthening the Global Trade Architecture for Geneva. Development." World Trade Review 1: 23­46. Hertel, Thomas W. 2000. "Potential Gains from Reducing Trade Hoekman, Bernard, and Carlos A. Primo Braga. 1997. "Protection Barriers in Manufacturing, Services and Agriculture." Federal and Trade in Services: A Survey." Open Economies Review 8: Reserve Bank of St. Louis Review 82: 77­99. 285­308. Hertel, Thomas, and Will Martin. 2000. "Liberalising Agriculture Hoekman, Bernard, and Peter Holmes. 1999. "Competition Policy, and Manufactures in a Millennium Round: Implications for Developing Countries and the WTO." World Economy 22 (6, Developing Countries." World Economy 23 (4, April): 455­69. August): 875­94. Hertel, T., K. Anderson, J. F. Francois, and W. Martin. Forthcoming. Hoekman, Bernard, and Michel M. Kostecki. 2001. The Political "Agriculture and Non-agricultural Liberalization in the Millen- Economy of the World Trading System: The WTO and Beyond. 2d nium Round." In M. D. Ingco and L. A. Winters, Agriculture and ed. New York: Oxford University Press. the New Trade Agenda from a Development Perspective. Cam- Hoekman, Bernard, and Michael P. Leidy. 1992. "Holes and Loop- bridge and New York: Cambridge University Press. holes in Regional Trade Arrangements and the Multilateral Hill, H., and P. Arthukorala. 1998. "Foreign Investment in East Trading System." In Kym Anderson and Richard Blackhurst, Asia." Asian Pacific Economic Literature 12 (2): 23­50. eds., Regional Integration and the Global Trading System, Hill, T. P. 1977. "On Goods and Services." Review of Income and 325­60. New York: St. Martin's Press. Wealth (23): 315­38. Hoekman, Bernard, and Will Martin, eds. 2001. Developing Coun- Hindley, Brian, and Alasdair Smith. 1984. "Comparative Advan- tries and the WTO: A Pro-active Agenda. Oxford, U.K.: Blackwell tage and Trade in Services." World Economy 7: 369­90. Publishers. Hines, J. R. 1993. "Altered States: Taxes and the Location of FDI in Hoekman, Bernard, and Aaditya Mattoo. 2000. "Services, Eco- America." NBER Working Paper 4397. Cambridge, Mass.: nomic Development and the Next Round of Negotiations on National Bureau of Economic Research. Services." Journal of International Development 12 (2, March): Hirschman, Albert O. 1958. The Strategy of Economic Development. 283­96. New Haven, Conn.: Yale University Press. Hoekman, Bernard, and Petros Mavroidis. 1994. "Competition, ------. 1981a. Essays in Trespassing: Economics to Politics and Competition Policy, and the GATT." World Economy 17: Beyond. Cambridge, U.K.: Cambridge University Press. 121­50. ------. 1981b. "Three Uses of Political Economy in Analyzing ------, eds. 1997. Law and Policy in Public Purchasing: The WTO European Integration." In Albert O. Hirschman, Essays in Tres- Agreement on Government Procurement. Ann Arbor: University passing, ch. 12. London: Cambridge University Press. of Michigan Press. Hobday, Michael. 1995. Innovation in East Asia: The Challenge to ------.2000. "WTO Dispute Settlement, Transparency and Sur- Japan. Cheltenham, U.K.: Edward Elgar. veillance." World Economy 23 (4): 527­42. Hodge, J. 2000. "Liberalising Communications Services in South Hoekman, Bernard, and Patrick A. Messerlin. 2000. "Liberalizing Africa." Development Southern Africa 17 (3, September): Trade in Services: Reciprocal Negotiations and Regulatory 373­87. Reform." In Pierre Sauvé and Robert M. Stern, eds., GATS ------. 2001. "Estimating the Cost of Services Protection in a 2000: New Directions in Services Trade Liberalization. Washing- Developing Economy: The Case of South Africa." In Robert M. ton, D.C.: Brookings Institution Press and Harvard University. Stern, ed., Services in the International Economy. Ann Arbor: Hoekman, Bernard, and Kamal Saggi. 1999. "Assessing the Case University of Michigan Press. for WTO Disciplines on Investment Related Policies." Journal of Hodge, J., and H. Nordas. 1999. "Trade in Services: The Impact on Economic Integration 15: 588­610. Developing Countries." Report to Norwegian Foreign Ministry. Hoekman, Bernard, H. L. Kee, and Marcelo Olarreaga. 2001. Processed. "Markups under Domestic and Foreign Entry." Background Hoekman, Bernard. 1993. "Rules of Origin for Goods and Ser- paper for World Development Report 2001. World Bank, Wash- vices." Journal of World Trade (August): 82­99. ington, D.C. 614 Bibliography Hoekman, B., P. Low, and P. Mavroidis. 1998. "Regulation, Com- ICRIER (Indian Council for Research on International Economic petition Policy, and Market Access Negotiations: Lessons from Relations). 1999. [Sectoral reports on health services, legal the Telecommunications Sector." In Einar Hope and Per Mae- services, accountancy services, construction and engineering leng, eds., Competition and Trade Policies, 115­140. London: services, and software services.] Prepared for the Ministry of Routledge. Commerce, New Delhi. Hoekman, Bernard, Francis Ng, and Marcelo Olarreaga. 2001. ILO (International Labour Organization). 1998. Labor and Social "Eliminating Excessive Tariffs in the QUAD and Least Devel- Issues Relating to Export Processing Zones. Geneva. oped Country Exports." Policy Research Working Paper Series IMF (International Monetary Fund). 1999. Annual Report on 2604. World Bank, Washington, D.C. Exchange Arrangements and Exchange Restrictions, 1999. Wash- Hoff, Karla, and Joseph E. Stiglitz. 2000. "Modern Economic Theo- ington, D.C. ry and Development." In Gerald M. Meier and Joseph E. IMF. 2000. International Capital Markets: Developments, Prospects Stiglitz, eds., Frontiers of Development Economics: The Future in and Key Policy Issues. World Economic and Financial Surveys. Perspective. New York: Oxford University Press. September. Washington, D.C. Holmes, P., J. Kempton, and F. McGowan. 1996. "International IMF. Various issues. Direction of Trade Statistics. Washington, D.C. Competition Policy and Telecommunications: Lessons from the Ingco, Merlinda, John Nash, and others. Forthcoming. "Agriculture, EU and Prospects for the WTO." Telecommunications Policy 20 the New Trade Agenda, and the WTO Round: Creating A Trad- (10): 755­67. ing System for Development." World Bank, Washington, D.C. Honglin Zang, K., and J. Markusen. 1999. "Vertical Multinationals ITCB (International Textiles and Clothing Bureau). 1999. "Experi- and Host-Country Characteristics." Journal of Development Eco- ence with the Implementation of the ATC: Main Areas of Con- nomics 59: 233­52. cern, Article-by-Article." April. Geneva. Hope, Einar, and Per Maeleng, eds. 1998. Competition and Jackson, John H. 1997. The World Trading System. 2d ed. Cam- Trade Policies. London: Routledge. Available at bridge, Mass.: MIT Press. . Policy." In Simon J. Evenett, Alexander Lehmann, and Benn Hudec, Robert E. 1978. Adjudication of International Trade Disputes. Steil, eds., Antitrust Goes Global: What Future for Transatlantic London: Trade Policy Research Centre. Cooperation? 29­56. Washington, D.C.: Brookings Institution ------. 1987. Developing Countries in the GATT Legal System. Lon- Press and Harvard University. don: Trade Policy Research Centre. JETRO (Japanese External Trade Relations Organization). 2001. ------. 1990. The GATT Legal System and World Trade Diplomacy. White Paper, "Foreign Direct Investment 2001: Accelerated Cor- 2d ed. Salem, N.H.: Butterworths. porate Realignment through Mergers and Acquisitions." Tokyo. ------. 1993. Enforcing International Trade Law. Salem, N.H.: But- Jha, Veena, and René Vossenaar. 2001. "Breaking the Deadlock: A terworths. Positive Agenda on Trade, Environment and Development?" In Hufbauer, Gary Clyde, and Kimberly Ann Elliott. 1993. Measuring Gary P. Sampson and W. Bradnee Chambers, eds., Trade, Envi- the Costs of Protection in the United States. Washington, D.C.: ronment, and the Millennium. 2d ed. Tokyo: United Nations Institute for International Economics. University. Hufbauer, Gary Clyde, and Tony Warren. 1999. "The Globalization Jones, P., and J. Hudson. 1996. "Standardization and the Costs of of Services: What Has Happened? What Are the Implications?" Assessing Quality." European Journal of Political Economy 12: Working Paper 99-12. October. Institute for International Eco- 355­61. nomics, Washington, D.C. Junta del Acuerdo de Cartagena. 1995. "Proyecto de resolución Hugenholtz, P. Bernt. 2000. "Caching and Copyright: The Right of sobre criterios y procedimientos para fijar requisitos específicos Temporary Copying." European Intellectual Property Review 22 de origen (REOS)." August. Lima. (10): 482­93. Kalirajan, Kaleeswaran, Greg McGuire, Duc Nguyen-Hong, and Hull, Cordell. 1948. The Memoirs of Cordell Hull. New York: Michael Schuele. 2001. "The Price Impact of Restrictions on Macmillan. Banking Services." In Christopher Findlay and Tony Warren, Ianchovichina, Elena, W. Martin, and E. Fukase. 2000. "Modelling eds., Impediments to Trade in Services: Measurement and Policy the Impact of China's Accession to the WTO." Prepared for the Implications. London and New York: Routledge. Third Annual Conference on Global Economic Analysis, Kanbur, Ravi. 2001. "Economic Policy, Distribution and Poverty: Monash University, Adelaide, Australia, June 27­30. Available The Nature of Disagreements." Cornell University, Ithaca, New at < www.monash.edu.au/policy/conf2000.htm >. York. Processed. Available at Ianchovichina, Elena, Aaditya Mattoo, and Marcelo Olarreaga. . 2001. "Unrestricted Market Access for Sub-Saharan Africa: Kang, Nam-Hoog, and Sara Johansson. 2000. "Cross-Border How Much Is It Worth and Who Pays for It?" Policy Research Mergers and Acquisitions: Their Role in Industrial Globaliza- Working Paper Series 2595. Washington, D.C., World Bank. tion." STI Working Paper 2000/1. OECD, Paris. Ianchovichina, Elena, Robert McDougall, and Thomas Hertel. Kapur, Devesh, and Richard Webb. 2000. "Governance-Related 1999. "A Disequilibrium Model of International Capital Mobili- Conditionalities of the International Financial Institutions." ty." Presented at the Second Annual Conference on Global Group of 24 Discussion Paper 6. Available at Economic Analysis, Ebberuk, Denmark, June 20­22. . ICPAC (International Competition Policy Advisory Committee). Karsenty, Guy. 2000. "Just How Big Are the Stakes? An Assessment 2000. "Final Report." Available at of Trade in Services by Mode of Supply." In Pierre Sauvé and . Robert M. Stern, eds., GATS 2000: New Directions in Services 615 D E V E L O P M E N T, T R A D E , A N D T H E W T O Trade Liberalization. Washington, D.C.: Brookings Institution Krueger, Anne O. 1984. "Trade Policies in Developing Countries." Press and Harvard University. In R. W. Jones, P. B. Jones, and P. B. Kenen, eds. Handbook of Kaufmann, Daniel, Aart Kraay, and Pablo Zoido-Lobatón. 1999. International Economics. Amsterdam: North-Holland. "Governance Matters." Policy Research Working Paper Series ------. 1990. "Asian Trade and Growth Lessons." American Eco- 2196. World Bank, Washington, D.C. nomic Review: Papers and Proceedings 80: 108­12. Kaul, Inge, Isabelle Grunberg, and Marc Stern, eds. 1999. Global ------. 1993. "Free Trade Agreements as Protectionist Devices: Public Goods: International Co-operation in the 21st Century. Rules of Origin." NBER Working Paper 4352. National Bureau New York: Oxford University Press. of Economic Research, Cambridge, Mass. Keesing, Donald B. 1982. Exporting Manufactured Consumer Goods ------. 1995. "Partial Adjustment and Growth in Turkey." In Rudi- from Developing to Developed Economies: Marketing by Local ger Dornbusch and Sebastian Edwards, eds., Reform, Recovery, Firms and Effects of Developing Country Policies. Washington, and Growth: Latin America and the Middle East, ch. 11. A D.C.: World Bank. National Bureau of Economic Research Project Report. Chica- ------. 1998. Improving Trade Policy Reviews in the World Trade go: University of Chicago Press. Organization. Policy Analysis in International Economics. Wash- ------. 1997. "Free Trade Agreements versus Customs Unions." ington, D.C.: Institute for International Economics. Journal of Development Economics 54: 169­87. Kehoe, Timothy. 1999. "Capital Flows and North American Eco- Krueger, Anne O., Maurice Schiff, and Alberto Valdés. 1988. "Agri- nomic Integration." In Richard E. Baldwin and Joseph F. Fran- cultural Incentives in Developing Countries: Measuring the cois, eds., Dynamic Issues in Commercial Policy Analysis. Effects of Sectoral and Economy-wide Policies." World Bank Cambridge, U.K.: Cambridge University Press. Economic Review 2 (3, September): 255­71. Kelegama, Saman, and Fritz Foley. 1999, "Impediments to Pro- Krugman, Paul. 1989. "Is Free Trade Passe?" Journal of Economic moting Backward Linkages from the Garment Industry in Sri Perspectives 1: 131­44. Lanka." World Development 27 (8): 1445­60. ------. 1992. "Does the New Trade Theory Require a New Trade Kemp, M. C. 1964. The Pure Theory of International Trade. Engle- Policy?" World Economy 15: 423­42. wood Cliffs, N.J.: Prentice-Hall. ------. 1996. Rethinking International Trade. Cambridge, Mass.: Kennan, Jane, and Christopher Stevens. 1997. "From Lomé to the MIT Press. GSP: Implications for the ACP of Losing Lomé Trade Prefer- Kubota, Keiko. 2000. "Fiscal Constraints, Collection Costs and ences." Institute of Development Studies, University of Sussex, Trade Policies." Policy Research Working Paper Series 2366. Brighton, U.K. World Bank, Washington, D.C. Khan, Mohsin, and Malcolm Knight. 1988. "Import Compression Kur, Annette. 2000. "The WIPO Recommendations for the Protec- and Export Performance in Developing Countries." Review of tion of Well-Known Marks." International Review of Industrial Economics and Statistics 70 (May): 315­21. Property and Copyright Law 31: 824­45. Kindleberger, C. 1983. "Standards as Public, Collective and Private Labán, Raúl, and Felipe Larraín. 1995. "Continuity, Change, and Goods." Kyklos 36: 377­96. the Political Economy of Transition in Chile." In Rudiger Dorn- King, Robert G., and Ross Levine. 1993. "Finance and Growth." busch and Sebastian Edwards, eds., Reform, Recovery, and Quarterly Journal of Economics 108: 717­38. Growth: Latin America and the Middle East, ch. 4. A National Kirk, Michael. 1997. "Issues Concerning the Implementation of Bureau of Economic Research Project Report. Chicago: Univer- the TRIPS Agreement from the Point of View of the Private Sec- sity of Chicago Press. tor." In WIPO, WIPO Asian Regional Roundtable on Implementa- Lacarte-Muro, Julio, and Petina Gappah. 2000. "Developing tion of the Agreement on Trade-Related Aspects of Intellectual Countries and the WTO Legal and Dispute Settlement System: Property Rights, Singapore, 1st­17th January. A View from the Bench." Journal of International Economic Law Kirmani, N., and others. 1996. "The Uruguay Round and Interna- 3 (3): 395­401. tional Trade in Textiles and Clothing." In Said El-Naggar, ed., Lafer, Celso. 1996. "O sistema de solucao de controversias da The Uruguay Round and the Arab Countries. Washington, D.C.: Organizacao Mundial do Comercio." Estudos, FIESP/IRS/CNI, IMF. Brazil. Kitchenman, Walter F. 1999. "U.S. Credit Reporting: Perceived Laffont, Jean Jacques. 1999. "Competition, Information, and Benefits Outweigh Privacy Concerns." The Tower Group. Avail- Development." Annual World Bank Conference on Development able at Economics 1998. Washington, D.C.: World Bank. . Laird, Sam. 1997. "WTO Rules and Good Practice on Export Poli- Knauth, B. 2000. "Foreign Owned Enterprises." Statistics in Focus. cy." Staff Working Paper TPRD9701. World Trade Organiza- Brussels:Eurostat. tion, Geneva. Komuro, Norio. 1995. "The WTO Dispute Settlement Mechanism: ------. 1999a. "Issues for the Forthcoming Multilateral Negotia- Coverage and Procedures of the WTO Understanding." Journal tions in Agriculture." Prepared for a seminar on Agricultural of World Trade 29 (4, August): 5­95. Marketing in Developed Countries, Insituto Agronomico Konan, Denise E., and Keith Maskus. 2000. "Service Liberalization Mediterraneo de Zaragoza. October 4. in WTO 2000: A Computable General Equilibrium Model of ------. 1999b. "Millennium Round Market Access Negotiations in Tunisia." Processed. Goods and Services." CREDIT Research Paper 00/4. University Krishna, Kala, and Anne O. Krueger. 1995. "Implementing Free of Nottingham, U.K. Trade Areas: Rules of Origin and Hidden Protection." NBER ------. 1999c. "Multilateral Approaches to Market Access Negoti- Working Paper 4983. National Bureau of Economic Research, ations." In Miguel Rodríguez Mendoza, Patrick Low, and Bar- Cambridge, Mass. bara Kotschwar, eds., Trade Rules in the Making: Challenges in 616 Bibliography Regional and Multilateral Negotiations. Washington, D.C.: tries." Background paper for World Development Report 2001. Brookings Institution Press and Organization of American World Bank, Washington, D.C. States. Levin, Richard C., Alvin K. Klevorick, Richard Nelson, and Sidney Lall, Sanjaya. 1980. "Vertical Inter-Firm Linkages in LDCs: An G. Winter. 1987. "Appropriating the Returns from Industrial Empirical Study." Oxford Bulletin of Economics and Statistics 42: Research and Development." Brookings Papers on Economic 203­06. Activity 3: 783­820. ------. 1994. "The East Asian Miracle: Does the Bell Toll for Liebl, Maureen, and Tirthankar Roy. 2000. Handmade in India. Industrial Strategy?" World Development 22: 645­54. New Haven, Conn.: Policy Sciences Center. Lang, W. 1996. "WTO Dispute Settlement: What the Future Limão, Nuno, and Anthony J. Venables. 1999. "Infrastructure, Holds." In Simon S. C. Tay and Daniel C. Esty, eds., Asian Drag- Geographical Disadvantage and Transport Costs." Policy ons and Green Trade: Environment, Economics, and International Research Working Paper Series 2257. World Bank, Trade, Law. Singapore: Times Academic Press. Development Research Group, Washington, D.C. Langhammer, Rolf J., and M. Lucke. 1999. "WTO Accession Lin, Justin Yifu, and Jeffrey B. Nugent. 1995. "Institutions and Eco- Issues." Kiel Working Paper 905. February. Kiel Institute of nomic Development." In Jere R. Behrman and T. N. Srinivasan, World Economics, Germany. eds., Handbook of Economic Development, vol. 3A. Amsterdam: Langhammer, Rolf, and André Sapir. 1988. Economic Impact of North-Holland. Generalized Tariff Preferences. London: Trade Policy Research Lipsey, R., and K. Lancaster. 1956. "The General Theory of Second Centre. Best." Review of Economic Studies 25: 11­32. Lanjouw, Jean O., and Iain Cockburn. 2000. "Do Patents Matter? Lloyd, P. J. 1974. "A More General Theory of Price Distortions in Empirical Evidence after GATT." NBER Working Paper 7495. Open Economies." Journal of International Economics 4: National Bureau of Economic Research, Cambridge, Mass. 365­86. La Porta, R., F. Lopez de Silanes, and A. Shleifer. 2000. "Govern- ------. 1993. "A Tariff Substitute for Rules of Origin in Free Trade ment Ownership of Banks." HIER Discussion Paper 1890: 1­55. Areas." World Economy 16 (November): 699­712. Harvard Institute of Economic Research, Cambridge, Mass. López, Ramón, John Nash, and Julie Stanton. 1995. "Adjustment Lawrence, Robert Z., ed. 1998. Brookings Trade Forum: 1998. and Poverty in Mexican Agriculture: How Farmers' Wealth Washington, D.C.: Brookings Institution Press. Affects Supply Response." Policy Research Working Paper Lawson, Cina, and Natalie Meyenn. 2000. "Bringing Cellular Phone Series 1494. World Bank, International Economics Department, Service to Rural Areas: Grameen Telecom and Village Pay Phones Washington, D.C. in Bangladesh." Public Policy for the Private Sector Note 205. Lopez-de-Silanes, Florencio, James R. Markusen, and Thomas F. March. World Bank Group, Finance, Private Sector, and Infra- Rutherford. 1993. "Anti-Competitive and Rent-Shifting Aspects structure Network, Washington, D.C. Available at . NBER Working Paper 4512. National Bureau of Economic Leamer, Edward. 1994. "American Regionalism and Global Free Research, Cambridge, Mass. Trade." NBER Working Paper 4753. National Bureau of Eco- Low, Patrick. 1995. "Preshipment Inspection Services." World nomic Research, Cambridge, Mass. Bank, Washington, D.C. Processed. Leary, V. 1996. "Workers' Rights and International Trade: The Low, Patrick, and Aaditya Mattoo. 2000. "Is There a Better Way? Social Clause." In Jagdish N.Bhagwati and Robert E. Hudec, Alternative Approaches to Liberalization under GATS." In Pierre eds., Fair Trade and Harmonization: Prerequisites for Free Trade? Sauvé and Robert M. Stern, eds., GATS 2000: New Directions in 2 vols. Cambridge, Mass.: MIT Press. Services Trade Liberalization. Washington, D.C.: Brookings Insti- Lee, Jong-Wha. 1996. "Government Interventions and Productivi- tution Press and Harvard University. ty Growth in Korean Manufacturing Industries." Journal of Eco- Low, Patrick, A. Mattoo, and A. Subramanian. 1997. "Government nomic Growth 1: 391­414. Procurement in Services." In Bernard Hoekman and Petros Leiberman, Ira. 1990. Industrial Restructuring: Policy and Practice. Mavroidis, eds., Law and Policy in Public Purchasing: The WTO Policy and Research Series 9. Washington, D.C.: World Bank. Agreement on Government Procurement, 225­42. Ann Arbor: Leidy, Michael, and Bernard Hoekman. 1993. "What to Expect University of Michigan Press. from Regional and Multilateral Trade Negotiations: A Public Luke, David F. 2000. "OAU/AEC Member States, the Seattle Choice Perspective." In Kym Anderson and Richard Blackhurst, Preparatory Process and Seattle: A Personal Reflection." Journal eds., Regional Integration and the Global Trading System. Lon- of World Trade 34 (3, June): 41­46. don: Harvester-Wheatsheaf. Lundberg, Mattias, and Lyn Squire. 2000. "The Simultaneous Evo- Lengyel, M. 2000. "Los países latinoamericanos y la agenda com- lution of Growth and Inequality." October. World Bank, Wash- ercial multilateral: Implicancias y perspectivas." Processed. ington, D.C. Lengyel, Miguel, and Diana Tussie. 2000. "Promoción de exporta- Maamoun Abdel-Fattah, M., and H. Kheir el Din. 2000. "Textiles ciones y disciplinas multilaterales: Posibilidades e interro- and Clothing in the Mediterranean Region: Opportunities and gantes." In M. Olarreaga and R. Rocha, eds., La Nueva Agenda Challenges of Returning Textiles and Clothing to GATT Disci- de Comercio en la OMC. Santa Fé de Bogotá: World Bank Insti- plines." Presented to the Mediterranean Development Forum tute and Universidad del Rosario. (MDF3), March. Lerner, A. 1936. "The Symmetry between Import and Export Madani, Dorsali. 1999. "A Review of the Role and Impact of Taxes." Economica 3 (11, August): 306­13. Export Processing Zones." Policy Research Working Paper Levenstein, Margaret C., and Valerie Y. Suslow. 2001. "Private Series 2238. World Bank, Development Research Group, Wash- International Cartels and Their Effect on Developing Coun- ington, D.C. 617 D E V E L O P M E N T, T R A D E , A N D T H E W T O Malan, Rian. 2000. "In the Jungle." Rolling Stone (May 25). Martin, Will, and Devashish Mitra. 2001. "Productivity Growth Malueg, David A., and Marius Schwartz. 1994. "Parallel Imports, and Convergence in Agriculture and Manufacturing." Econom- Demand Dispersion, and International Price Discrimination." ic Development and Cultural Change 49 (2, January): 403­22. Journal of International Economics 37: 167­96. Martin, Will, and Mari Pangestu. Forthcoming. Options for Global Mann, Catherine. 1999. Is the U.S. Trade Deficit Sustainable? Wash- Trade Reform: A View from the Asia-Pacific. New York: Cam- ington, D.C.: Institute for International Economics. bridge University Press.Martin, W., and P. Urban. 1984. "Mod- Mann, Catherine, and Sarah Cleeland Knight. 2000. "Electronic eling Producer Response under Support Price and Stabilization Commerce in the World Trade Organization." In Jeffrey Schott, Schemes." Paper presented to the 28th Annual Conference of ed., The WTO after Seattle. Washington, D.C.: Institute for the Australian Agricultural Economics Society, Sydney. International Economics. Martin, Will, and L. Alan Winters, eds. 1996. The Uruguay Round Mann, Catherine L., Sue E. Eckert, and Sarah Cleeland Knight. and the Developing Economies. Cambridge, U.K.: Cambridge 2000. Global Electronic Commerce: A Policy Primer. Washington, University Press. D.C.: Institute for International Economics. Martin, Will, and K. Yagashima.1993. "Concerted Trade Liberaliza- Mansfield, Edwin. 1994. Intellectual Property Protection, Foreign tion and Economic Development in the Asia-Pacific Region." Direct Investment, and Technology Transfer. International Background paper for East Asia's Trade and Investment: Region- Finance Corporation Discussion Paper 19. Washington, D.C.: al and Global Gains from Liberalization (Washington, D.C.: World Bank. World Bank, 1994). World Bank, International Economics ------. 1995. Intellectual Property Protection, Direct Investment, Department, Washington, D.C. and Technology Transfer: Germany, Japan, and the United States. Martinez de Prera, Josefina. 2000. "Revenue-Neutral Tariff Reform: International Finance Corporation Discussion Paper 27. Wash- Welfare Effects of Uniform Tariffs in 13 Developing Countries." ington, D.C.: World Bank. Ph.D. dissertation, University of Colorado, Boulder. Available at Mansfield, E., and Anthony Romeo. 1980. "Technology Transfer to < http://ussub.colorado.edu/~martindp />. Overseas Subsidiaries by U.S. Based Firms." Quarterly Journal of Mashayeki, M. 2000. "Trade-Related Investment Measures." In Economics 95: 737­49. UNCTAD, Positive Agenda and Future Trade Negotiations. Mansfield, E., M. Schwartz, and S. Wagner, 1981. "Imitation Costs Geneva. and Patents: An Empirical Study." Economic Journal 91: Maskus, Keith. 1997. "Should Core Labor Standards Be Imposed 907­18. through International Trade Policy?" Policy Research Working Manual on Statistics of International Trade in Services. 2000. Pre- Paper 1817, World Bank, Washington, D.C. pared by the Commission of the European Union, the IMF, the ------. 1998. "The Role of Intellectual Property Rights in Encourag- OECD, the United Nations, UNCTAD, and the WTO. ing Foreign Direct Investment and Technology Transfer." Duke Marko, M. 1998. "An Evaluation of the Basic Telecommunications Journal of Comparative and International Law 9 (1): 109­61. Services Agreement." CIES Policy Discussion Paper 98/09. Cen- ------. 2000a. Intellectual Property Rights in the Global Economy. tre for International Economic Studies, University of Adelaide, Washington, D.C.: Institute for International Economics. Australia. ------. 2000b. "Regulatory Standards in the WTO: Comparing Markusen, James R. 1983. "Factor Movements and Commodity Intellectual Property Rights with Competition Policy, Environ- Trade as Complements." Journal of International Economics 13: mental Protection and Core Labor Standards." World Bank, 341­56. Washington, D.C. Available at ------. 1989. "Trade in Producer Services and in Other Specialized ------. 1995. "The Boundaries of Multinational Enterprises and ------. 2000c. "Strengthening Intellectual Property Rights in the Theory of International Trade." Journal of Economic Perspec- Lebanon." In Bernard Hoekman and Jamel Zarrouk, eds., tives 9: 169­89. Catching Up with the Competition: Trade Opportunities and Chal- ------. 2001. "Multilateral Rules on Foreign Direct Investment: lenges for Arab Countries, 251­84. Ann Arbor: University of The Developing Countries' Stake." Review of International Eco- Michigan Press. nomics. 9 (2): 287­302. ------. 2001. "Options for Raising Access to Essential Medicines: Markusen, James R., Thomas F. Rutherford, and Linda Hunter. Report to Working Group 4 of the Commission on Macroeco- 1995. "Trade Liberalization in a Multinational Dominated nomics and Health." World Health Organization, Geneva. Industry." Journal of International Economics 38: 95­17. Maskus, Keith E., and Yongmin Chen. 2000. "Vertical Price Con- Markusen, James, Thomas Rutherford, and David Tarr. 2000. "For- trol and Parallel Imports: Theory and Evidence." Policy eign Direct Investment in Services and the Domestic Market Research Working Paper Series 2461. World Bank, Washington, for Expertise." Policy Research Working Paper Series 2413. D.C. World Bank, Washington, D.C. Maskus, Keith E., and Mohamed Lahouel. 2000. "Competition Marsden, P. 2000. "The Divide on Verticals." In Simon J. Evenett, Policy and Intellectual Property Rights in Developing Coun- Alexander Lehmann, and Benn Steil, eds., Antitrust Goes Glob- tries." World Economy 23: 595­611. al: What Future for Transatlantic Cooperation? Washington, Maskus, Keith E., and Christine McDaniel. 1999. "Impacts of the D.C.: Brookings Institution Press. Japanese Patent System on Productivity Growth." Japan and Martin, W. 1996. "The Abolition of the Multi-Fiber Arrangement the World Economy 11: 557­74. and Its Implications for Fiber Markets." Prepared for the con- Maskus, Keith, and Mohan Penubarti. 1995. "How Trade-Related ference, "The WTO and the Uruguay Round Agreement: Impli- Are Intellectual Property Rights?" Journal of International Eco- cations for South Asian Agriculture," Katmandu, April. nomics 39: 227­48. 618 Bibliography Maskus, Keith E., and John S. Wilson. 2001. Quantifying the Impact and Tony Warren, eds., Impediments to Trade in Services: Mea- of Technical Standards to Trade: Can It Be Done? Ann Arbor: Uni- surement and Policy Implications. London and New York: Rout- versity of Michigan Press. ledge. Maskus, Keith E., Sean M. Dougherty, and Andrew Mertha. 1998. McGuire, Greg, Michael Schuele, and T. Smith. 2001. "Restrictive- "Intellectual Property Rights and Economic Development in ness of International Trade in Maritime Services." In Christo- China." Presented at the Southwest China Regional Confer- pher Findlay and Tony Warren, eds., Impediments to Trade in ence on Intellectual Property Rights and Economic Develop- Services: Measurement and Policy Implications. London and New ment, Chongqing, September 16. York: Routledge. Maskus, K. E., T. Otsuki, and J. S. Wilson. 2000. "Quantifying the McKibbin, Warwick, and P. J. Wilcoxen. 1996. "The Role of Ser- Impact of Technical Barriers to Trade: A Framework for Analy- vices in Modeling the Global Economy." Asia-Pacific Economic sis." Policy Research Working Paper Series 2512. World Bank, Review 2: 2­13. Washington, D.C. Mehta, P. 2000. "Taming Unilever in Bhutan." Consumer Unity & Masuyama, S., D. Vanderbrink, and S. Y. Chia. 1997. Industrial Trust Society (CUTS), Jaipur. Policies in East Asia. Nomura Institute and Institute of Southeast Mendelsohn, R., and M. J. Balick. 1995. "The Value of Undiscov- Asian Studies, Tokyo Club Foundation for Global Studies. ered Pharmaceuticals in Tropical Forests." Economic Botany 49 Mathis, J. 2000 "Towards a Positive Agenda for Multilateral Nego- (2): 223­28. tiations on Competition Policy: Interests of Developing Coun- Merges, R. P., and R. R. Nelson. 1990. "On the Complex Econom- tries." In UNCTAD, Positive Agenda and Future Trade ics of Patent Scope." Columbia Law Review 90 (4, May): Negotiations. Geneva. 839­916. Mattoo, Aaditya. 1999. "Financial Services and the WTO: Liberal- Messerlin, Patrick A. 2001. Measuring the Costs of Protection in ization Commitments of the Developing and Transition Europe: European Commercial Policy in the 2000s. Washington, Economies." Policy Research Working Paper Series 2184. World D.C.: Institute for International Economics. Bank, Washington, D.C. Messerlin, Patrick, and Karl Sauvant. 1990. The Uruguay Round: ------. 2000. "Developing Countries in the New Round of GATS Services in the World Economy. Washington, D.C.: World Bank. Negotiations: Towards a Pro-active Role." World Economy 23 Michaely, Michael. 1996. "Trade Preferential Agreements in Latin (4): 471­89. America: An Ex-Ante Assessment." Policy Research Working ------. Forthcoming. "Shaping Future Rules for Trade in Services: Paper Series 1583. World Bank, Washington, D.C. Lessons from the GATS." In Takatoshi Ito and Anne Krueger, Michaely, Michael, Demetris Papageorgiou, and Armeane M. ed., Trade in Services. Cambridge, Mass.: National Bureau of Choksi. 1991. Liberalizing Foreign Trade: Lessons of Experience in Economic Research. the Developing World, vol. 7. Oxford, U.K.: Blackwell. Mattoo, Aaditya, and Marcelo Olarreaga. 2001. "Reciprocity Michalopoulos, Constantine. 1998a. "Developing Countries' Par- across Modes of Supply in the World Trade Organization: A ticipation in the World Trade Organization." Policy Research Negotiating Formula." Policy Research Working Paper Series Working Paper Series 1906. World Bank, Washington, D.C. 2373. World Bank, Washington, D.C. ------. 1998b. "WTO Accession for Countries in Transition." Post- Mattoo, Aaditya, and Ludger Schuknecht. 2000. "Trade Policies Soviet Prospects 6 (3. June). for Electronic Commerce." Policy Research Working Paper ------. 1999a. "The Developing Countries in the WTO." World Series 2380. World Bank, Development Research Group, Wash- Economy 22 (1, January): 117­43. ington, D.C. ------. 1999b. "The Integration of Transition Economies into the Mattoo, Aditya, and Robert Stern. Forthcoming. India and the WTO. World Trading System." Policy Research Working Paper Series Mattoo, Aaditya, and Arvind Subramanian. 1998. "Regulatory 2182. World Bank, Washington, D.C. Autonomy and Multilateral Disciplines." Journal of International ------. 1999c. "Trade Policy and Market Access Issues for Devel- Economic Law 1: 303­22. oping Countries: Implications for the Millennium Round." Poli- Mazzoleni, Robert, and Richard R. Nelson. 1998. "The Benefits cy Research Working Paper Series 2214. World Bank, and Costs of Strong Patent Protection: A Contribution to the Washington, D.C. Current Debate." Research Policy 27: 273­84. ------. Developing Countries and the WTO. New York: Palgrave. McAfee, R. Preston, and John McMillan. 1989. "Government Pro- Michalopoulos, Constantine, and L. Alan Winters. 1997. "Summary curement and International Trade." Journal of International Eco- and Overview." In Peter D. Ehrenhaft and others, eds., Policies nomics 26: 291­308. on Imports from Economies in Transition. Studies of Economies in McCulloch, Neil, L. Alan Winters, and Xavier Cirera. 2001. Trade Transformation 22. Washington, D.C.: World Bank. Liberalization and Poverty: A Handbook. London: Centre for Eco- MITI (Ministry for Industry and Trade), Japan. 1999. 1999 Report nomic Policy Research. on the WTO Consistency of Trade Policies by Major Trading Part- McDougall, Robert, Aziz Elbehri, and Truong P. Truong. 1998. ners. Tokyo, Industrial Structure Council, MITI. "Global Trade, Assistance, and Protection: The GTAP 4 Data Mitra, Pradeep. 1992. "The Coordinated Reform of Tariffs and Base." Center for Global Trade Analysis, Purdue University, Indirect Taxes." World Bank Research Observer 7 (2): 195­218. West Lafayette, Ind. ------. 1994. "Protective and Revenue Raising Trade Taxes: Theo- McGuire, Greg. 1998. Australia's Restrictions on Trade in Financial ry and Application to India." Economic Studies Quarterly 45 (3): Services. Staff Research Paper. Productivity Commission, Can- 265­87. berra. Mochebelele, Motsamai. 1998. "Trade in Services: Potential and McGuire, Greg, and Michael Schuele. 2001. "Restrictiveness of Prospects for the Economy of Lesotho." Prepared for CAPAS International Trade in Banking Services." In Christopher Findlay regional meeting, Mauritius, December 7­9. Processed. 619 D E V E L O P M E N T, T R A D E , A N D T H E W T O Mody, Ashok. 1999. "Industrial Policy after the East Asian Crisis: The Promise and Problems of Trade Negotiations in Latin America. From Outward Orientation to New Internal Capabilities?" Poli- Macmillan Palgrave. cy Research Working Paper Series 2112. World Bank, Washing- Nash, John, and Wendy Takacs, eds. 1998. Trade Policy Reform: ton, D.C. Lessons and Implications. Washington D.C.: World Bank. Moenius, Johannes. 1999. "Information versus Product Adaptation: Naughton, Barry. 2001. "Problems of Lagging Regions in China." The Role of Standards in Trade." International Business & Mar- Background paper submitted to the World Bank Resident Mis- kets Research Center Working Paper. Kellogg School of Man- sion, Beijing. agement, Northwestern University, Evanston, Ill. Processed. Neary, J. Peter. 1997. "Pitfalls in the Theory of International Trade Mogens, Peter Carl. 2001. "Towards Basic Rules on Trade-Related Policy: Concertina Reforms of Tariffs and Subsidies to High Competition Policy." Speech by the Director General, DG Technology Industries." Scandinavian Journal of Economics 100 TRADE, European Commission, Brussels, March 2. (March): 187­206. Mohan, Rakesh. 2000. "Enhancing Competitiveness through Nelson, Richard R., and Howard Pack. 1999. "The Asian Miracle Industrial Restructuring: Options for Asia-Pacific Countries." and Modern Growth Theory." Economic Journal 109: 416­36. Presented at United Nations Industrial Development Organiza- Ng, Francis, and Sandy Yeats. 1997. "Open Economies Work Bet- tion Asia-Pacific Regional Forum on Industrial Development, ter! Did Africa's Protectionist Policies Cause Its Marginalization Shanghai, December 4­5. in World Trade?" World Development 25 (6): 889­904. Montaña Mora, Miguel. 1997. La OMC y el reforzamiento del sis- NIAAA (National Indigenous Arts Advocacy Association Inc.). tema GATT: Una nueva era en la solución de diferencias comer- 1997. Discussion Paper on the Proposed Label of Authenticity. ciales. Madrid: McGraw-Hill. Royal Exchange, NSW, Australia. Moran, Theodore. 1998. Foreign Direct Investment and Develop- Nicolaidis, K. 1997. "Managed Mutual Recognition: The New ment. Washington, D.C.: Institute for International Economics. Approach to the Liberalization of Professional Services," Work- Morkre, Morris, and David Tarr. 1980. Effects of United States ing Paper. John F. Kennedy School of Government, Harvard Restriction on Imports: Five Case Studies and Theory. Washing- University, Cambridge, Mass. ton, D.C.: Government Printing Office. Niggemann, H. 2000. "Die volkswirtschaftliche Bedeutung von Morrissey, Oliver, and Yogesh Rai. 1995. "The GATT Agreement Dienstleistungen und der Beitrag des Mannheimer Innovation- on Trade-Related Investment Measures: Implications for Devel- spanels und der ZEW/Creditreform Konjunkturumfrage zu ihrer oping Countries and Their Relationship with Transnational Cor- Darstellung." In Statistisches Bundesamt, Wirtschaft und Statis- porations." Journal of Development Studies 31 (5): 702­24. tik 7/2000. Mosley, Paul. 1999. "Micro-Macro Linkages in Financial Markets: Nivola, Pietro, ed. 1997. Comparative Disadvantages? Social Regu- The Impact of Financial Liberalization on Access to Rural Credit lations and the Global Economy. Washington, D.C.: Brookings in Four African Countries." Journal of International Development Institution Press. 11: 367­84. Nogués, Julio J., Andrzej Olechowski, and L. Alan Winters. 1986. ------. 2000 "Globalisation, Economic Policy and Convergence." "The Extent of Nontariff Barriers to Industrial Countries' World Economy 23 (5, May): 613­34. Imports." World Bank Economic Review 1 (1): 181­99. Mukherjee, N. 1996. "Exporting Labour Services and Market Noland, Marcus. 1997. "Has Asian Export Performance Been Access Commitments under GATS in the WTO." Journal of Unique?" Journal of International Economics 43: 79­101. World Trade 30 (5): 21­42. Norchi, Charles H. 2000. "Indigenous Knowledge as Intellectual ------. 1998a. "General Agreement on Trade in Services and Property." Policy Sciences 33: 387­98. India's Exports." New Delhi: Indian Council for Research on Nordhaus, W. D. 1969. Invention, Growth and Welfare. Cambridge, International Economic Relations (ICRIER). Mass.: MIT Press. ------. 1998b. "Non-Tariff Barriers and Trade in Services: A Com- North, Douglass C., and R. Thomas. 1973. The Rise of the Western parative Assessment of Capital and Labour Mobility in the World: A New Economic History. Cambridge, U.K.: Cambridge GATS under the WTO." World Competition, Law and Economics University Press. Review 21 (5): 79­91. North, Douglass C., and Barry Weingast. 1989. "Constitutions and Mundell, R. A. 1957. "International Trade and Factor Mobility." Commitment: The Evolution of Institutions Governing Public American Economic Review 47: 321­35. Choice in Seventeenth Century England." Journal of Economic Munk, Bernard. 1969. "The Welfare Costs of Content Protection: History 49: 803­32. The Automotive Industry in Latin America." Journal of Political Noyelle, T. 1995. "Barriers to Trade in Labour Services: Proposals Economy 77 (1, January­February): 85­98. for Strengthening GATS Commitments on the Temporary Mwega, Francis, and Kefa Muga. 2000. "Africa and the World Movement of Natural Persons from the Perspective of Devel- Trading System: A Case Study of Kenya." African Economic oping Countries." May. UNCTAD, New York. Research Consortium, Nairobi. Processed. Noyelle, T., K. Blocke, D. Chandra, and B. Redfield. 1991. "Trade Najam, Adil, and Nick Robins. 2001. "Seizing the Future: The in Labour Services and Temporary Movement of Personnel: South, Sustainable Development and International Trade." Elements for a Request on Labour Mobility by the Developing International Affairs 77 (1): 93­111. Countries to the Group of Negotiations on Services." Working Narlikar, Amrita. 2000. "Bargaining Together in Trade: Developing Paper 91-08. Eisenhower Center for the Conservation of Countries in Coalitions, 1982­98." Ph.D. thesis. University of Human Resources, Columbia University, New York. Oxford, Oxford, U.K. O'Connor, Bernard, and Paolo Vergano. 2000. "Procedures for EU- Narlikar, Amrita, and Ngaire Woods. 2002. "Sectoral Coalitions: US Trade Dispute on FSCs Sparks Debate on DSU Interpreta- The Case of Services." In Diana Tussie, ed., On Shifting Ground: tion Issues." World Trade Agenda 00/17 (September). Geneva. 620 Bibliography OECD. 1989. Competition Policy and Intellectual Property Rights. Oyejide, T. A. 1997. "Africa's Participation in the Post-Uruguay Paris. Round." PSIO Program for the Study of International Organiza- ------. 1992. Integration of the Developing Countries into the Inter- tions. Occasional Paper WTO Series 6. Geneva. national Trading System. Paris. ------. 2000. "Interests and Options of Developing and Least- ------. 1995. Regional Integration and the Multilateral Trading Sys- Developed Countries in a New Round of Multilateral Trade tem: Synergy and Divergence. Paris. Negotiations." G-24 Discussion Paper Series. May. United ------. 1996. Regionalism and Its Place in the Multilateral Trading Nations, New York.Oyejide, Ademola, and Dominique Njinkeu. System. Paris. 2000. "African Countries' Proposals and Objectives in the Post- ------. 1997a. "Assessing Barriers to Trade in Services: A Pilot Seattle Framework of WTO Trade Negotiations." Prepared for Study on Accountancy Services." TD/TC/WP(97)26. Working the workshop on Trade Negotiation Issues in the Cotonou Party of the Trade Committee. Paris. Partnership Agreement: Agriculture and Economic Partnership ------. 1997b. Indicators of Tariff and Non-Tariff Barriers. Paris. Agreements, Brussels, November 21­22. African Economic ------. 1997c. "Market Access for Least-Developed Countries." Research Consortium, Nairobi. Processed. WTO, WT/LDC/HL:19. Paris. Ozawa, Terutomo. 1974. Japan's Technological Challenge to the ------. 1997d. The OECD Report on Regulatory Reform. Vols. 1 and West, 1950­1974: Motivation and Accomplishment. Cambridge, 2. Paris. Mass.: MIT Press. ------. 1997e. The World in 2020: Towards a New Global Age. Ozdem, C., and O. Demirkol. 1994. "The Implications of the WTO Paris. Uruguay Round on Turkish Economy." Arab Experts Meeting ------. 1999a. "Aide-Memoire of the Discussion at the Confer- on WTO Impacts, Analysis on Arab Economies, League of Arab ence on Trade and Competition (29 and 30 June 1999)." Avail- States, Cairo, July. Processed. able at . opment: A Case Study in Textiles. A World Bank Research Publi- ------. 1999b. "An Assessment of the Costs for International cation. New York: Oxford University Press. Trade in Meeting Regulatory Requirements." Pack, Howard, and Kamal Saggi. 1997. "Inflows of Foreign Tech- TD/TC/WP(99)8/FINAL. Paris. nology and Indigenous Technological Development." Review of ------. 1999c. "Competition Elements in International Trade Development Economics 1 (1): 81­98. Agreements." Available at ------. 1999. "Vertical Technology Transfer, Diffusion, and Com- . Bank, Washington, D.C. ------.1999d. Trade, Investment and Development: Policy Coher- Palim, Mark R. A. 1998. "The Worldwide Growth of Competition ence Matters. Paris. Law: An Empirical Analysis." Antitrust Bulletin 105: 111­15. ------. 2000. Tariffs and Trade: OECD Query and Reporting System. Palmeter, David. 1993a. "Pacific Regional Trade Liberalization and CD-ROM. Paris. the Rules of Origin." Journal of World Trade (October): 49­62. ------. 2001. "Government Procurement in Member and Non- ------. 1993b. "Rules of Origin in Customs Unions and Free Member Countries." Paris. Processed. Trade Areas." In Kym Anderson and Richard Blackhurst, eds., Ogunkola, E. O. 1999. "African Capacity for Compliance and Regional Integration and the Global Trading System, 326­43. Defense of WTO Rights." Prepared for conference, "Africa and New York: St. Martin's Press. the World Trading System," African Economic Research Con- ------. 1998. "The WTO Antidumping Agreement and the sortium, Yaoundé, April 17­18. Economies in Transition." In Thomas Cottier and Petros C. Ohiorhenuan, John F. E. 1998. "Capacity Building Implications of Mavroidis, eds., State Trading in the Twenty-first Century. Ann Enhanced African Participation in Global Trade Rules-Making Arbor: University of Michigan Press. and Arrangements." Prepared for the AERC collaborative Palmeter, David, and Petros C. Mavroidis. 1999. Dispute Settle- research project, "Africa and the World Trading System." ment in the World Trade Organization: Practice and Procedure. African Economic Research Consortium, Nairobi. Processed. The Hague: Kluver Law International. Olarreaga, M., and R. Rocha, eds. 2000. "La nueva agenda del Panagariya, Arvind. 1992. "Import Tariffs, Duty Drawbacks and comercio en la OMC." World Bank Institute and Universidad Tariff Reforms." Journal of International Economics 32: 131­47. del Rosario, Bogotá. Panagariya, Arvind, and Dani Rodrik. 1993. "Political-Economy Ostry, Sylvia. 2000. "The Uruguay Round North-South Grand Bar- Arguments for a Uniform Tariff." International Economic Review gain: Implications for Future Negotiations." Prepared for con- 34 (3, August): 685­703. ference on the Political Economy of International Trade Law, Parente, Stephen L., and Edward C. Prescott. 1994. "Barriers to University of Minnesota, September. Technology Adoption and Development." Journal of Political Otsuki, T., and J. S. Wilson. 2001. "Global Food Safety Regulation Economy 102: 298­321. and Trade: An Empirical Analysis." Work in progress. World Park, Walter. 2000. "International Patenting, Patent Rights and Bank, Washington, D.C. Technology Gaps." March. American University, Washington, Otsuki, T., M. Sewadeh, and J. S. Wilson. 2000. "Saving Two in a D.C. Processed. Billion: A Case Study to Quantify the Trade Effect of European Park, Walter G., and Carlos Ginarte. 1997. "Intellectual Property Food Safety Standards on African Exports." Work in progress. Rights and Economic Growth." Contemporary Economic Policy World Bank, Washington, D.C. 15: 51­61. Oum, Tae Hoon, and Chunyan Yu. 1998. Winning Airlines. Boston: Pauwelyn, Joost. 2000. "Enforcement and Countermeasures in Kluwer Academic. the WTO: Rules Are Rules--Toward a More Collective 621 D E V E L O P M E N T, T R A D E , A N D T H E W T O Approach." American Journal of International Law 94 (2, April): Jeffrey A. McNeely, eds., Protection of Global Biodiversity: Converg- 335­47. ing Strategies. Durham, N.C.: Duke University Press. PECC (Pacific Economic Cooperation Council). 1995. Survey of Imped- Rege, Vinod. 1999. "Developing Country Participation in Negotia- iments to Trade and Investment in the APEC Region. Singapore. tions Leading to the Adoption of the WTO Agreement on Cus- Petri, Peter A. 1997. "Foreign Direct Investment in a Computable toms Valuation and Preshipment Inspection: A Public Choice General Equilibrium Framework." Prepared for the conference Analysis." World Competition 22 (1, March): 37­117. "Making APEC Work: Economic Challenges and Policy Alterna- Reichman, Jerome H. 1994. "Legal Hybrids between the Patent tives," Keio University, Tokyo, March 13­14. and Copyright Paradigms." Columbia Law Review 94: Petrazzini, B. A., and P. Lovelock. 1996. "Telecommunications in 2432­558. the Region: Comparative Case Studies." Presented at the Inter- ------. 1998. "Securing Compliance with the TRIPS Agreement national Institute for Communication Telecommunications after US v India." Journal of International Economic Law 1 (4, Forum, Sydney, Australia, April 22­23. December): 603­06. Phillips, David A. 2001. "Implementing the Market Approach to ------. 2000. "Of Green Tulips and Legal Kudzu: Repackaging Enterprise Support: An Evaluation of Ten Matching Grant Rights in Subpatentable Invention." Vanderbilt Law Review 53: Schemes." World Bank, Washington, D.C. Processed. 1743­98. Pinheiro A. 2000. "The Brazilian Privatisation Experience: What's ------. 2001. "IPR as Related to Crafts in India: Compensatory Next?" Prepared for the Global Development Network Confer- Liability Legislation." Presented to the Crafts Workshop, India, ence, Tokyo, December 10­13. World Bank, Washington, D.C., January 9. Pistor, Katharina. 2000. "The Standardization of Law and Its Effect Reichman, J. H., and Pamela Samuelson. 1997. "Intellectual Prop- on Developing Economies." Group of 24 Discussion Paper 4. erty Rights in Data?" Vanderbilt Law Review 50: 51­166. Available at . al Institutional Measures." In Lakshman D. Guruswamy and Jef- Poapongsakorn, N. 2000. "The Thai Automotive Industry." Pre- frey A. McNeely, eds., Protection of Global Biodiversity: pared for the PECC-ASEAN Auto Project, School of Advanced Converging Strategies. Durham, N.C.: Duke University Press. Management Technology, Manila. Reinbothe, Jörg, Maria Martin-Prat, and Silke Von Lewinsky. 1997. Porges, A. 2000. "The Banana War: Whose Market Access?" World "The New WIPO Treaties: A First Résumé." European Intellectual Bank, Washington, D.C. Processed. Property Review 4: 171­76. Preeg, E. 1994. "Traders in a Brave New World: The Uruguay Revesz, R. 1992. "Rehabilitating Interstate Competition: Rethink- Round and the Future of the International Trading System." ing the `Race to the Bottom' Rationale for Federal Environmen- Center for Strategic and International Studies, Washington, tal Regulation." New York University Law Review 67: 1210­54. D.C. Processed. Rhee, Yung Whee, Bruce Ross-Larson, and Garry Pursell. 1984. Productivity Commission. 1998. Inquiry into the Automotive Indus- Korea's Competitive Edge: Managing Entry into World Markets. try. Canberra: Australian Government Publishing Service. Baltimore, Md.: Johns Hopkins University Press. Pursell, Garry. 1999. "Export Policies and Institutions in Develop- Rhodes, Sylvia A. 2000. "The Article 21.5/22 Problem: Clarifica- ing Countries: The Role of the World Bank." World Bank, tion through Bilateral Agreements?" Journal of International Washington, D.C. Processed. Economic Law 3 (3, September): 553­58. ------. 2001. "The Australian Experience with FDI and Local Con- Richardson, J. David. 1995. "Income Inequality and Trade: How to tent Programs in the Auto Industry." Journal of World Trade [..]. Link, What to Conclude." Journal of Economic Perspectives (9, Rajapathirana, S., C. Lusthaus, and M. Adrien. 2000. "Review of Summer): 33­56. the Integrated Framework for Technical Assistance for Trade Richardson, Martin. 1991. "The Effects of a Content Requirement Development of Least Developed Countries." World Bank, on a Foreign Duopsonist." Journal of International Economics Washington, D.C. Processed. 31: 143­55. Ramachandran, Vijaya. 1993. "Technology Transfer, Firm Owner- ------. "Content Protection with Foreign Capital." Oxford Eco- ship, and Investment in Human Capital." Review of Economics nomic Papers 45: 103­17. and Statistics 75: 664­70. Ricupero, Rubens. 1998. "Integration of Developing Countries Ramos, Raul, and Juan Rosellón. 1991. "La economía elemental de into the Multilateral Trading System." In Jagdish Bhagwati and las reglas de origen." El Trimestre Económico 481­96. Matthias Hirsch, eds., The Uruguay Round and Beyond: Essays in Ramsey, F. P. 1927. "A Contribution to the Theory of Taxation." Honor of Arthur Dunkel. Ann Arbor: University of Michigan Economic Journal 37: 47­61. Press. Rauch, J. E. 1999. "Networks versus Markets in International ------. "A Development Round: Converting Rhetoric into Sub- Trade." Journal of International Economics 48: 7­35. stance." Presented at John F. Kennedy School of Government Ravallion, Martin. 1999. "Protecting the Poor in Crisis." PREM symposium, "Efficiency, Equity and Legitimacy: The Multilater- Note 12. World Bank, Washington, D.C. al Trading System at the Millennium," Harvard University, Ravallion, Martin, and Dominique van de Walle. 1991. "The Cambridge, Mass.,June 1­2. Impact on Poverty of Food Pricing Reforms: A Welfare Analysis Rivera-Batiz, L., and P. Romer. 1991. "International Integration for Indonesia." Journal of Policy Modeling 13 (2): 281­99. and Endogenous Growth." Quarterly Journal of Economics 106: Raven, John. 2000. Trade and Transport Facilitation: An Audit 531­56. Methodology. Washington, D.C.: World Bank. Robinson, Sherman, Zhi Wang, and Will Martin. 2002. "Capturing Raven, Peter H., and Jeffrey A. McNeely. 1998. "Biological Extinction: the Implications of Services Trade Liberalization." Economic Sys- Its Scope and Meaning for Us." In Lakshman D. Guruswamy and tem Research 14 (March). 622 Bibliography Rodriguez, Francisco, and Dani Rodrik. 2001. "Trade Policy and Satapathy, C. 1994. "Valuation of Goods Traded Internationally, Economic Growth: A Skeptic's Guide to the Cross-National Evi- with Special Reference to Implementation of GATT Agreement dence." In Ben S. Bernanke and Kenneth Rogoff, eds., NBER in Developing Countries." University of Bath. Macroeconomics Annual 2000. Cambridge, Mass.: MIT Press. ------. 2000. "Implementation of WTO Agreement on Customs Rodriguez-Clare, A. 1996. "The Division of Labour and Economic Valuation." Economic and Political Weekly (Mumbai, India) 35 Development." Journal of Development Economics 49: 3­32. (25, June 17­23). Rodrik, Dani. 1987. "The Economics of Export-Performance Sauvé, Pierre, and James Gillespie. 2000. "Financial Services and Requirements." Quarterly Journal of Economics 102: 633­50. the GATS 2000 Round." In Robert E. Litan and Anthony M. ------. 1996. "Labor Standards in International Trade: Do They Santomero, eds., Brookings-Wharton Papers on Financial Services: Matter, and What Do We Do about Them?" Overseas Develop- 2000, 412­52. Washington, D.C.: Brookings Institution Press. ment Council, Washington, D.C. Processed. Sauvé, Pierre, and Brenda Gonzalez-Hermosillo, 1993. "Implications ------. 1997. Has Globalization Gone Too Far? Washington, D.C.: of the NAFTA for Canadian Financial Institutions." C. D. Howe Institute for International Economics. Commentary, no. 44 (April). Toronto: C. D. Howe Institute. ------. 1998. "Why Do More Open Economies Have Bigger Gov- Sauvé, Pierre, and Robert M. Stern. 2000. GATS 2000: New Direc- ernments?" Journal of Political Economy 106 (October): tions in Services Trade Liberalization. Washington, D.C.: Brook- 997­1032. ings Institution Press and Harvard University. ------. 1999. Making Openness Work: The New Global Economy Sauvé, Pierre, and Christopher Wilkie. 2000. "Investment Liberal- and the Developing Countries. Washington, D.C.: Overseas ization in GATS." In Pierre Sauvé and Robert M. Stern, eds., Development Council. GATS 2000: New Directions in Services Trade Liberalization, ------. 2000a. "Can Integration into the World Economy Substi- 331­63. Washington, D.C.: Brookings Institution Press and tute for a Development Strategy?" June. Harvard University, Harvard University. Cambridge, Mass. Processed. Scherer, F. M. 1977. The Economic Effects of Compulsory Patent ------. 2000b. "Institutions for High-Quality Growth: What They Licensing. Monograph Series in Finance and Economics. New Are and How to Acquire Them." Studies in Comparative Inter- York: New York University. national Development 35 (3, fall): 3­31. ------. 1994. Competition Policies for an Integrated World Economy. Rollo, Jim, and L. Alan Winters. 2000. "Subsidiarity and Gover- Washington, D.C.: Brookings Institution Press. nance Challenges for the WTO: Environmental and Labour ------. 2000. Competition Policy: Domestic and International. Chel- Standards." World Economy 23 (4): 561­76. tenham, U.K.: Edward Elgar. Romer, David. 1993. "Openness and Inflation: Theory and Evi- Scherer, F. M., and J. Watal. 2001. "Post-TRIPS Options for Access dence." Quarterly Journal of Economics 108 (4, November): to Patented Medicines in Developing Countries." Harvard Uni- 870­903. versity, Cambridge, Mass. Processed. Rose, Nancy L. 1985. "The Incidence of Regulatory Rents in the Schiff, Maurice, and Alberto Valdés. 1992. The Plundering of Agri- Motor Carrier Industry." RAND Journal of Economics 16 (3, culture in Developing Countries. Washington, D.C.: World autumn): 299­318. Bank. Rose-Ackerman, Susan. 1999. Corruption and Government. Cam- Schive, Chi. 1990. The Foreign Factor: The Multinational Corpora- bridge, U.K.: Cambridge University Press. tions' Contribution to the Economic Modernization of the Republic Sachs, Jeffrey. 2001. "A New Paradigm for Development." In of China. Stanford, Calif.: Hoover Institution Press. Roger B. Porter, Pierre Sauvé, Arvind Subramanian, and Ameri- Schott, Jeffrey J., and Johanna W. Buurman. 1994. The Uruguay co Beviglia-Zampetti, eds., Efficiency, Equity, and Legitimacy: The Round: An Assessment. Washington, D.C.: Institute for Interna- Multilateral Trading System at the Millennium. Washington, tional Economics. D.C.: Brookings Institution Press and Harvard University. Schott, J. J., and J. Watal. 2000. "Decision-making in the WTO." Sachs, Jeffrey, and Felipe Larraín. 1999. "Why Dollarization Is International Economics Policy Briefs 00-2 (March). More Straitjacket Than Salvation." Foreign Policy (fall): 80­92. Scollay, R. 1996. "The Closer Economic Relations Agreement." In Sachs, Jeffrey, and Andrew Warner. 1995. "Economic Reform and B. Bora and C. Findlay, eds., Regional Integration and the Asia the Process of Global Integration." Brookings Papers on Eco- Pacific. Melbourne: Oxford University Press. nomic Activity 1: 1­118. Scotchmer, Suzanne. 1991. "Standing on the Shoulders of Giants: Safadi, R., and S. Laird. 1996. "The Uruguay Round Agreements: Cumulative Research and Patent Law." Journal of Economic Per- Impact on Developing Countries." World Development 24 (7): spectives 5 (1, winter): 29­41. 1223­42. Sedjo, Roger A. 1992. "Property Rights, Genetic Resources, and Sampson, G. 1987. "Safeguards." In J. Michael Finger and Andrzej Biotechnological Change." Journal of Law and Economics 35 Olechowski, eds., The Uruguay Round: A Handbook for the Mul- (April): 199­213. tilateral Negotiations. Washington, D.C.: World Bank. Sharer, Robert, and others. 1998. "Trade Liberalization in IMF- ------. 1999. "Trade, Environment and the WTO." Overseas Supported Programs." World Economic and Financial Surveys. Development Council, Washington, D.C. Washington, D.C.: IMF. Sapir, André. 1997. "The Political Economy of EC Regionalism." Shatz, Howard J., and David G. Tarr. 2000. "Exchange Rate Over- European Economic Review 42: 717­32. valuation and Trade Protection: Lessons from Experience." Pol- ------. 1998. "GATS 1994­2000." Journal of World Trade 33: 51­66. icy Research Working Paper Series 2289. World Bank, Sartar, S. 2000. "Privatisation and Beyond: The Indian Experience in Washington, D.C. Infrastructure Services." Prepared for the Global Development Shiells, Clint, Robert Stern, and A. Deardorff. 1986. "Estimates of Network Conference, Tokyo, December 10­13. Processed. the Elasticities of Substitutions between Imports and Home 623 D E V E L O P M E N T, T R A D E , A N D T H E W T O Goods for the United States." Weltwirtschaftliches Archiv 122: Stern, Robert M., Jonathan Francis, and Bruce Schumacher. 1976. 497­519. Price Elasticities in International Trade: An Annotated Bibliogra- Simpson, John P. 1994. "North American Free Trade Agreement: phy. London: Macmillan for the Trade Policy Research Centre. Rules of Origin." Journal of World Trade (February): 33­41. Stevens, Christopher, and Jane Kennan. 2000. "Analysis of EU Simpson, R. David, Roger A. Sedjo, and John W. Reid. 1996. "Valu- Trade Arrangements with Developing and Transition ing Biodiversity for Use in Pharmaceutical Research." Journal of Economies." Institute of Development Studies, University of Political Economy 104 (1): 163­84. Sussex, Brighton, U.K. Singh, A. 1996. "The Post-Uruguay Round World Trading System, Stewart, Richard B. 1992. "Environmental Regulation and Interna- Industrialization, Trade and Development: Implications for the tional Competition." Yale Law Journal 102 (8): 2039­2106. Asia-Pacific Developing Countries." In UNCTAD, Expansion of Stiglitz, Joseph E. 1998. "Towards a New Paradigm for Develop- Trading Opportunities to the Year 2000 for Asia-Pacific Develop- ment: Strategies, Policies and Processes." Raúl Prebisch Lec- ing Countries: Implications of the Uruguay Round and Adaptation ture, October 19, UNCTAD, Geneva. of Export Strategies, 147­90. Geneva. Stone, Christopher D. 1998. "What to Do about Biodiversity: The Singh, A., and R. Dhumale. 1999. "Competition Policy, Develop- Earth's Biological Resources in Law and Economics." In Laksh- ment and Developing Countries." South Centre, Global Biodiversity: Converging Strategies. Durham, N.C.: Duke Sittenfield, Ana M., and Annie Lovejoy. 1998. "Biodiversity University Press. Prospecting Frameworks: The INBio Experience in Costa Rica." Subramanian, Arvind. 1992. "Genetic Resources, Biodiversity and In Lakshman D. Guruswamy and Jeffrey A. McNeely, eds., Pro- Environmental Protection: An Analysis and Proposal towards a tection of Global Biodiversity: Converging Strategies. Durham, Solution." Journal of World Trade 26 (5): 105­10. N.C.: Duke University Press. Subramanian, Arvind, and Jayashree Watal. 2000. "Can TRIPS Smarzynska, Beata K. 1999. "Technological Leadership and the Serve as an Enforcement Device for Developing Countries in Choice of Entry Mode by Foreign Investors: Evidence from Tran- the WTO?" Journal of International Economic Law 3 (3): sition Economies." World Bank, Washington, D.C. Processed. 403­16. Smith, Adam. 1976 [1776]. An Inquiry into the Nature and Causes Subramanian, Arvind, Ali Ibrahim, and Luis A. Torres-Castro. 1993. of the Wealth of Nations. Chicago: University of Chicago Press. "Optimal Tariffs: Theory and Practice." IMF Working Paper. Smith, P. 1995. "End of the Line for the Local Loop Monopoly." June. Washington, D.C. Public Policy for the Private Sector Note 63. December. World Summers, Robert, and Alan Heston. 1991. "The Penn World Table Bank Group, Finance, Private Sector, and Infrastructure Net- (Mark 5): An Expanded Set of International Comparisons, work, Washington, D.C. 1950­1988." Quarterly Journal of Economics 106 (2, May): Snape, Richard H., Lisa Gropp, and Tas Luttrell. 1998. Australian 327­68. Trade Policy 1965­1997. A Documentary History. St. Leonards, Swann, P., P. Temple, and M. Shurmer. 1996. "Standards and NSW: Allen and Unwin. Trade Performance: The UK Experience." Economic Journal 106: Srinivasan, T. N. 1998. Developing Countries and the Multilateral 1297­313. Trading System: From GATT to the Uruguay Round and the Sykes, Alan. 1995. Product Standards for Internationally Integrated Future. New York: HarperCollins. Goods Markets. Washington, D.C.: Brookings Institution Press. ------. 2001. "Policy Issues in International and Regional Trade: ------. 1998. "Antidumping and Antitrust: What Problems Does An Overview." Department of Economics, Yale University. Each Address?" In Robert Z. Lawrence, ed., Brookings Trade Processed. Forum: 1998, 1­43. Washington, D.C.: Brookings University Srinivasan, T. N., and Jagdish Bhagwati. 1999. "Outward-Orienta- Press. tion and Development: Are Revisionists Right?" Discussion Takacs, Wendy E. 1994. "Domestic Content and Compensatory Paper 806. September. Economic Growth Center, Yale Univer- Export Requirements: Protection of the Motor Vehicle Industry sity, New Haven, Conn. in the Philippines." World Bank Economic Review 8 (1, January): Srivastava, Vivek. 2000. "India's Accession to the Government Pro- 127­49. curement Agreement: Identifying Costs and Benefits." Avail- Tamms, V. 2001. "Frontier Cost Estimates of the Impact of Restric- able at . Processed. Tony Warren, eds., Impediments to Trade in Services: Measure- Stahl, D. O., and A. H. Turunen-Red. 1995. "Tariff Games: Cooper- ment and Policy Implications. London and New York: Routledge. ation with Random Variation in Political Regimes." European Tanzi, Vito. 1989. "The Impact of Macroeconomic Policies on the Journal of Political Economy 11: 215­38. Level of Taxation and the Fiscal Balance in Developing Coun- Stegemann, Klaus. 1992. Price Undertakings to Settle Antidumping tries." IMF Staff Papers 36 (September): 633­56. Cases. Ottawa: Institute for Research on Public Policy. Tanzi, Vito, and Ludger Schuknecht. 2000. Public Spending in the Steiner, Henry J., and Philip Alston. 1996. International Human 20th Century: A Global Perspective. Cambridge, U.K.: Cam- Rights in Context. Oxford: Clarendon Press. bridge University Press. Stephenson, Sherry M. 2000. "GATS and Regional Integration." In Tarr, David G. 1990. "Second-Best Foreign Exchange Policy in the Pierre Sauvé and Robert M. Stern, eds., GATS 2000: New Direc- Presence of Domestic Price Controls and Export Subsidies." tions in Services Trade Liberalization. Washington, D.C.: Brook- World Bank Economic Review 4 (2, May): 175­93. ings Institution Press and Harvard University. ------. 1998. "On the Design of Tariff Policy for Russia." In Harry Stern, Robert, ed. 2001. Services in the International Economy. Ann Broadman, ed., Russian Trade Policy: Reform for WTO Accession. Arbor: University of Michigan Press. Washington, D.C.: World Bank. 624 Bibliography Tarullo, Daniel K. 2000. "Norms and Institutions in Global Compe- Tyson, Laura D'Andrea. 1992. Who's Bashing Whom: Trade Conflict tition Policy." American Journal of International Law 94: in High-Technology Industries. Washington, D.C.: Institute for 478­504. International Economics. Tavares, Jose De Araujo, Jr. 2001. "Competition Policy." In Jose UNCTAD. 1992. World Investment Report: Transnational Corpora- Manuel Salazar-Xirinachs and Maryse Robert, eds., Toward Free tions as Engines of Growth. New York. Trade in the Americas. Washington, D.C.: Brookings Institution ------. 1995a. "In-Depth Study on UNCTAD's ASYCUDA Pro- Press. grams." December 4. Geneva. Taylor, M. S. 1999. "Trade and Trade Policy in Endogenous ------. 1995b. Technological Capacity-Building and Technology Growth Models." In J. Piggott and A. Woodland, eds., Interna- Partnership: Field Findings, Country Experiences and Programmes. tional Trade Policy and the Pacific Rim, ch. 15. London: Macmil- Geneva. lan for the IAE (International Economic Association). ------. 1995c. World Investment Report 1995: Transnational Corpo- Taylor, R., and others. 1984. "EC and EFTA in the 1980s." Brussels: rations and Competitiveness. UN sales no. E.95.II.A.9. New York European Research Associates. and Geneva. Teece, David J. 1976. The Multinational Corporation and Resource ------. 1996. The TRIPS Agreement and Developing Countries. UN Cost of International Technology Transfer. Cambridge, Mass.: sales no. 96.II.D.10. Geneva. Ballinger. ------. 1997. World Investment Report: Transnational Corporations, ------. 1977. "Technology Transfer by Multinational Firms: The Market Structure, and Competition Policy. New York and Geneva. Resource Cost of Transferring Technological Know-How." Eco- ------. 1998a. "Coordinated African Programme for Assistance nomic Journal 87: 242­61. on Services (CAPAS): Phase III Synthesis." CAPAS Paper. ------. 1986. The Multinational Corporation and the Resource Cost of December. Geneva. International Technology Transfer. Cambridge, Mass.: Ballinger. ------. 1998b. World Investment Report: Trends and Determinants. Templeman, L. S. 1998. "Intellectual Property." Journal of Interna- New York and Geneva. tional Economic Law 1 (4, December): 585­602. ------. 1999a. "Preparing for Future Multilateral Trade Negotia- Tharakan, P. K. M. 1995. "The Political Economy of Contingent tions: Issues and Research Needs from a Development Perspec- Protection." Economic Journal 105 (433): 1550­64. tive." UN sales no. E.99.II.D.17. Geneva. Thomas, Vinod, John Nash, and associates. 1991. Best Practices in ------. 1999b. TRAINS Database. Geneva. Trade Policy Reform. New York: Oxford University Press. ------. 1999c. World Investment Report 1999: FDI and the Chal- Thompson, R. 2000. "Formula Approaches to Improving GATS lenge of Development. UN sales no. E.99.II.D.3. New York and Commitments." In Pierre Sauvé and Robert M. Stern, eds., Geneva. GATS 2000: New Directions in Services Trade Liberalization, ------. 2000a. "Building Confidence: Electronic Commerce and 473­86. Washington, D.C.: Brookings Institution Press and Development." UNCTAD/SDTE/MISC.11, 2000. New York. Harvard University. ------. 2000b. Positive Agenda and Future Trade Negotiations. New Transparency International. 1997. The TI Source Book: Applying the York. Framework. Available at . Geneva. Trebilcock, Michael, and Robert Howse. 1998. The Regulation of ------. 2001a. Duty and Quota Free Market Access for LDCs: An International Trade. London: Routledge. Analysis of QUAD Initiatives. Geneva. Trujillo, Lourdes, and Antonio Estache. 2001. "Surfing a Wave of ------. 2001b. World Investment Report 2001. New York and Fine Tuning Reforms in Argentina's Ports." World Bank, Wash- Geneva. ington, D.C. Processed. UNCTAD and WHO (World Health Organization). 1998. Interna- Trujillo, Lourdes, and Gustavo Nombela. 1999. "Privatization and tional Trade in Health Services: A Development Perspective. Gene- Regulation of the Seaport Industry." Policy Research Working va: UNCTAD. Paper Series 2181. World Bank, Washington, D.C. UNCTAD and World Bank. 1994. Liberalizing International Transac- Tumlir, J. 1983. "Strong and Weak Elements in the Concept of tions in Services: A Handbook. Geneva: United Nations. European Integration." In F. Machlup, G. Fels, and H. Müller- UNIDO (United Nations Industrial Development Organization). Groeling, eds., Reflections on a Troubled World Economy. Lon- Various years. International Yearbook of Industrial Statistics Series don: Macmillan. (CD-ROM version). Vienna. Tussie, Diana 1987. The Less Developed Countries and the World United Nations. 1975. Standard Industrial Trade Classification Sys- Trading System. A Challenge to the GATT. London: Pinter. tem, Revision 2. Statistical Papers Series M, no. 34/Rev.2. New Tussie, Diana, and David Glover. 1993. "Developing Countries in York. World Trade: Implications for Bargaining." In Diana Tussie and Unter, Brian D. 1998. "Maximizing Customer Benefits: A Global David Glover, eds., The Developing Countries in World Trade: Model for Regulatory Reform." Prepared for the U.S.-China Stan- Policies and Bargaining Strategies. Boulder, Colo.: Lynne Rien- dards, Testing, and Certification Workshop, Information Technol- ner. ogy Industry Council, Washington, D.C., February 17­18. Tybout, James. 2000. "Manufacturing Firms in Developing Coun- U.S. Bureau of the Census. 1975. Historical Statistics of the United tries: How Well Do They Do, and Why?" Journal of Economic States: Colonial Times to 1970. Part I, Bicentennial Edition. Wash- Literature 38 (March): 11­44. ington, D.C. Tyndall, Paramjit. 2000. "The Malaysian Automotive Industry." U.S. International Services. 1999. "Cross-Border Trade in 1998 Prepared for the PECC-ASEAN Auto Project, School of and Sales through Affiliates in 1997." Survey of Current Business Advanced Management Technology, Manila. (October): 48. 625 D E V E L O P M E N T, T R A D E , A N D T H E W T O USITC (U.S. International Trade Commission). 1987. "Standard- Ward, H. 1996. "Common but Differentiated Debates: Environ- ization of Rules of Origin." Report to the Committee on Ways ment, Labour, and the World Trade Organization." Internation- and Means of the U.S. House of Representatives. Washing- al and Comparative Law Quarterly 45 (July): 592­632. ton, D.C. Warr, P. G. 1978. "The Case against Tariff Compensation." Aus- ------. 1997. The Dynamic Effects of Trade Liberalization: An Empir- tralian Journal of Agricultural Economics 22 (2, August): 85­98. ical Analysis. Investigation 332-375. Publication 3069. October. Warren, Tony. 2001a. "The Application of the Frequency Washington, D.C. Approach to Trade in Telecommunications Services." In ------. 1998. Global Assessment of Standards Barriers to Trade in Christopher Findlay and Tony Warren, eds., Impediments to the Information Technology Industry. Publication 3141. Novem- Trade in Services: Measurement and Policy Implications. London ber. Washington, D.C. and New York: Routledge. Valles, Cherise, and Brendan McGivern. 2000. "The Right to Retal- ------. 2001b. "Quantity Impacts of Trade and Investment iate under the WTO Agreement: The Sequencing Problem." Restrictions in Telecommunications." In Christopher Findlay Journal of World Trade 34 (2): 63­84. and Tony Warren, eds., Impediments to Trade in Services: Mea- Van, P. H., and H. Y. Wan. 1999. "Emulative Development through surement and Policy Implications. London and New York: Rout- Trade Expansion: East Asian Evidence." In A. Woodland and J. ledge. J. Piggott, eds., International Trade Policy and the Pacific Rim. Warren, Tony, and Christopher Findlay. 2000. "How Significant Sydney: Macmillan. Are the Barriers? Measuring Impediments to Trade in Services." Vandermerve, S., and M. Chadwik. 1989. "The Internalisation of In Pierre Sauvé and Robert M. Stern, eds., GATS 2000: New Services." Service Industries Journal 9: 79­93. Directions in Services Trade Liberalization. Washington, D.C.: Varian, Hal R. 1985. "Price Discrimination and Social Welfare." Brookings Institution Press and Harvard University. American Economic Review 75: 870­75. Watal, Jayashree. 1995. "MNEs, Market Structure and Price Com- Vaubel, R. 1986. "A Public Choice Approach to International petition in Patentable Drug Markets in India." Presented at the Organization." Public Choice 51: 39­57. seminar on Technology and Globalization, Institute for Eco- Verikos, George, and Keven Hanslow. 1999. "Modeling the Effects nomic Growth, Delhi, April 3. of Implementing the Uruguay Round: A Comparison Using the ------. 1998. "The TRIPS Agreement and Developing Countries: GTAP Model under Alternative Treatments of International Strong, Weak or Balanced Protection?" Journal of World Intellec- Capital Mobility." Presented at the Second Annual Conference tual Property 1 (2, March): 281­307. on Global Economic Analysis, Ebberuk, Denmark, June 20­22. ------. 2000a. Intellectual Property Rights in the World Trade Orga- Vermulst, Edwin A. 1994. "Rules of Origin as Commercial Policy nization: The Way Forward for Developing Countries. New Delhi: Instruments? Revisited." In Edwin A. Vermulst, Jacques Bour- Oxford University Press. geois, and Paul Waer, eds., Rules of Origin in International ------. 2000b. "Pharmaceutical Patents, Prices and Welfare Trade: A Comparative Study, 433­84. Ann Arbor: University of Losses: A Simulation Study of Policy Options for India under Michigan Press. the WTO TRIPS Agreement." World Economy 23 (5, May): Vermulst, Edwin, Jacques Bourgeois, and Paul Waer, eds. 1994. 733­52. Rules of Origin in International Trade: A Comparative Study. Ann ------. 2001. Intellectual Property Rights in the WTO and Develop- Arbor: University of Michigan Press. ing Countries. The Hague: Kluwer Law International. Vinje, Thomas C. 1997. "The New WIPO Copyright Treaty: A Watson, Peter. 2000. "Export Processing Zones: Has Africa Missed Happy Result in Geneva." European Intellectual Property Review the Boat? Not Yet!" World Bank, Washington, D.C. Processed. 5: 230­36. WCO (World Customs Organization). 1999. "Survey of Customs Vogel, David. 1995. Trading Up: Consumer and Environmental Regu- Reform and Modernization: Trends and Best Practices." Avail- lation in a Global Economy. Cambridge, Mass.: Harvard Univer- able at . sity Press. Wellenius, Bjorn. 1997. "Telecommunications Reform: How to Vossenaar, Rene. 1999. "Process and Production Methods: Sizing Succeed." Public Policy for the Private Sector Note 130. World up the Issues from the South." In Halina Ward and Duncan Bank Group, Finance, Private Sector, and Infrastructure Net- Brack, eds., Trade, Investment and the Environment. Washing- work, Washington, D.C. ton, D.C.: Brookings Institution Press. ------. 2001. "Chile: Extending Telecommunications to Rural Vousden, Neil. 1990. The Economics of Trade Protection. Cam- Areas." World Bank, Washington, D.C. Processed. bridge, U.K.: Cambridge University Press. Weston, Ann, and Valentina Delich. 2000. "Settling Trade Disputes Waller, Spencer Weber. 2000. "Anti-Cartel Cooperation." In Simon after the Uruguay Round: Options for the Western Hemi- J. Evenett, Alexander Lehmann, and Benn Steil, eds., Antitrust sphere." LATN Working Paper 10. Latin American Trade Net- Goes Global: What Future for Transatlantic Cooperation? Wash- work, Buenos Aires. ington, D.C: Brookings Institution Press. Westphal, L. E. 1990. "Industrial Policy in an Export-Propelled Walmsley, Terrie Louise. 1999. "Incorporating International Capi- Economy: Lessons from South Korea's Experience." Journal of tal Ownership into the GTAP Mode: Results for Asia-Pacific Economic Perspectives 4: 41­60. Trade Liberalisation." Presented at the Second Annual Confer- Whalley, John. 1996. "Developing Countries and System Strength- ence on Global Economic Analysis, Ebberuk, Denmark, June ening in the Uruguay Round." In Will Martin and L. Alan Win- 20­22. ters, eds., The Uruguay Round and the Developing Economies. Wang, Z. K., and L. Alan Winters. 1997. "Africa's Role in Multilat- Cambridge, U.K.: Cambridge University Press. eral Trade Negotiations: Past and Future." Journal of African ------. 1999. "Special and Differential Treatment in the Millenni- Economies 7 (supp. 1). um Round." Centre for the Study of Globalisation and Region- 626 Bibliography alisation (CSGR) Working Paper N 30/99. May. University of ------. 1998. Intellectual Property Reading Material. No. 476 (E): Warwick, Coventry, U.K. March. 2d ed. Geneva. Wheeler, David, and Ashoka Mody. 1992. "International Invest- Wiseman, Leanne. 2001. "The Protection of Indigenous Art and ment Decisions: The Case of U.S. Firms." Journal of Internation- Culture in Australia: The Labels of Authenticity." European Intel- al Economics 33: 57­76. lectual Property Review 23 (1, January): 14­25. Whichard, Obie. 2002. Measurement and Classification of Service Wolf, M. 1984. " Two Edged Sword: Demands of Developing Coun- Sector Activity: Data Needs for GATS 2000." In Robert M. tries and the Trading System." In Jagdish N. Bhagwati and John Stern, ed., Services in the International Economy. Ann Arbor: Gerard Ruggie, eds., Power, Passions, and Purpose: Prospects for University of Michigan Press. North-South Negotiations. Cambridge, Mass.: MIT Press. Whish, R., and D. Wood. 1994. "Merger Cases in the Real World: Woo, Yuen Pau, and John S. Wilson. 2000. Cutting through Red A Study of Merger Control." OECD, Paris. Tape: New Directions for APEC's Trade Facilitation Agenda. Van- White, Howard, and Edward Anderson. 2000. "Growth versus Dis- couver, B.C.: Asia Pacific Foundation of Canada. tribution: Does the Pattern of Growth Matter?" Institute of World Bank. 1992. The East Asian Miracle: Economic Growth and Development Studies, University of Sussex. Brighton, U.K. Public Policy. New York: Oxford University Press. Processed. ------. 1995. "ARE: Egypt into the Next Century." Report 14048- WHO (World Health Organization). 1996. The World Health Report EGT. May. Washington, D.C. 1996. Fighting Disease, Fostering Development. Geneva. ------. 1996a. "Container Transport Services and Trade: Frame- ------. 2000. The World Health Report 2000. Health Systems: work for an Efficient Container Transport System." Report Improving Performance. Geneva. 15303-CHA. Washington, D.C. Wilson, E. O. 1992. The Diversity of Life. Cambridge, Mass.: Belk- ------. 1996b. World Development Indicators 1996. Washington, D.C. nap Press of Harvard University. ------. 1997. "Ethiopia Export Development Strategy." Wilson, J. S. 1995. Standards and APEC: An Action Agenda. Wash- Processed. World Bank. ington, D.C.: Institute for International Economics. ------. 1998. Assessing Aid: What Works, What Doesn't, and Why. ------. 1999. "The Post-Seattle Agenda of the WTO in Standards New York: Oxford University Press. and Technical Barriers to Trade: Issues for the Developing ------.1999. World Development Indicators 1999. Washington, D.C. Countries." Washington, D.C.: World Bank. ------. 2000a. Can Africa Claim the 20th Century? Washington, D.C. Wilson, John. 2000a. "The Development Challenge in Trade, Sani- ------. 2000b. "India: Policies to Reduce Poverty and Accelerate tary and Phytosanitary Standards." Prepared for the WTO Sustainable Development." Report 19471-IN. Washington, D.C. Committee on Sanitary and Phytosanitary Standards. June. ------. 2000c. Trade Blocs. Washington, D.C. Wilson, J. S. 2000b. "Technical Barriers to Trade and Standards: ------. 2000d. World Development Indicators 2000. Washington, D.C. Challenges and Opportunities for Developing Countries." ------. 2000e. World Development Report 2000/01: Attacking World Bank, Washington, D.C. Processed. Poverty. New York: Oxford University Press. Wilson, J. S. 2001. "Standards, Regulation, and Trade: Recommenda- ------. 2001a. Finance for Growth: Policy Choices in a Volatile tions from a Development Perspective." Background paper for World. World Bank Policy Research Report. New York: Oxford World Development Report 2002. World Bank, Washington, D.C. University Press. Winham, Gilbert R. 1986. International Trade and the Tokyo Round ------.2001b. Global Economic Prospects and the Developing Coun- Negotiation. Princeton, N.J.: Princeton University Press. tries 2001. Washington, D.C. ------. 1998. "Explanations of Developing Country Behavior in ------. 2001c. Global Economic Prospects 2002: Making Trade the GATT Uruguay Round Negotiation." World Competition Law Work for the World's Poor. Washington, D.C. and Economics Review 21 (3): 109­34. ------.2001d. "Information and Communications Technology Winters, L. Alan. 1997. "Assessing Regional Integration Agree- Sector Strategy Paper." Washington, D.C. Available at . ------. 1998."Regionalism versus Multilateralism." In Richard Worldscope. 1998. "Corporate Financial and Ownership Indica- Baldwin, Daniel Cohen, André Sapir, and Anthony Venables, tors." Bethesda, Md. eds., Market Integration, Regionalism and the Global Economy. WTO. 1994a. Compilation of Horizontal Commitments, All Sectors. London: Centre for Economic Policy Research. Geneva. ------. 1999. "Regionalism and the Rest of the World." World ------. 1994b. Country-wise Schedules of Commitments. Geneva. Bank, Washington, D.C. Processed. ------. 1996a. "Japan: Taxes on Alcoholic Beverages. Report of ------. 2000a. "Trade and Poverty: Is There a Connection?" In B. the Appellate Body." Available at . parity and Poverty, 43­49. Geneva: WTO. ------. 1996b. Trade Policy Review: Korea. Geneva. ------. 2000b. "Trade Policy as Development Policy: Building on ------. 1997a. Opening Markets in Financial Services and the Role Fifty Years' Experience." Proceedings of the High-Level Round of the GATS. Special study. Geneva. Table on Trade and Development, 10th United Nations Confer- ------. 1997b. "A Review of Statistics on Trade Flows in Services." ence on Trade and Development, Bangkok, February. Available at WTO Council for Trade in Services. Geneva. . ------. 1997c. WTO Annual Report 1997. Geneva. WIPO (World Intellectual Property Organization). 1997. Asian ------. 1998a. "Japan: Measures Affecting Consumer Photo- Regional Round Table on Implementation of the Agreement on graphic Film and Paper, Report of the Panel, 31 March." Avail- Trade-Related Aspects of Intellectual Property Rights, Singapore, able at . 627 D E V E L O P M E N T, T R A D E , A N D T H E W T O ------. 1998b. "Presence of Natural Persons." Background note. Yang, Guifang, and Keith E. Maskus. 2001. "Intellectual Property WTO Council for Trade in Services. Geneva. Rights and Licensing: An Econometric Investigation." ------. 1998c. "WTO Adopts Disciplines on Domestic Regulation Weltwirtschaftliches Archiv 137: 58­79. for the Accountancy Sector." WTO Working Party on Profes- Yeats, Alexander. 1995. "Are Partner Country Statistics Useful for sional Services. Geneva. Estimating `Missing' Trade Data?" Policy Research Working ------. 1999a. "Indonesia--Certain Measures Affecting the Auto- Paper Series 1501. World Bank, Washington, D.C. mobile Industry: Report of the Panel." WT/DS54/R, Young, L., and J. Romero. 1994. "A Dynamic Dual Model of the WT/DS55/R, WT/DS59/R, WT/DS64/R. Geneva. North American Free Trade Agreement." In Joseph F. Francois ------. 1999b. "Preparations for the 1999 Ministerial Confer- and Clinton R. Shiells, eds., Modeling Trade Policy: Applied Gen- ence." Job (99)/479/Rev. 3. Geneva. eral Equilibrium Assessments of North American Free Trade. Cam- ------. 1999c. "Technical Note on the Accession Process." bridge, U.K.: Cambridge University Press. WT/ACC/7. March. Geneva. Youssef, H. 1999. "Special and Differential Treatment for Develop- ------. 2000a. "Communication from Mauritius on Behalf of the ing Countries in the WTO." TRADE Working Paper 2. South African Group." Doc. Ref. S/CSS/W/7. Geneva. Centre, Geneva. ------. 2000b. International Trade in Services. Geneva. Yu, E. S. H., and C. C. Chao. 1998. "On Investment Measures and ------. 2000c. "World Trade in 1999." WTO Secretariat. Geneva. Trade." World Economy 21 (Global Trade Policy Issue): 549­61. ------. Various issues. Trade Policy Review. Geneva. Zeiler, Thomas W. 1992. American Trade and Power in the 1960s. ------. Various years (1997, 1998). Miscellaneous Documents on New York: Columbia University Press. Professional Services. Working Party on Professional Services. Zutshi, B. K. 1998. "Bringing TRIPs into the Multilateral Trading Geneva. System." In Jagdish Bhagwati and Matthias Hirsch, eds., The WTO Secretariat. 1999. Guide to the Uruguay Round Agreements. Uruguay Round and Beyond: Essays in Honor of Arthur Dunkel. The Hague: Kluwer Law International. Ann Arbor: University of Michigan Press. 628 Index access control, 392 ALADI. See Latin America Integration Association (ALADI) Agree- accounting services, 251, 295­96b ment ACP. See African, Caribbean, and Pacific (ACP) countries Andean Community, 337­38, 339 ad valorem tariff rates, 249, 250t, 531 Andean Pact Open Skies Agreement, 279 adjustment, phased, 36 Andean Trade Preferences Act (ATPA), 116b, 118 adjustment assistance, 33, 54 Anderson, Kym, 11­16 Advance Cargo Information System, 141 anticartel policies, 456­62 advertising, 404, 405 anticompetitive practices, 216, 279, 293­94; IP ownership and, Advisory Centre on WTO Law, 75b 405; IPRs and, 374; maritime trade, 294; spillover effects, aflatoxin levels, 431, 433b 448 Africa: access to credit, 301b; bank failures, 299b; capacity-build- antidumping actions, 68, 70n6, 73­74b, 95, 203, 205nn8­9; anti- ing efforts, 509­15; EPZs, 166­67; export diversification in, circumvention of, 126; industrial policy and, 154; overview, 523­25b; financial sector liberalization, 223b; growth in serv- 199­200; typewriters, 211­12n1; WTO rules, 201b ices trade, 238; infrastructure, 226; JITAP, 510­15; regulatory antidumping and countervailing duty (AD/CVD) cases, compensa- reform, 230; RIAs listing, 549­50b; World Music, 390, 400n1 tion for, 84, 85­86, 91nn10­12, 206­7, 208­11, 212n2 African, Caribbean, and Pacific (ACP) countries, 87­88, 106­7b antitrust laws, 207­8, 457 African Growth and Opportunity Act , 115­16b APEC. See Asia-Pacific Economic Cooperation (APEC) agenda-setting, GATT process, 489­90b Appellate Body (AB), 71, 72, 80n4; Bananas case, 87­88b; deci- Agreement on Agriculture, 102­3 sions, 478­80; role of, 78, 80n8 Agreement on Antidumping, 126 applied tariffs, 98­99, 541, 542t, 545­47 Agreement on Basic Telecommunications, 293­94b, 318 arbitrage in markets, 7 Agreement on Customs Valuation, 128­36, 140 Argentina, 173, 223­24b, 487; biotechnology, 363; electricity dis- Agreement on Government Procurement (GPA), 422­24, tribution, 303; labor force, 228; overvaluation of exchange 425­26b rate, 19­20; parallel imports, 363­64; port sector, 227b; SPS Agreement on Implementation of Article VI, 201b standards, 497 Agreement on Import Licensing Procedures, 140 Asia, 76, 93, 238; competition in telecommunications, 233­34b; Agreement on Preshipment Inspection (PSI), 132­33, 137­38n7, cross-border mergers and acquisitions, 457, 458t; financial cri- 140 sis, 4, 5; mergers and acquisitions, 456; RIAs listing, 550b Agreement on Rules of Origin, 122­23, 140 Asia-Pacific Economic Cooperation (APEC), 142, 248, 323, 344 Agreement on Safeguards, 103, 327­28 ASYCUDA. See Automated System for Customs Data and Manage- Agreement on Sanitary and Phytosanitary Measures (SPS), 430b, ment (ASYCUDA) 434, 435f, 435t, 436, 437, 473, 495­97, 499 asymmetric information, 291, 298, 351­52, 405, 407 Agreement on Subsidies and Countervailing Measures (SCM), 72, ATC. See Agreement on Textiles and Clothing (ATC) 154, 159n9, 167, 168­69b, 331­32 ATPA. See Andean Trade Preferences Act (ATPC) Agreement on Technical Barriers to Trade (TBT), 434, 436, 437, Australia, 179­85, 278, 363, 393­94, 434b 476, 506 Australia­New Zealand Closer Economic Relations Agreement, Agreement on Textiles and Clothing (ATC), 69, 103, 174, 178n8, 176 186­94 Australian Productivity Commission, 278 agricultural industries, 102­3; discrimination against, 520; effect Austria, 189­92 of exchange rate distortions on, 13­14; export subsidies, 100; Automated System for Customs Data and Management (ASYCU- Mexico, 35; standards for, 438b; support to under WTO, 67, DA), 134, 141, 144 tariffs and, 55, 59n8, 256, 257t, 533; taxation of, 11­12 automobile industry, 175; Australia, 179­85; effect of procure- agricultural supports, 12­13 ment bans on, 421; Mexico, 355; overview, 179; tariff struc- air transport services, 247, 278­79 ture, 539n3 629 D E V E L O P M E N T, T R A D E , A N D T H E W T O balance of payments: problems of, 506; services trade and, 236b, trade policy review of, 45b; TRIMs agreement and, 174; TRIPs 240, 241, 243t, 246n4; statistics and GAT, 236b; use of tariffs disputes, 173; WTO forums, 48 and, 529 Canada-U.S. Shared Border Accord, 142 Balance of Payments Manual, 245b capacity-building efforts, African countries, 509­15 Bananas case, 53, 72, 76, 87­88, 89, 367 capital account liberalization, 299­300b bananas, duties on, 107b capital markets, 372 Bangladesh, 34, 300 Caribbean Basin Initiative, 99 banking services, 249, 251, 265; liberalization of, 223b; market Caribbean countries, offshore financial services, 292b access commitments, 270­72t; price-based measure of barriers CARICOM, 339, 340 to, 251; structural reform in, 231, 234n8; under GATS, 261. cartels: anticartel policies, 456­62; price-fixing by, 457, 458t See also financial services CBD. See Convention on Biological Diversity (CBD) bans on foreign purchases, 420­21, 427n1 CEE. See Central and Eastern Europe (CEE) barriers to trade, 1; electronic commerce, 318, 319b; implications ceiling binding, 344 for research and policy, 256, 258; measurement of, in trade Center for Facilitation of Procedures and Practices for Administra- and FDI in services, 249­53; modeling of, 253­56; technology, tion, Commerce, and Transportation (CEDACT-UN/ECE), 351; types of in FDI, 248, 258n1; types of in services trade, 141­42 247­48; welfare effects of, 257t. See also nontariff barriers; Central American Common Market (CACM), 117, 118­19, 120, technical barriers to trade 555 BDV. See Brussels Definition of Value (BDV) Central and Eastern Europe (CEE), 192, 194n5 Bed Linen case, 73­74b Central and Eastern European countries (CEECs), 452­53 Beef Hormone case, 53, 89, 473 Central European Free Trade Area (CEFTA), 309b beef imports, 432 certification: procedures, 429­31; requirements for, 248, 305, behind-the-border agenda, 1, 413­14 431­42; tires in Brazil, 433b Berne Convention for the Protection of Literary and Artistic Works, certification marks, 394, 401nn11­14 391, 400n2 CFA zone countries, 19; GDP growth, 22, 23n8; mock devalua- Bhagwati, Jagdish, 452b, 552 tions, 22; overvaluation of exchange rate, 21­22, 23n6 bidding on contracts, 420, 422 CGE. See computable general equilibrium (CGE) bilateral agreements: cross-border cartels and, 449; disciplines on Chad, 21, 22 monopoly practices, 339; EU creation of, 453 Chauvet, P., 235­46 bilateral investment treaties (BITs), 357, 440 child labor, 467 bilateral trade flows, 239 Chile, 299b, 531, 532; duty drawback experience, 164b; MFN tar- binding tariffs, 43, 55, 56, 67, 540­47; of industrial and develop- iffs, 31; overvaluation of exchange rate, 19­20; price bands, ing economies, 98t, 101; market access and, 541­42, 545­47; 32; Scallops case, 73b; telecommunications, 302b; TRIMs overview, 541, 545t; rates, 541, 542t agreement and, 173; uniform tariffs, 530 biodiversity, 382­84, 385, 388n3, 389nn9­10, 389n12, 477, China: accession to the WTO, 9, 68­69; control rights, 4; exports 481n9 from coastal regions of, 225­26b; growth of services trade, biopiracy, 383 238; health care spending, 381b; Internet use, 317; MFN sta- biosafety, 480 tus, 70n1; R&D facilities, 370; trademark infringement, 371 biotechnology, 363, 382, 384 clothing and apparel, 115­116b. See also Agreement on Textiles BITs. See bilateral investment treaties (BITs) and Clothing (ATC) black market for currency, 22n2 CLSs. See core labor standards (CLSs) Border Industrialization Program, 355 coalition diplomacy, 489­90b border prices, 29b Codex Alimentarius, 433b, 434, 437n1 bourbon vanilla cartel, 27b Colombia, 173, 223b brand names, 404, 405, 406, 411n3 Columbus Declaration, 141 Brazil, 299b, 363; imported tires, 433b; industrial property law, commercial policies, 546­47 375b; mergers, 460; MFN tariffs, 31; patented drugs in, commercial presence, 221, 222, 266, 334, 339, 419­20; commit- 366­67b; pharmaceutical industries, 374­75b; TRIPS agree- ment by sector, 308t; foreign, 280; GATT safeguards and, 327, ment and, 76­77, 80n5 328, 329, 332, 334; inadequacies of statistical domains for, Brussels Definition of Value (BDV), 128­29, 132b, 137n4 236, 237t; market access and, 281t; restrictions on, 281t, 305, business-to-business exchanges, 316 314n2; taxation and, 283; telecommunications, 279n3 businesses, establishment of, 35­36 commercial services, 238, 239t, 241t commitments, 43; agreement on, 341, 345n5; by sector and CACM. See Central American Common Market (CACM) mode of supply, 304, 308t; freezing existing regimes, 344; Café au Lait diplomacy, 489­90b horizontal, 307, 310; levels of, 344­45; list or lose, 345; mar- Cameroon, 21, 22 ket access negotiations, 280­89; on movement of natural per- Canada: changes in export revenues, 111, 112b; GSP and, 109, sons, 254f, 306­10; ratcheting, 344­45; schedules of, 339; 110t; negative list approach to NAFTA, 337; rights of indige- structure of, 310; telecommunications, 272; under GATS, nous peoples, 393; Shared Border Accord, 142; tariff peaks, 260­65, 273­76t, 279n2; under TRIMs agreement, 174 105­6, 108­9, 111; textile and clothing restrictions, 189­192; Committee on Regional Trade Agreements, 555 630 Index Committee on Trade and Environment (CTE), 472­73, 476 in developing countries, 495; reform and modernization pro- Common Market of the South (MERCOSUR), 117­18, 119, gram, 141; reform projects, 495, 496t; use of technology in, 337­40, 531, 554, 556 146 Common Trust Fund, 511 customs duties, 319­20b communications, 249, 280, 281t customs origin rules, 125 compensation, 197; compensatory liability regime, 398­99; in dis- customs union, 79, 553 pute settlements, 84­86; as GATT safeguard, 52, 59n5; and customs valuation, 140, 484, 494­95; agreement provisions, intra-industry trade, 54; policies for developing countries, 33­35 129­30; fallback method of, 131­32; harmonization of proce- competition, 161, 214, 232, 291, 317, 447; contract rigging, dures, 135­36; historical overview of agreement, 128­29, 420; data base collection and, 398; effect of labor standards 137n2; implementation of agreement, 130­36; Kyoto Conven- on, 466b; ESM concerns and, 328, 329; free trade and, 452b; tion, 135­36b; provisions of, 128­30, 137n2 improvement of position in markets, 206, 211­12n1; interna- tional rules for, 447­55; intrabrand, 407; laws, 452­53; licens- dairy products, 432, 433­34b ing agreement, 376; link with growth, 451; misleading origin data collection: customs valuations, 133; data-processing services, indications, 394­95; policy and market access relationship, 292b, 320b; electronic data interchange, 141­42; European 450b; precommitment to, 283­85; procompetitive issues, Database directive, 398; genetic resources, 388, 389n15; PERI- 218­19t, 228, 297­98b; spillovers and, 465, 467; telecommu- NORM database, 431; services trade, 244­45b, 252 nications, 189nn2­3, 233­34b, 281­82, 293­94b; trade liber- de minimis clause, 118, 121n3 alization and, 230; unfair, 394­95, 398, 400; use of IPRs and, de Wulf, Luc, 160­70 374. See also anticompetitive practices decentralization, 488 competition law, 451, 455n5, 457 Declaration Concerning Integrity in Customs, 141 complementary policies, 1, 35­36, 519, 521­23 deforestation, 382­83 compulsion, 467­69 Delich, Valentina, 71­80 computable general equilibrium (CGE) model, 222, 229b, 531 delivery logistics, electronic commerce, 318 computer and information services, 240, 241 delivery systems, 315, 316 concessions, 43, 56, 59nn9­10, 488; adjustments of, 52; on Department of Commerce, U.S., 208­10 import tariffs, 180; intra-industry trade and, 54; market access Department of Justice, U.S., 449, 457 to SSA countries, 115­16b; as means to control free-riding, deregulation, 222 53­54; suspension of, 72, 77­78; and Uruguay Round, 57t, Devarajan, Shantayanan, 21­22, 23n5 58t; use of tariffs to obtain, 529 differential pricing, 378­79b conformity assessment, 432 differentiated tariff protection, 527 consumer products, 404, 405 Dillon Round, 53 Consumer Unity & Trust Society (CUTS), 34, 451 direct insurance services, 249, 251 consumption, 32­33, 221, 222 discretionary allocation mechanisms, 33 consumption abroad, 236, 237t, 308t, 327, 332 disease treatments, 366­67b content schemes, 175 Dispute Settlement Body (DSB), 71­72, 73, 74, 77 contracts, 420, 422 dispute settlement mechanisms, 39, 41, 52­53, 314n7; adequacy Convention on Biological Diversity (CBD), 385, 389nn9­10, 480 of WTO system, 81­91; aim of, 59n5; as benefit of WTO mem- copyright, 360b, 361, 391­92, 400, 400­401nn4­5; definition, bership, 62; changes in, 46; developing countries capacity to 411n1; digital form of works, 392­93, 401nn7­10; infringe- use, 487; effects of trade and environmental debate on, ments of, 347, 394; "life plus 70" rule, 401n5; protection of, 478­80; IPRs, 361­62; market access commitments and, 43; 371, 372, 404, 411n3, 498; time limits on, 404, 411n3 reform of, 478­80; telecommunications, 294b; under TBT and core labor standards (CLSs), 464­65b, 466­67b SPS, 436; under TRIMs, 172­73; use of to date, 365, 367 Cornejo, Rafael, 114­15 Dispute Settlement Understanding (DSU), 71­73, 76­77, 78, 79, Costa Rica, 388, 397 80n4, 468; cross-retaliation under, 87, 89, 91n14; reform Côte d'Ivoire, 21­22, 23n6 made by, 82; remedy under, 83; Shrimp case, 479­80 Cotonou Agreement, 99, 106b, 109 disputes: action prior to filing trade case, 208; environmental pro- counterfeit firms, 372 tection, 473; injury determination, 208­11; IPRs, 76­77, 80n5; countervailing measures, 154, 159n9, 165, 167, 168­69b, minimizing final decisions, 211; nonpolitical remedies for, 209­11, 332, 335n3 206­7; notification of, 207­8; political remedies for, 207; post- crafts industry, 393­96, 401nn11­14 petition filing, 208­11, 212n2; trademarks, 410, 412n10 credibility-enhancing institutions, 444 distortions: costs of, 531, 534n6, 534n11; of exchange rates, credit markets, 35 13­14; in FDI policy, 441b, 446n1; of prices, 11; sectoral cross-border supply, in services, 321b approaches to liberalization, 536­37; source of, 528­29 cross-border trade, 221, 222, 232, 236, 237t, 327­29, 449 distributional policy, 521 cross-subsidization, 282, 300, 331 Doha Development Agenda, 548, 555 CTE. See Committee on Trade and Environment (CTE) Doha Ministerial Declaration, 104, 368, 414 Cuba, 286b, 412n10 domestic regulations, 428; accountancy sector, 295­96b; customs administration, 33­34, 94, 139, 520; costs, 144; Jamaica, approaches to, 291b; asymmetric information, 298; costs of, 145­46b; postrelease audits, 146­47; practices and problems 298b; efficiency and equity conflicts, 298, 300­303; electronic 631 D E V E L O P M E N T, T R A D E , A N D T H E W T O commerce, 292b, 317­18; monopolies, 297; overview of regu- European Commission, 448, 453, 457 lation in services, 290­91, 293; regional agreements and, 338, European Common Market, 448 345n2 European Community (EC), 74, 85, 91n11 domestic service suppliers, 328 European Database Directive, 398 domestic suppliers, favoritism, 418­19 European Economic Community (EEC), 87­88b, 538, 554 domestically prohibited goods (DPGs), 476, 477­78 European Patent Office, 396 Dominican Republic, 166, 167, 339 European Union (EU), 7, 142, 199, 251, 363, 377; aflatoxin levels downstream markets, 226, 228, 230 standard, 431; Bananas case, 87­88, 89; bilateral trade flows, drawback systems, 164­65, 166, 179n1 239; changes in export revenues, 111, 112b; competition and, DSB. See Dispute Settlement Body (DSB) 453­54; customs valuation, 129; electronic transactions, 322; DSU. See Dispute Settlement Understanding (DSU) "Everything but Arms" initiative, 106­7b; fish and fish prod- Dunkel, Arthur, 51 ucts, 430b; growth of services trade, 238; GSP and, 109, 110t; duty drawback scheme, 163­65, 166, 170n1, 533 liberalization of financial sector, 299b; national treatment of investors, 443­44; peanut standards, 433b; privacy, electronic Earth Summit, 475, 481n4 commerce, 292b; RIAs, 444; safeguard actions, 68; Single East African Community (EAC), 555 Market Program , 431; tariff peaks, 105­6, 108­9, 111; textile Eastern Caribbean Telecommunications Authority, 298b and clothing restrictions, 189­92, 193­94b; trade policy Ebrill, Liam, 24­27 review of, 45b; trademark protection, 412n10; waste regula- EC. See European Community (EC) tions, 436; WTO forums, 48 ecolabeling, 473, 475, 476­77 Evenett, Simon J., 417­27, 421, 422, 456­62 Economic Coal and Steel Community, 554 "Everything but Arms" initiative, 106­7b Ecuador, 76, 88, 89­90, 91n17 ex ante assurance, 288 EDI. See electronic data exchange (EDI) exchange rates, 2, 22n1: devaluations of, 17­18, 22, 23n7, 26; Edwards, Sebastian, 32 effects of distortions of, 13­14; local content and, 183; man- EEC. See European Economic Community (EEC) agement of, 17­22; misalignment of, 19; overvaluation of, effective rate of protection (ERP), 37n3, 151b 18­22; trade protection and, 17, 22­23n3 Egypt, 33 exclusive marketing rights (EMRs), 362, 367 Ehrlich, Paul, 384 exemptions, tariff, 25 electronically delivered products, 215 exhaustion regimes, 377 electronic commerce, 218t, 292b, 316­24, 392­93, 401nn7­10, export cartels, 448­49, 455n4 411 export clearance, 145­46b electronic data interchange (EDI), 141­42 export facilitation, 180, 183­84, 185n6, 185n8 electronic payments infrastructure, 317­18 export subsidies, 80n8, 154, 159nn8­10, 167, 333, 335n4; agri- embargoes, 478­79 cultural, 100; use of, 506, 507 emergency actions, under GATT, 197­98, 199, 200f, 205n6 export taxes, 26, 27b, 427n1 emergency safeguard mechanism (ESM), 327­28, 334­35 export-processing zones (EPZs), 7­8, 94, 165­67, 466­67b, 533 emigration, 62 exporters, value of open trade regimes to, 526, 534n1 employment, 183, 185nn4­5, 228, 303, 303nn3­4 exports, 94, 317; concessions and, 56, 59nn9­10; developing EMRs. See exclusive marketing rights (EMRs) countries, 101t, 111; diversification in Africa, 523­25b; duty Enabling Clause (GATT), 486, 505 drawback scheme, 163­65, 166, 170n1; financial sector and, Endangered Species Act, 478 16­18, 169; from coastal regions of China, 225­26b; goods enforcement, 294; anticartel, 462n2; of GATT dispute settlement and services, 235, 237­38; LDCs, 109, 111; matching grant system, 81­83; IPRs, 410; trademarks, 371, 409; TRIPS agree- programs, 161, 162­63b; natural resource-based, 224, 234n3; ment, 498 promotion of, 150, 152, 154, 155­56t, 159n9; services trade, English, Philip, 160­70 240t, 286­87b; software services, 224; tariff peaks and, 109, enquiry points, 434­35 111; temporary admission scheme, 163­65, 166, 170n2; TPOs enterprise support, 161, 163 and, 160­63; VERs, 198t, 199, 200t, 205n7, 205n9 entry requirements, 285, 286b, 307 extended balance of payments services, 245b entry tax, 282 external balance of trade, 18 environmental protection, 15, 472­78 external market interventions, 150 environmental services, 287 externalities, 150 EPZs. See export-processing zones (EPZs) factor market interventions, 150­51, 152b, 372 ERP. See effective rate of protection (ERP) factor prices, 14 escape clause of GATT, 52, 207 fallback method of valuation, 131­32 ESM. See emergency safeguard mechanism (ESM) FATS. See foreign affiliate trade statistics (FATS) ethnobotanicals, 396­99 FDI. See foreign direct investment (FDI) EU. See European Union (EU) Federal Communications Commission, U.S., 298b EU-Mexican Free Trade Agreement, 453 Federal Register, 209­10 EU-South Africa Free Trade Agreement, 453 financial services, 249, 261, 266, 318; commitments in, 266; Europe, 6, 549b export policies and, 167­68, 169; grandfather provisions, 280, 632 Index 289n1; liberalization of, 223­24b, 234n5, 299­300b, 301b; eralization programs and, 283­85; main provisions of, 260t; market access, 249, 251, 267b, 280, 281t; offshore, 292b; maritime negotiations in the WTO, 277­78; market access safeguard mechanisms, 329­30; South Africa, 232, 234n5; issues, 285­88, 340; market failure issues, 293­94, 296­97; under GATS, 265­67, 265­72, 270­71t, 272, 318 movement of natural persons and, 305­10, 311­13; negotia- Financial Services Agreement, 318 tions and domestic policy issues of, 218­19t; positive list financial-based measures of services barriers, 252 approach, 423; principles of, 259­60; service exclusions and Finger, J. Michael, 8, 50­60, 131, 136, 137n5, 493­503 reservations under, 340; services statistics and, 236b; specific Fink, Carsten, 293­94b, 403­12 commitments under, 260­65, 279n2; telecommunications, Finland, 85, 189­92 272­77, 279n3; treatment of investors, 439, 440; view of fish and fish products, 430b domestic regulations in, 290 fixed exchange rate, 17, 20 General Electric, 457 folklore, 391­92, 393, 400­401nn3­4 Generalized System of Preferences (GSP), 99, 105, 106b, 109, food products sector, 371 487; exclusions from, 115b; origin regimes and, 114; S&D food safety standards, 432 treatment of LDCs, 505­6 foreign affiliate trade statistics (FATS), 240, 241­43, 245b, 246n4 genetic resources: overview, 382­83; patent applications for, 396; foreign capital, 228 proprietary protection of, 385­86, 389nn9­11 foreign currency revolving funds, 167 genetically modified organisms (GMOs), 473, 475 foreign direct investment (FDI), 14, 166, 171, 439, 440; activities geographical indications: multinational agreement on, 410­11; of foreign affiliates and, 242, 243t; gains from as means of overview, 403­4; rules of origin, 124­25, 395­96, 400, trade, 228; impact on RIAs, 553­54; indexes of restrictions, 401nn15­16; trademarks and, 409 249, 250t; international spillovers and, 442­43, 446n3; IPRs Germany, 87­88b, 126 and, 364, 369­70; market access benefits of, 280­81; mergers Global Competition Forum, 454 and acquisitions, 456; procurement policies and, 333; services Global Environment Facility, 389n12 trade and, 222, 226, 247­48; technology and, 150, 151, 352, global warming, 474b 353­54, 356­58, 358n4; technology transfer and, 352, globalization, 8 356­58, 358n4; TRIPS and, 157; WTO rules on, 441­42b GMOs. See genetically modified organisms (GMOs) Foreign Investment Review Act (FIRA), 174 governance, in trade relations, 488, 490­91 foreign labor content entitlements, 287­88 government assistance, 207 foreign sales corporations (FSCs), 170 government procurement, 219t, 306, 314n5, 333­34, 339, 417; foreign services providers, 305, 311­12, 327­28, 345n1 corruption and collusion scope, 420; cost overruns and rene- foreign-owned companies, 240, 241, 242­43, 246n4 gotiation, 420; discrimination in, 420­22; GPA, 326, 422­24; forward-looking remedies under GATT, 85 India, 425b; Korea, 424t; market for, 418; reform, 424, France, 87­88b, 431 426­27 Francois, Joseph, 193b, 251, 418, 540­47 Government Procurement Agreement (GPA), 326, 422­24 free trade agreements (FTAs), 109, 114. See also names of specific government services, 239 agreements grandfather clauses, 344 Free Trade Area of the Americas (FTAA), 120, 142, 323 gravity-model measures, 251­52 free trade, competition policy and, 452b green-box policies, 103, 104 free-rider problem, 53­54, 399, 404 green-room practice, 491 frequency measures of services barriers, 249, 251 Gropp, Reint, 24­27 FTAs. See free trade agreements (FTAs) Group of Three, 119, 337, 339 GSP. See Generalized System of Preference (GSP) Gamberale, Carlo, 290­303 Gulf Cooperation Council, 554 Garay, Luis Jorge, 114­21 General Agreement on Tariffs and Trade (GATT), 41, 49n1, 52, Harberger triangle, 545 140, 486; customs valuation, 128­38; developing countries' harmonization: of customs procedures, 135­36b; tariff levels and, involvement in, 485­87; dispute settlement provisions, 52­53, 538 59n5, 81­90; export subsidies, 169b; function, 49; historical Harmonization Work Program, 122­27 review of, 44­46, 47t; multilateral versus unilateral liberaliza- Harmonized Commodity Description and Coding System (HS), tion, 535, 539n1; negotiations under, 50­52, 59n2; principles 108, 116­19, 126, 140 of, 44­45; reciprocity rules in, 50­53; S&D treatment, 505; Havana Charter, 447­48, 453 safeguard actions under, 52, 59n4, 195­200, 204nn1­6; struc- Heads of Customs Conference, 142 ture of, 46­47; textile and clothing industry, 187; TRIMs health services, 286­87b, 380­81b agreement and, 171­78, 174, 177n4, 177nn6­7 Hewlett-Packard Company, 432 General Agreement on Trade in Services (GATS), 47, 158, 168b, high-technology industries, foreign workers in, 309b 176, 213­15, 439, 451; air transport services, 278­79; Hirschman, Albert O., 352 Bananas case, 89, 91n17; disciplines on subsidies, 331­32; HIV/AIDS, 366­67b, 375b, 378b, 381b environmental issues, 480; financial services, 265­72; frequen- Hoekman, Bernard, 98, 259­79, 425b, 439­46; 48­58, 558n4; on cy measures and, 249; government procurement and, 333; barriers to trade, 247, 249, 252; on dispute settlements, 79; hierarchy of instruments of protection in services, 282­83; lib- on procurement, 421; on RIAs, 444 633 D E V E L O P M E N T, T R A D E , A N D T H E W T O hold-back problem, 288 industrialization, 15, 98 Holden automobile, 179­80 Industries Assistance Commission, Australia, 180­85, 184, 185n1, Holmes, Peter, 447­55 185n6, 185n9 Honeywell, 457 infant industries, 149­50, 159n6, 281; production subsidies and, Hong Kong (China), 451 534n4; protection of, 157, 202; restructuring protection and, Horlick, Gary N., 206­12 528, 534n4 horticulture, 35, 37n6 inflation, 19­20 households, trade effects on members of, 29b information, exchange of, 43­44, 449 HS. See Harmonized Commodity Description and Coding System information markets, 399 (HS) information technology, 133, 228, 324, 432 Hudec, Robert, 53, 55 Information Technology Agreement (ITA), 318, 536 human capital, 372, 380b infrastructure, 35, 231; Africa, 226; China, 225­26b; electronic human security, 469 commerce, 317­18, 321­22; EPZs and, 167; technology, Hume, David, 19 372­73; telecommunications, 272 Hungary, 299b injury, 154, 159n10, 207, 328; determination of, 208; investiga- tion of, 203b; to domestic industries, 327, 332 ILO. See International Labour Organization (ILO) inspection of goods, 132­33, 137­38n7 immigration, 305, 309b institutional reform, 1, 7­9 impairment test, 401n10 instruments, government, 318­19 implementation audit, 51­52 insurance industry, 265, 268­69t implementation, costs of, 484, 502b integrated circuits, 498, 503n3 import compression syndrome, 22­23n3 Integrated Framework for Least-Developed Countries, 216, 492, import protection, 526, 534n1 510­15 import-substituting industrial policies, 20­21, 99, 355 Integrated Framework Trust Fund (IFTF), 513 imports, 163, 417; Australia, 180­181; bans on fish and fish prod- integration program, textiles and clothing, 187­88, 189t, 190t ucts, 430b; escape clause relief, 52, 207; ESM concerns and, intellectual property (IP), 405; definition, 347, 359; music indus- 327-28; licensing systems, 140, 180; Quad economies, 108-9; try, 391; protection standards under TRIPS, 361 quotas on, 95; restrictions of, 150, 152b; Shrimp case, 478­79; intellectual property rights (IPRs), 347, 361, 369, 497­98, 499t, textiles and clothing, 186. See also antidumping actions 503nn1­4; agreement on protection of, 386­88; building mar- in-house production, 182, 185n2 kets for, 371; competition rules and, 374; as dispute settlement Inama, Stefano, 122­27 theme, 76­77, 80n5, 361­62; domestic innovation of, 370­71; incentives, 443, 446nn4­5, 474b economic implications of, 364­65; enforcement of, 45, 361, income distribution, 534n3 410; extension of, 498, 503n3; factor market flexibility, 372; India, 317, 396, 555, 558n4; antidumping measures and, human capital development for, 372; of indigenous peoples, 73­74b; copyright protection in, 371; crafts industry, 393, 393­94; international convention, 498; licensing agreements, 401n13; FDI in, 442b; growth of services trade, 238; mango 375­76; open market access, 373­78; overview, 360­61b, pulp, 433b; milk products, 433­34b; procompetitive regula- 369­70; protection of, 5, 157, 507; quality assurance for, 371; tion, 297­98b; public sector enterprises, 425b; regulatory technology infrastructure, 372­73; treatment of parallel reform, 230; software industry, 224, 286; standards-related imports and, 376­78; versus trademark protection, 404. See problems, 433­34b; steel company, 434b; telecommunica- also technology transfer tions, 233b; 89n3, 297­98b, 303n3; TRIMs agreement and, internal balance of trade, 18 174, 175, 177­78n5 international accounting standards, 296 indigenous crafts, 393, 394 International Accounting Standards Committee, 296 indigenous peoples, 396­99 International Animal Health Code, 433b indirect tax rebates, 165, 170n1 International Bar Association, 460 Indonesia, 27b, 173, 174, 175, 177­78n5 International Chamber of Commerce, 142­43 industrial countries, 93, 103, 496; Appellate Body rulings, 80n9; International Competition Policy Advisory Committee, 449, 454 compensation in dispute settlements, 85; gains from trade lib- International Convention on the Simplification and Harmonization eralization, 222, 225­26; IPR regime, 373; search for genetic of Customs Procedures, 135­36b, 141 resources, 382, 388n2; share or services production, 235; tar- International Electrotechnical Commission, 429 iffs and, 55, 59n8, 105; Uruguay Round outcomes, 58; valua- International Express Carriers Conference, 146 tion agreement, 130­31; view of developing countries International Federation of Accountants, 296 involvement in trade talks, 486 International Labour Organization (ILO), 463, 464­65b industrial designs, rights to, 360b International Monetary Fund (IMF), 154, 159n5, 509 industrial policy: definition, 149, 159n1; East Asia, 151­54, International Organization of Securities Commissions, 296 155­56t; evolution of, 151­54; government procurement and, International Organization for Standardization (ISO), 161, 428, 426b; implications of WTO rules for, 158; instruments of used 429 in Korea and Japan, 152b; multilateral rules regarding use of, International Standard Industrial Classification (ISIC), 245b, 249 154, 157­58; objectives and scope of, 149­51, 152b International Trade Centre, 160, 509 industrial property rights, 360­61b, 375b, 396­98 International Trade Commission (ITC), 207, 208, 210 634 Index International Trade Organization (ITO), 41, 53; 86, 91n14, 203b, 285, 303; telecommunications industry, 303n4; wage parity, 469; competition and, 447­48; proposal for, 200­202, 469 311­12; work-related and academic qualifications, 312 Internet marketplace, 315, 411, 412n11; classification system for labor rights, 465, 468, 470, 471n7 transactions, 322; digital form of works and, 392­93, Laird, Sam, 97­104 401nn7­10; WTO work program on, 323 language barriers, 434b Internet service providers (ISPs), 315, 316 Latin America, 4, 6, 93, 317; Bananas case, 87­88b; cross-border Interventions, 28, 32; arguments for, 150; categories of, 150­51, mergers and acquisitions, 457, 458t; dispute settlement 152b; infant industries, 528, 534n4; for protection of genetic involvement, 76; growth in services trade, 238; mergers and resources and traditional knowledge, 384­85 acquisitions, 456; regulatory reform, 130; telecommunications intracorporate transferees, 307 industry, 289n2 intra-industry trade, 54 Latin American Integration Association (ALADI) Agreement, intrasectoral resources, 13 117­18, 119 inventions, novelty of, 396, 397, 401nn18­19 laws and legislation: electronic commerce, 319­21b; U.S. trade inverse elasticity rule, 529 laws, 206­12 investments, 439­40; in advanced technologies, 370; incentives least-developed countries (LDCs), 493; capacity-building efforts, for, 332; policies, 440, 442­45, 446n1; restrictions on, 356; 509­15; changes in export revenues, 112b; customs adminis- subregional agreements, 339, 345n4; underinvestment in tration, 495; dispute settlements, 75; market access, 94, innovation, 399. See also foreign direct investment (FDI) 106­7b, 111t, 112b; participation in WTO, 47; S&D treat- IP. See intellectual property (IP) ment, 504­8; tariffs and, 99­100, 105, 109, 111; transition IPRs. See intellectual property rights (IPRs) periods under TRIPS, 362; valuation system, 131 Ireland, 299b Lebanon, 371, 405 ISIC. See International Standard Industrial Classification (ISIC) Lengyel, Miguel F., 485­92 ISO. See International Organization for Standardization (ISO) Lerner symmetry theorem, 534n8 ISPs. See Internet service providers (ISPs) Lesotho, 301b ITA. See Information Technology Agreement (ITA) licensing, 352, 370, 393; Australia, 180; compulsory, 363, 365, Italian Copyright Law of 1941, 399 367b, 374­75b; import licensing procedures, 140; IPRs, ITO. See International Trade Organization (ITO) 375­76; IP, 405, 411n4; Korean policy, 356­57, 358n8; recog- Jamaica, 145­46b, 166, 502b nition of in trade agreements, 338; requirements for, 248, Japan, 6, 48, 363, 370; automobile industry, 185n7; bilateral 305, 440; service workers, 312­13 trade flows, 239; changes in export revenues, 111, 112b; com- Lisbon Agreement for the Protection of Appellations of Origin and petition laws, 452­53; GSP and, 109, 110t; industrial policy Their International Registration, 395, 401n16 instruments used in, 151, 155­56t; quarantine treatment, 436; list-or-lose technique, 336­37 tariff peaks, 105­6, 108­9, 111; technology transfer, 355­56; lobbying, by exporters, 15 telecommunications, 293­94b; trade policy review of, 45b local content policies, 179­85, 440, 446n2 Jaramillo group, 489­90b local presence requirement, 343 Joint Integrated Technical Assistance Programme for Selected locational services, 290­91 Least-Developed and Other African Countries, (JITAP), 510­15 locational subsidies, 439 joint ventures, technology transfer, 356­57 Lomé Convention, 106b judicial reviews of AC/CVD investigations, 210 Long-Term Textile Arrangement, 198 Luke, David F., 509­15 Kennedy Round, 33, 53­54, 59n6, 535, 536 luxury content, 534n3 Kenya, 301b, 430b Keynes, John Maynard, 5 macroeconomic policy, 521­22 Kheir-el-din, Hanaa, 186­94 macroeconomic stabilization, 5­6 knowledge, exploitation of, 399 Madagascar, 27b knowledge transfers, 352, 355 Madrid Agreement for the Repression of False or Deceptive Indica- Kono, Masamichi, 259­79 tions of Source on Goods, 395 Korea, 5, 363; antidumping regulations, 126; financial services lib- Madrid Convention, 408 eralization, 223­24b; government procurement, 424t; indus- majority-owned foreign affiliates (MOFAs), 242, 243t trial policy instruments used in, 151, 155­56t; net foreign Malawi, 301b exchange earnings, 166; SAROK, 424, 427n2; technology Malaysia, 173, 175, 309b transfer, 355­56; use of subsidies, 154 Mann, Catherine L., 315­25 Krueger, Anne O., 12, 13, 15, 31 Manual on Statistics of International Trade in Services (MSITS), 237, Krugman, Paul, 527 245b Kyoto Convention, 134­36, 141 manufacturing, 102t, 257t margin of preference, 418 labeling, 123, 394 marginal propensity to import, 417 labor force, 35, 228, 353; child labor, 467; foreign workers, margins, 252­53 287­88, 309b; higher-skilled workers, 228, 234n5, 307, 309b; maritime services, 216, 251, 277­78, 294, 303n1 labor flows, 311; labor market policies, 305; skilled workers, market access, 15­16, 62, 439, 448; agricultural industries, 635 D E V E L O P M E N T, T R A D E , A N D T H E W T O 103­4; air transport services, 279; binding tariffs and, 541­42, rates, 531; application of, 215­16; as benefit of WTO member- 545­47; commitments to, 43, 262­65, 280­89, 293; competi- ship, 62; exemptions under GATS, 260, 266; FDI restrictions tion policy and, 450b; concessions to Sub-Saharan Africa and, 442b; GATT tariff cuts and, 197, 204n1; included as prin- countries, 115­16b; economic needs test and, 313; environ- ciple in agreements, 341, 343, 346n6; maritime transport sec- mental protection issues, 477; expansion due to quota increas- tor, 277, 279nn4­5; movement of natural persons and, 305, es, 190t; financial services, 267b, 270­71t; intellectual 314n3; as necessity for services trade agreements, 337­38; property, 373­78; LDCs, 94, 106­7b, 112b; limitations to, safeguards application to, 326­27; tariffs and, 97­98, 99­100, 306, 314n4; restrictions on commercial presence, 281t; S&D 540 treatment and, 508; services trade, 249, 251, 281t; skilled movement of natural persons: broadening GATS framework on, labor and, 215; subregional agreements, 339­40; to European 311­13; commitment by sector, 308t; commitments by WTO Union, 106­7b; to LDCs, 94; under GATS, 218t, 260; the WTO members, 265f; constraints on, 304­5; GATS and, 305­10; and, 97­104. See also barriers to trade GATT safeguards and, 327, 332, 335n1; overview, 304, 3141; market economy, 5, 6, 67, 70n5 proposals for liberalization of, 310­13; services exports, market institutions, 35 286­87b market power, 405, 406, 457 MRAs. See mutual recognition agreements (MRAs) marketing boards, 27b MSIT. See Manual on Statistics on International Trade in Services marketing, trademarks, 412n7 (MSITS) markets, obstacles to operation of, 521­22 Multifibre Arrangement (MFA), 58, 186, 187, 188, 192­93, marks of origin, 125 194n1, 198 Marrakech Agreement, 46, 50­51, 510 Multilateral Agreement on Investment, 331 Martin, Will, 540­47 multilateral environmental agreements, 476 Maskus, Keith, 229b, 367b, 370, 377, 432, 464 Multilateral Initiative on Malaria, 366b matching grant programs, 161, 162­63b multilateral negotiations, 95, 97, 487, 488; domestic reforms and, Mattoo, Aaditya, 259­79, 280­89, 290­303 215­16, 218­19t; issues of IPRs, 399­400; trademark protec- Maurer, A., 135­46 tion, 410 Mauritius, 7­8, 162­63b, 166­67 multilateral trade agreements (MTAs), 174, 331, 444, 476 medical care: disease treatments, 366­67b; medicines and, multilateral trade rules, TRIMs and, 175­76 380­81b; pharmaceutical industries, 374­75b multimodal transport, 278 medicines: derived from plants, 383­84, 389n7; differential pric- multinational firms, 440 ing of, 378­79 multiple tariff bands, 533 Mercedes, 405 music industry, 390­91 Merck & Co., 366­67b, 382, 387­88, 388n2, 397 mutual recognition agreements (MRAs), 219t, 437 MERCOSUR. See Common Market of the South (MERCOSUR) Myanmar, 463, 465b, 471n2 mergers and acquisitions, 448, 449; cross-border, 457, 458t; noti- fication and review regimes, 457, 459­61; overview of 1990s Nairobi Convention, 133 trend, 456­57 national business plans, 524b Mexico, 173, 547, 453; agriculture, 35; automobile industry, 355; National Community Gene Fund, 388 imported tires, 433b; maquiladora sector, 355; NAFTA and 7, National Indigenous Arts Advocacy Association (NIAAA), 394 337; telecommunications, 293­94b national interest, 488 MFA. See Multifibre Arrangement (MFA) national resources, 442, 446n3 Michalopoulos, Constantine, 61­70, 519­25 national treatment, 278, 290, 296, 304, 337­38, 343 microfinance institutions, 301b nationalism, 417 Microsoft, 405 natural resource-based exports, 224, 234n3 Middle East, 371, 401, 550b natural resources, protection of, 478 Milpurrurru v. Indofurn Pty Ltd, 394 necessity test, 295b, 296­97 Ministry of International Trade and Investment (MITI), Japan, 356 negative list approach, 336­37 mobile cellular technology, 300 negotiations: electronic commerce and, 316­24; horizontal modes of delivery, 240­42, 287­88 approach to, 323; market access enhancement, 285­88; modes of services, 340 modalities of in Western Hemisphere, 336­37; movement of modes of supply, 304; commitments to, 308t; GATS agreement, natural persons, 310; multilateral, 95, 97; reciprocity, 43, 259, 261, 262, 264­65; telecommunications, 276­77, 279n3 50­52, 59n2; tariffs, 97­100, 101­3; under GATT, 50­52, MOFAs. See majority-owned foreign affiliates (MOFAs) 59n2 Mongolia, 15 net foreign exchange earnings, 166 monopolies, 222, 290­91, 297, 404; cost of for IPRs, 399; net interest margins, 251 monopoly suppliers, 339; prices and, 452b; regulating new generation trade agreements, 119, 120 monopoly prices, 374­75; telecommunication, 282, New Zealand, 85, 176, 363, 434b 289nn2­3; universal service and, 301­2 Ng, Francis, 105­13 monopsony power, 527 Nigeria, 76 Morocco, 353 nominal rate of protection, 151b most-favored-nation (MFN) status, 31, 53, 278; ad valorem tariff nonpaper, on bindings, 56, 60n11 636 Index nontariff barriers (NTBs), 24, 25, 103, 105­6, 520, 526; agricul- Philippines, 173, 175, 288 tural, 98, 105; as trade obstacles, 93 plant breeders' rights, 360b, 411n1 nonuniform tariffs, arguments for, 527­30 Plant Protection Convention, 434 North American Free Trade Agreement (NAFTA), 142, 443­44; plant resources, 383­84, 388n1, 388n4 domestic regulation and, 338, 345n2; government procure- plurilateral agreements, 67 ment, 339; Mexico's reforms and, 7; negative list approach to Poland, 299b trade liberalization, 337; regime for origin rules, 117, 118, Policy Sciences Center, 390, 393, 397, 401n13 119, 120; tariff refunds and, 90­91n6; textile products market, politics, 50; foreign control of services, 282; and reciprocity in 54 negotiations, 59n2; RIAs and, 552­53, 554­55; trade policy North American Trade Automation Prototype, 142 and, 32 North, Douglass C., 4 pollution, 474b Northern Triangle group, 337 port sector, 227b Norway, 189­92, 494 Portugal, 299b NTBs. See nontariff barriers (NTBs) positive list approach, 336, 337, 423 nullification of benefits, 154, 159n10 poverty, 28­30, 30b, 380­81b poverty alleviation, 2, 513 OECD countries, 445, 454; average tariff on imports, 101; govern- PPMs. See production and process methods (PPMs) ment procurement market, 417, 418; immigration patterns in, preferential access, 105, 109­11, 113n1, 114­21 309b; Industry Committee, 331 preferential trade agreements (PTAs), 175­76, 452­53 Office Internationale des Epizootic, 434 presence of natural persons, 221, 222, 236, 237t offshore activities, 161 preshipment export finance guarantee schemes, 167 OIE International Animal Health Code, 434b preshipment inspection (PSI), 132­33, 137­38n7, 140, 143­44b Olarreaga, Marcelo, 105­13 pressure valve actions, 197­200 oligopolies, 290­91, 440, 527 price preferences, 422, 426 open skies agreements, 279 price-based instruments, 248, 251 openness to trade, 3, 7, 8, 9­10, 10n1; growth and, 14, 31­32, price-control systems, 377­78 73n2; IPRs and, 373 price-fixing conspiracy, 457, 458t option value of resources, 384 prices, 11­12; acquisition of market power and, 457; agricultural Osaka Action Agenda, 142 products, 12; automobile industry, 183, 185n5; determination outgrowers, 35, 37n6 of, 300; differential pricing, 378­79b; discrimination in, 377, outsourcing, 192, 425b 378b; effect of procurement bans on, 421; imports relative to Oyejide, T. Admola, 483, 504­8 exports, 12­13; internal, 29b; maritime trade, 303n1; monop- olies and, 452b; regulation of, 297, 303n2; stability of, Pacific Economic Cooperation Council (PECC), 249, 258n3 through bindings, 546­47 Pack, Howard, 351, 372 Prieto, Javier Francisco, 336­46 packaging, 434b privacy issue, electronic commerce, 292b Pakistan, 77, 78, 175, 479, 555 private sector, trade facilitation and, 142­43 Panagariya, Arvind, 535­39 privatization, 4, 214, 230, 300; electricity distribution, 303; of Pangestu, Mari, 149­59 state-owned enterprises, 417­18; telecommunications, 302b parallel trade, 363­64, 365, 376­78, 407 process standards, 428 Paris Convention, 395, 408, 498 procurement practices, 417­21, 427n1 participation issues, 485 product development, 370­71 partner countries, 552 production: in-house, 182, 185n2; methods of, 434; services used pasta, 431 in, 226; transition periods, 33 patents, 360b, 400, 411n1; abuse of, 452; enforcement of, 405; production goods, poverty and, 32­33 global demand for, 407, 408f; granting of, 387; increase in use production and process methods (PPMs), 468, 476 of, 407­8; infringements of, 207, 347; IPR standards and, productive factors, 14, 15 497­98, 503n2; Japanese system of, 370; notification require- productivity, 358n2, 420­21; automobile industry, 180­81, ment for patentability, 396; pharmaceutical industries, 185n1; effects of foreign involvement in, 353­54, 358n6; 374­75b; plant resources, 383, 388n4; protection of, 362­64, technology, 370 366b, 375, 388, 389n14, 404, 411n2; relationship to R&D, protection: of automobile industry, 181­85; nominal and effective 370; revocation of, 396; traditional knowledge, 397 rates of, 151b peanuts, 433b protectionism, 15, 417, 465, 467­69 PECC. See Pacific Economic Cooperation Council (PECC) PTAs. See preferential trade agreements (PTAs) Penna, Frank J., 390­402 public goods, 384 performers, protection of recorded performances, 392­93, public policy, genetic resources and traditional knowledge, 401n10 384­85 PERINORM database, 431 public utilities, 222, 228, 230, 231 Peru, 73b, 302b Punta del Este declaration, 490b pharmaceutical industries, 366­67b, 374­75b, 378­79b, 389n13 Pursell, Garry, 179­85, 355 637 D E V E L O P M E N T, T R A D E , A N D T H E W T O rules of origin, 111­12, 114­21, 122­27, 140, 191, 192 Quad economies. See Canada; European Union (EU); Japan; Unit- Russia, 4, 62, 70n1 ed States quality assurance, 371 safeguard actions, 52, 59n4, 85, 95, 103; emergency measures, quantitative-restriction (QR)-type policies, 3, 247­48, 251­52, 326­30; EMS in GATS, 328­29, 334­35; EU, 68, 107b; GATT 338, 344 and, 195­200, 204nn1­6, 327­28; guidelines for procedures, quotas, 24, 211: automobile industry, 182; Bananas case, 87­88b; 204; implementation of policies, 202­4; industrial policy and, elimination of, 103; market access and, 190t; movement of 154; injury investigation, 203b, 204; overview, 195; subre- natural persons and, 287b; textiles and clothing, 186, 187­92, gional agreements and, 339, 345n3; under ATC, 188­89 194nn3­4; under GATS, 261; yarn, 192, 194n3 safety nets, 34­35 safety valves, 44 Radio Shack, 355 Saggi, Kamal, 351­58, 439­46 real exchange rate, 2, 19, 20, 22, 164, 523b Salmon case, 72 real income, 527, 534n2 sanctions, 465 reciprocal tariff liberalization, formula approaches to, 535­39 sanitary and phytosanitary (SPS) measures, 140, 429, 430b, reciprocity, 43, 54, 337, 488; cross-sectoral, 288, 289n6; defini- 434­37, 473, 495­97, 499 tion, 50, 59n1; domestic, 54; in GATT rules, 50­53; as influ- Sauvé, Pierre, 326­35 ence on trade agreements, 53­55; procurement markets, 418; Scallops case, 73b under GATT/WTO negotiations, 285, 289n5 Schiff, Maurice, 12, 13, 15, 519­25, 548­58 recognition requirements, 305, 312­13, 314n8 SCM. See Agreement on Subsidies and Countervailing Measures redundancy payments, 34, 37n5 (SCM) reforms, 14, 33; challenges to implementation of, 493­94; com- Schuler, Philip, 126, 131, 137n5, 493­503 plementary policies, 519, 521­23; costs of, 496t; credibility of, sea turtles, 478 523­25b; criteria for evaluation of, 3­4; customs administra- Seattle Ministerial Conference, 75b, 473, 488, 512, 514­15 tion, 128­38, 495, 496t; poverty issues, 28­30; regulatory, secondary rules of origin, 125­26 230­31; timing and sequencing of, 522; views of, 3; in WTO sectoral agreements, 341, 342t, 345n5, 536­37, 539 dispute settlement mechanisms, 83­90, 90n3 security, 318, 554­55 Rege, Vinod, 128­38 Senegal, 22, 23n6 regional content value calculation, 119­20 serious prejudice, 154, 159n10 regional integration agreements (RIAs), 443­44, 518; economic service marks, 360b aspects of, 550­54; listing, 549­50b; North-South versus services, government spending on, 418 South-South, 553­54, 558n3; overview, 548­49, 558n1; polit- services trade. See trade in services ical aspects of, 554­55; WTO rules and, 555­56, 558n4 SERVIMED, 286b regional trade agreements, 327, 487; approaches to trade in serv- Shatz, Howard J., 17­23, 164 ices, 336­46; gap between bound and applied rates in, 99­100 Shea, Eleanor, 206­12 regional trade policy, 512 shipping services, 247, 277 regionalism, 524b, 548, 556­57 Shrimp case, 473, 478­79 regulatory policies, benefits of, 429 shrimp imports, 478 renegotiations, 52, 197­98, 199, 200t Singapore, 309b, 363 research and development (R&D), 224, 351, 358; China, 370; dis- Singapore ministerial conference, 512 ease treatments, 366­67b; implications of services barriers for, Smarzynska, Beata K., 403­12 256, 258; patents and, 370; share of in multinational firms, smuggling, 531 352; technology, 372­73 social issues, 465; gains from GPA membership, 426b; link residual rules of origin, 125­26 between trade and labor rights and, 469­70, 471n6; telecom- resource allocation, 12­13, 14, 532, 533, 534n9 munications markets, 302b; value of innovations in IPRs, 364; restructuring industries, 528 value of social programs, 380­81b retaliation: against United States, 82; Bananas case, 88, 89­90, software industry, 224, 286b 91n17; forms of, 86­87, 91n14; as GATT safeguard, 52­53; in sound recordings, 392­93 TRIPS, 77; under WTO, 72 South Africa, 364, 451, 453; AIDS-controlling drugs, 366­67b; revenues, 2; customs, 131, 137n6; from digitizable media prod- electronic commerce, 317; financial sector, 232, 234n5; phar- ucts, 320b; from exports, 111; from tariffs, 537; impact of maceutical industries, 374­75b; privatization in, 230 trade liberalization on, 24­26; trade reform effects on, 29b; Spain, 299b, 431 trade taxes, 529 special and differential (S&D) treatment, 73­74, 79­80n3, 157, RIAs. See regional integration agreements (RIAs) 158, 484, 504­8; new approach to, 485; TRIMs agreement, rice, duties on, 107b 174, 178n6 rights of indigenous peoples, 393 special prosecutors, 79 risk assessment, 430b species extinction rate, 382­83 Rodrik, Dani, 3­10, 31, 467b spillovers, 4; anticompetitive practices and, 448; due to regional Rome Convention, 498 integration, 443­44; environmental concerns and, 474b; from royalties, 352, 388, 390­91, 397, 400n3 FDI, 353­54, 358n5; international, 442­43; labor rights and, 638 Index 465, 467; poverty issues and, 29b; technology, 356, 358n5; technology, 351, 536; industrial policy and, 150; use in customs wage inequality and, 466b administration, 146 SPS. See sanitary and phytosanitary measures (SPS) technology transfer, 397; channels for diffusing, 352­55, Standard Industrial Classification, 431 358nn1­4; costs of, 372; facilitation of, 379­80; government standards, 428; conformity assessment, 429; developing countries policies related to, 355­57; IPRs and, 369­70; mode of, issues with, 436­37; enquiry points and notifications, 434­35; 356­57; vertical, 352­53, 354­55 future of, 438b; impact of meeting, 431; implementation of, telecommunications, 251, 318, 451; competition in, 189nn2­3, 429­30; labor, 463; merger review issues, 460­61; problems 233­34b, 281­82, 293­94b; developing country precommit- encountered by Indian exporters, 433­34b; product certifica- ments to liberalize, 284­85t; employment in, 303; existence of tion, 429; role of, 429­31; SPS agreement, 495­97; standards- natural monopoly in, 291; India, 289n3; interconnection fees, developing bodies, 428­29. See also core labor standards 294b; Latin America, 289n2; procompetitive regulation, Staples, Brian Rankin, 139­48 297­98b; Reference Paper, 297; social objectives in competi- steel, 434b tive markets, 302B; under GATS, 272­77, 279n3 Stephenson, Sherry M., 336­46 temporary admission scheme, 163­65, 166, 179n2, 533 Stern, Robert M., 247­58, 251, 256 tendering procedures, 419, 425b Stotsky, Janet, 24­27 Texas Instruments, 355 strategic trade theory, 527­28 textiles and clothing sector, 103, 371, 405, 558n4. See also Agree- structural issues, 231, 307, 309­10 ment on Textiles and Clothing (ATC) Sub-Saharan Africa: copyright protection and, 371; health care Textiles Monitoring Body (TMB), 189, 191 system, 381b; market access concessions to, 115­16b TFP. See total factor productivity (TFP) Subramanian, Arvind, 76­77, 382­89 Thailand, 5, 173 subsidies, 103­4, 154, 159nn8­10, 165, 207, 219t; categories of, tiered pricing, 378­79b 168­69b; cross-subsidization, 300, 331; effect on movement of tires, 433b natural persons, 306, 314n5; GATS disciplines on, 331­32; pro- TMB. See Textiles Monitoring Body (TMB) duction, 534n4; WTO and, 168­69b. See also export subsidies Tokyo Round: customs valuation, 128, 129, 130; green-room Subsidies and Countervailing Measures model, 72 meetings, 48; S&D treatment, 505 sugar, duties on, 107b total factor productivity (TFP), 353, 369, 370, 373 sustainable development, 472 tourism, 224, 231 Sweden, textile and clothing restrictions, 189­92 Toyota, 181­82, 183 Switzerland, 102 TPOs. See trade promotion organizations (TPOs) trade agreements: collective nature of, 55; locking in, 55; noneco- Taiwan, 166, 355­56 nomic objectives influence on, 54­55; PTAs, 175­76, 452­53; Tanzania, 134 reciprocity influence on, 53­55; regional, 99­100, 327, tariff of abominations, 555 336­46, 487, 555; subregional, 339, 345n3. See also names of Tariff Act of 1930, 211 specific agreements Tariff Board, Australia, 180, 184, 185n1 trade deficit, 18 tariff equivalents, 56, 249, 250t, 252 trade deflection, 114­15 tariff rate quotas (TRQs), 102­3, 107b trade diversion, 551, 556 tariffs, 3, 25, 93, 211, 519, 520; anti-export bias of, 533, 534n10; trade facilitation, 139­48 arguments for, 527­30; automobile industry, 185n9; classifica- trade law, versus environmental law, 475 tion changes, 116­17; effective protection for certain sectors, trade laws, U.S., 206­12 538, 539n3; import-weighted, 97­98, 99; imported cars, 181; Trade Marks Act 1995, 401n11 industrial policy and, 154; on manufactures, 100; negotiation trade policy: design lessons, 26; external, 552­53; growth and, issues, 97­100, 101­3, 529­30; peaks, 105, 108­9, 110t, 111, 9­10; institutional reform and, 7­8; liberal regime elements, 113n2; rate structure, 102, 104n1; reduction of, 55­59, 519­21; politics and, 32; regionalism and, 556­57; status quo 59nn7­8, 60n2, 537; refunds of, 85, 90­91n6, 91nn9­12; roll of, 520 back, 538; simplification of, 533; unbound, by economy, 541, Trade Policy Review Mechanism, 45b 543­44t; uniform, 527, 530­33; U.S. reduction approach, trade preferences. See Generalized System of Preferences (GSP) 538; WTO membership and, 67. See also applied tariffs; bind- trade promotion organizations (TPOs), 160­63 ing tariffs trade reform. See reform Tarr, David G., 17­23, 164, 519­25, 531 trade regulations, publication of, 44 taxation: agricultural industries, 11­12; commercial presence and, trade secrets, 361b, 397, 398, 401­2nn20­23 283; export taxes, 26, 27b; foreign ownership limits and, 282, trade in services, 45, 221­31, 239, 520; approaches to in Western 283, 289n4; trade facilitation and, 147; trade taxes, 25­26 Hemisphere, 338­39; classification of service providers, Tay, Simon, 463­71 310­11; commitment patterns, 261t, 262­65, 307, 308t, TBT. See technical barriers to trade (TBT) 321b; customs duties and, 319­21b; domestic regulations technical assistance, 51, 134, 491­92 and, 290­303; emergency safeguards, 326­30; exports of technical barriers to trade (TBT), 140, 434, 436, 437, 476, 506 services, 240t; foreign affiliates, 240, 241, 242­43, 246n4; technical protection measures, 392, 401n7 gains from liberalization of, 225­26, 227, 230, 234n6; GATS, technical regulations, 428, 431­32, 434 overview of agreement, 259­62; GATT safeguards paradigm, 639 D E V E L O P M E N T, T R A D E , A N D T H E W T O 327­28; government procurement, 333­34; growth of, Uganda, 301b 238­39, 246n2; higher-level personnel in, 307, 309b; impor- U.K. Patents Act of 1977, 401n19 tance of individual sectors, 239­40; improving data on, U.N. Commission on International Trade Law (UNCITRAL), 142 244­45b; managing liberalization process, 232, 234; market U.N. Conference on Trade and Development (UNCTAD), 134, access, 281t; matrix of elements in agreements, 343f; meas- 141­42, 248, 509 urement of barriers to, 249­53; mode of delivery, 240­41, U.N. Development Assistance Framework (UNDAF), 513 242t; modes of supply. See specific mode of supply; reform in, U.N. Development Programme (UNDP), 469 213­17; regional liberalization of, 336­46; role of reciprocity U.N. Electronic Data Interchange for Administration, Commerce, in, 285, 287­88, 289nn5­7; safeguard actions for, 328­29; and Transport (EDIFACT), 142 sectoral treatment in agreements, 340­41, 342t; share of, U.N. Statistical Commission, 237 239t; statistical measurement of, 235, 237­38; structure and U.N. System of National Accounts, 245b content of regional agreements, 341­45; types of barriers to, unbound tariffs, by economy, 541, 543­44t 247­48; welfare effects of reduction in trade barriers, 257t unemployment, 30b, 37n4 Trade-Related Aspects of Intellectual Property Rights (TRIPS) agree- unfair competition, 394­95, 400 ment, 7, 47, 157, 158, 347, 403­4; artistic works and, 391; unfair trading practices, 402n24 cross-retaliation under, 87, 88­90; developing countries strategy uniform tariff, 527, 530­33 to change, 477; disputes under, 172­73; geographic indica- unilateral liberalization, 56 tions, 395­96, 401n16; implementation of, 77, 347­48, unilateralism, 467­69 359­68; key provisions of, 359­62; overview, 497­98, 499t, United Kingdom, 363, 431 503nn1­4; pharmaceutical industries and, 374­75b; reform of, United Nations, trade facilitation agencies in, 141­42 367­68; review process for, 473; social issues involved with, United States, 5, 6, 277, 322, 363, 454; antidumping measures 348; summary of selected issues of, 349t; violations of, 76, 80n5 and, 73­74b, 199; bilateral trade flows, 239; changes in Trade-Related Investment Measures (TRIMs): agreement, 95, 154, export revenues, 111, 112b; cross-sectoral reciprocity, 288; 157, 321; approaches to review of, 176­77; implementation depression, 21t; dispute settlements and, 365; extension of of, 172­76, 177n2, 177-78n5; overview, 171­72; structural MFN to Russia, 62, 70n1; FATS data, 242; GSP and, 109, 110t; adjustment as dispute defense, 175, 178nn9­10 information technology sector, 324; origin rules implementa- trade-supporting services, 514 tion, 126; privacy, electronic commerce, 292b; requests for trademarks, 127n1, 360b, 363, 395, 400; application costs, 407, retaliation against, 82; Shared Border Accord, 142; tariffs and, 412n7; enforcement of, 371, 372, 405, 409; global demand 53, 85, 91nn9­11, 105­6, 108­9, 111, 113n2; telecommuni- for, 407, 408f; increase in use of, 407­8; infringements of, cations, 293­94b; textile and clothing industry, 54, 189­92; 207, 371; international agreements on, 408­11; overview, trade laws, 206­12; trade policy review of, 45b; trade in serv- 403­4; parallel imports of, 377; protection of, 404­8; registra- ices data, 243t; TRIPS cases, 76; waste regulations, 436; WTO tions, 407­8, 409f, 412n8 and, 48, 70n1 trading partners, effects of trade liberalization on, 535­36 United States International Trade Commission 432 traditional knowledge, 382­86, 389nn9­11, 396 universal service, 222, 234n1, 282, 300­303 transactions costs, 139, 448 Uruguay, overvaluation of exchange rate, 19­20 transition economies, market access commitments, 267b Uruguay Round, 45, 98, 483, 486, 493; AD/CVD safeguards, 85, transition periods: adequacy of, 173; production during, 33; S&D 91n12; adjustment periods for liberalization, 33; Antidumping concessions, 507; textile and clothing industry, 186, 188­89; Agreement, 202, 204n3, 205n4; dispute settlement system, under TRIMs, 172, 173, 174, 175, 176; under TRIPS, 362 83­84; green-room meetings, 48; lessons learned, 499­500; transitional agreements, 187 preshipment inspections, 132b; S&D treatment history, 505­8; transparency, 43­44, 45, 45b, 344; environmental regulations safeguard actions, 52; single undertaking approach of, 485; and, 472; requirements under GPA, 425­26b; of technical reg- tariff peaks and, 105­6; tariff reductions lessons, 55­59, 59n7, ulations, 434 59n10; technical assistance provision, 51; textiles and clothing Transparency International, 425b trade, 186; TRIMs and, 175, 178n6 transport initiatives, 141 U.S. International Antitrust Enforcement Act, 449 transport systems, 225­26b, 241t U.S. International Trade Commission, 393 travel services, 239­40, 241t U.S. Trade Act of 1974, 402n24 TRIMs. See Trade-Related Investment Measures (TRIMs) U.S. Trade Adjustment Assistance Program, 37 TRIPS. See Trade-Related Aspects of Intellectual Property Rights U.S. Trade and Development Act of 2000, 115­16b (TRIPS) Agreement U.S. Trade Representative (USTR), 208, 402n24 tropical rain forests, 382, 387 TRQs. See tariff rate quotas (TRQs) value added criteria, 117, 126, 193n Tuna case, 473 value-added tax (VAT), 147 Tunisia, 229b VERs. See voluntary export restraints (VERs) Turkey, 20­21, 223b, 531, 558n4 vertical technology transfer, 352­53, 354­55 Tussie, Diana, 485­92 Vienna Convention on the Interpretation of Treaties, 463 Tuthill, Lee, 259­79 visas, 285, 286b, 311, 312 two-track multilateral trading system, 484 Visser, Coenraad J., 390­402 typewriters, 211­12n1 voluntary export restraints (VERs), 198t, 199, 200t, 205n7, 205n9 640 Index World Intellectual Property Organization (WIPO), 388, 392­93, wages, 29b, 228, 311­12, 466b 395, 396, 398, 407 Washington Treaty, 498 World Trade Organization (WTO), 44, 48, 187, 477, 491; acces- Washington Treaty on Integrated Circuits, 361 sion process, 62­69, 70nn2­5; accessions to as of May 1, Watal, Jayashree, 76­77, 359­68, 410 2001, 64t; ACWL and, 75b; costs of implementing agree- WCO. See World Customs Organization (WCO) ments, 502b; decisionmaking in, 48, 491­92; dispute settle- Webb-Pomerene Act, 455n4 ment system, 71­80, 83­84, 90n3; electronic commerce on weighted-average tariff equivalents, 249 agenda, 316, 318, 321­24; FDI rules, 441­42b; governance welfare, 536 of, 488, 490­91; historical review of, 44­46, 47t; market welfare costs, 319b access under, 97­104; membership, 7, 61­62, 510, 529­30; welfare effects, 531; of discriminatory procurement, 334; of RIAs, nonmember denial of benefits, 338; overview, 39­40, 140; 551, 558n2; of tariff bindings, 545­47 reciprocity as principle in negotiations, 285, 289n5; rules of, welfare-enhancing policies, 13, 16n1, 94, 229b, 377, 440 3­4, 10n1, 103­4, 104n3, 158, 555­56, 558n4; scope, func- welfare-reducing policies, 444­45 tions, and structure of, 41­42, 46­48; services schedules, Western Hemisphere, approaches to liberalization of services trade 262­65; Subsidies Agreement, 208; subsidy rules, 168­169b; in, 336­46 tariffs as negotiating tool for membership, 529­30; trade and WHO World Health Report 2000, 380b environment in, 472­75; trade-competition relationship, Wilson, J. S., 431, 432 447­55; TRIMs notification and, 172, 177n2; use of dispute Winters, L. Alan, 28­37, 50­60, 556 settlement procedures, 365. See also names of specific agree- WIPO. See World Intellectual Property Organization (WIPO) ments under, for example, Agreement on Rules of Origin Working Party on Professional Services, 295­96b World Bank, 509; country classification system, 507­8; customs yarn quotas, 192, 194n3 reform, 495, 496t; IPR projects, 498, 499t; mainstream trade efforts, 513, 514; SPS-related experience, 496­97 Zambia, 34, 35, 37n3 World Customs Organization (WCO), 123, 128, 140, 141, 145 Zimbabwe, 34, 35­36, 37n6, 76 641