EAST ASIA Integrates A TRADE POLICY AGENDA FOR SHARED GROWTH Editors Kathie Krumm and Homi Kharas east Asia integrates A Trade Policy Agenda for Shared Growth east Asia integrates A Trade Policy Agenda for Shared Growth Kathie Krumm and Homi Kharas, Editors A copublication of the World Bank and Oxford University Press © 2004 The International Bank for Reconstruction and Development / The World Bank 1818 H Street, NW Washington, DC 20433 Telephone 202-473-1000 Internet www.worldbank.org E-mail feedback@worldbank.org All rights reserved. 1 2 3 4 06 05 04 A co-publication of the World Bank and Oxford University Press. The findings, interpretations, and conclusions expressed here are those of the author(s) and do not necessarily 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 information 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 Copy- right 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, fax 202-522-2422, e-mail pub- rights@worldbank.org. ISBN 0-8213-5514-7 Cover photo: World Bank Library of Congress Cataloguing-in-Publication data has been applied for. CONTENTS Foreword xi Acknowledgments xiii Overview xv Kathie Krumm Homi Kharas PART I WIDENING OPPORTUNITIES IN TRADE ARRANGEMENTS 1 1. China's Accession to the WTO: Impacts on China 3 William J. Martin Deepak Bhattasali Shantong Li 2. Regional Impact of China's Accession to the WTO 21 Elena Ianchovichina Sethaput Suthiwart-Narueput Min Zhao vi East Asia Integrates 3. New Regionalism: Options for East Asia 39 Mari Pangestu Sudarshan Gooptu 4. Market Access Barriers and Poverty in Developing East Asia 59 Bijit Bora with additional material by Paul Brenton and Takako Ikezuki PART II DEVELOPMENT ORIENTATION FOR A BEHIND-THE-BORDER AGENDA 75 5. Trade and Logistics: An East Asian Perspective 77 Robin Carruthers Jitendra N. Bajpai David Hummels 6. Protecting Industrial Inventions, Authors' Rights, and Traditional Knowledge: Relevance, Lessons, and Unresolved Issues 95 Manjula Luthria Keith E. Maskus 7. Trade and Competitiveness Aspects of Environmental and Labor Standards in East Asia 115 Keith E. Maskus PART III REINFORCING SOCIAL STABILITY THROUGH BROAD SHARING OF BENEFITS 135 8. Household Welfare Impacts of China's Accession to the WTO 137 Shaohua Chen Martin Ravallion 9. Trade in Sectors Important to the Poor: Rice in Cambodia and Vietnam and Cashmere in Mongolia 149 Jehan Arulpragasam Francesco Goletti Tamar Manuelyan Atinc Vera Songwe 10. Trade and Labor Market Vulnerability in Indonesia, Republic of Korea, and Thailand 171 François Bourguignon Chor-ching Goh Authors and Their Affiliations 189 Index 191 Contents vii FIGURES Figure 1 Trends in Regional Trade xviii Figure 2 Intraregional Trade xix Figure 3 China's Trade in Third-Country Markets xx Figure 4 Reduced Agricultural Protection in China xxi Figure 5 Chinese Imports, 1995 and 2005 xxii Figure 6 Regional Trade Arrangements xxvi Figure 7 Inland Transport Costs for Remote Regions xxix Figure 8 Expanded Exports to APEC from Improvements in Trade Facilitation xxx Figure 9 Costs of Producer Services xxx Figure 10 Regulatory and Competitive Reforms, Selected Countries xxxii Figure 11 Geographic Distribution of Bilateral Trade Treaties xxxiii Figure 12 Environmental and Labor Standards, East Asia xxxiv Figure 13 Poverty and Inequality xxxvi Figure 14 Effects of China's Accession to WTO on Household Incomes xxxvii Figure 15 Trade and Stability in Labor Markets, Republic of Korea xxxix Figure 2.1 Exports to China from East Asia Compared with Exports to Other Partners 22 Figure 2.2 China's Imports, 1995 and 2005 23 Figure 2.3 China's Growing Role in Production Networks 24 Figure 2.4 Impact of China's WTO Accession on Japan and East Asia's NIEs, 2001­10 26 Figure 2.5 Impact of China's WTO Accession on Developing Economies of East Asia, 2001­10 27 Figure 4.1 Real Income Gains in Developing East Asia from Agricultural Liberalization 60 Figure 5.1 Potential Contribution of Transport to Economic Growth in East Asia 79 Figure 5.2 Poverty and Accessibility in Lao PDR 82 Figure 6.1 Comparison of Patenting Trends, Republic of Korea and Major Developing Countries, 1978­99 97 Figure 6.2 Comparison of Korean Semiconductor Patents with Worldwide Semiconductor Patents, 1976­99 98 Figure 6.3 Comparison of Korean Semiconductor Patents with Semiconductor Patents of Major Advanced Countries, 1976­99 98 Figure 6.4 Citations of Korean Patents as a Proportion of Total Patent Citations, by Types of Patent Applicants, 1986­99 101 Figure 6.5 Self-Citation Rates of Korean Private Corporations, 1980­99 102 Figure 6.6 Korean Patents Granted in the United States, Selected Industries, 1981­99 102 Figure 8.1 Poverty Incidence Curves: Rural 141 Figure 8.2 Poverty Incidence Curves: Urban 142 Figure 8.3 Mean Gains by Provinces: Absolute Gains in Yuan Per Capita 143 Figure 8.4 Mean Gains by Provinces: Proportionate Gains in Percent 143 Figure 8.5 Mean Gains by Provinces: Percentage of Gainers 144 Figure 8.6 Mean Gains in Yuan, by Income Percentile 144 Figure 8.7 Mean Percentage Gain, by Income Percentile 145 Figure 8.8 Percentage of Gainers, by Income Percentile 145 Figure 9.1 Profit Shares for Rice Value Chain, Cambodia and Vietnam 155 Figure 9.2 Profit Structure of Rice Value Chain, Cambodia and Vietnam 161 Figure 9.3 Share of Cashmere Products in Mongolia's Total Cashmere Exports, 1993­2001 166 viii East Asia Integrates Figure 10.1 Decade Averages of Import Tariffs for Manufacturing Industries: Indonesia, Republic of Korea, and Thailand, 1980s and 1990s 174 Figure 10.2 Average Log of Real Wages of Males by Broad Sectors (Manufacturing, Services and Construction, and Others) and by Trade Exposure (Low, Medium, and High): Republic of Korea, 1976­2000 177 Figure 10.3 Average Log of Real Wages of Males by Broad Sectors (Manufacturing and Services) and by Trade Exposure (Low, Medium, and High): Thailand, 1991­2000 178 Figure 10.4 Average Log of Real Wages of Males by Broad Sectors (Manufacturing and Services): Indonesia, 1986­1999 178 Figure 10.5 Standard Deviation of Log of Real Wages of a Male Cohort in Low, Medium, and High Trade Exposure Industries: Republic of Korea, 1976­2000 179 Figure 10.7 Average Monthly Hours Worked of a Male Cohort in Manufacturing and Services: Thailand, 1991­2000 180 Figure 10.6 Distribution of Employment Status and Employment Sectors of a Male Cohort: Republic of Korea, 1985­2000 180 Figure 10.8 Estimated by Birth Year Cohort: Republic of Korea, 1929­69 183 Figure 10.9 Vulnerability by Sectors and Trade Exposure Groups: Republic of Korea, 1976­2000 184 Figure 10.10 Vulnerability by Gender and Educational Attainment: Thailand, 1991­2000 185 TABLES Table 1 Effects on Economic Welfare of Various Regional Trade Proposals xxiv Table 2 Structure of Manufactured Exports by Country, 1985 and 1996 xxix Table 1.1 Some Measures of Import Protection in China's Agriculture 8 Table 1.2 Protection of Industrial Sectors in China 10 Table 1.3 Impacts of Reduction in Protection Required by WTO Accession from 2001 Tariff Levels, China 14 Table 2.1 Weighted Average Tariffs Facing Exports to China: Indonesia and Thailand, 2001­08 28 Table 2.2 Shares of Exports to China Affected by Lifting of Quantitative Restrictions (QRs): Indonesia and Thailand 28 Table 2.3 Market-by-Market and Product-by-Product Analysis of Indonesian Exports: China's Market Share and Closeness of Unit Value (UV) as Proxies for Potential Risk 29 Table 2.4 Market-by-Market and Product-by-Product Analysis of Thai Exports: China's Market Share and Closeness of Unit Value (UV) as Proxies for Potential Risk 30 Table 2.5 Share of Exports to China Affected by Lifting of Quantitative Restrictions (QRs): Cambodia and Lao PDR 33 Table 2.6 Market-by-Market and Product-by-Product Analysis of Cambodian Exports: China Market Share and Closeness of Unit Value (UV) as Proxies for Potential Risk 34 Table 2.7 Market-by-Market and Product-by-Product Analysis of Lao PDR's Exports: China Market Share and Closeness of Unit Value (UV) as Proxies for Potential Risk 35 Table 3.1 Proposed and Actual Regional Trading Arrangements Involving East Asian Countries 42 Contents ix Table 4.1 Frequency of HS4 Lines in Top 30 Exports of Developing East Asia by Value 61 Table 4.2 Key Markets for East Asian Agricultural Exports 62 Table 4.3 Commitments and Average Direct Domestic Support Levels, 1995­98 64 Table 4.4 Simple Average MFN Applied Tariff Rates by MTN Category, Selected Asian Countries and Major Trading Partners 66 Table 4.5 Trade Policy Structure of European Union, Japan, and United States for Imports from East Asian Countries 68 Table 4.6 Tariff Escalation for Selected Products in Selected Markets and ASEAN Free Trade Area 69 Table 4.7 Nontariff Measures Affecting Products of Interest to Developing East Asia Economies 71 Table 5.1 Availability of Transport Infrastructure 79 Table 5.2 Composition of Logistics Costs of Container Transport from Inland China (Chongqing) to U.S. West Coast 84 Table 5.3 Container Movements at Selected East Asian Ports, 1995­2001 85 Table 5.4 Customs Clearance Times 87 Table 5.5 Electronic Data Interchange and Transport E-Commerce in Selected Countries 88 Table 5.6 Ranking of Major Freight Airports in East Asia, 2001 89 Table 6.1 Research Productivity of Selected Countries 100 Table 7.1 Ratifications of Fundamental ILO Conventions Covering Worker Rights, Selected East Asian Countries 117 Table 7.2 Overall Ranking of Labor Standards 119 Table 7.3 Impacts of Labor Standards on Labor­Intensive Exports from Developing Countries 122 Table 7.4 Environmental Policy Indexes 127 Table 7.5 Impacts of Environmental Standards on Pollution-Intensive Exports from Developing Countries 128 Table 8.1 Predicted Aggregate Impacts on Welfare of Rural and Urban Households 141 Table 9.1 Rice in Cambodia and Vietnam 153 Table 9.2 Land Size and Rice Profit, Cambodia and Vietnam 154 Table 9.3 Rice and Income of the Poor, Cambodia and Vietnam 154 Table 9.4 Strategic Options for Rice, Cambodia and Vietnam 158 Table 9.5 Effects of Various Policy Options on Rice Production, Income, and Exports: Cambodia and Vietnam 159 Table 9.6 Qualitative Analysis of Strategic Options, Cambodia 159 Table 9.7 Marketing Costs and Margins for Rice in Cambodia, 2002 160 Table 9.8 Marketing Costs and Margins for Export Rice in Vietnam, 2002 160 Table 9.9 Marketing Margins of Farmers in Cambodia, Vietnam, Thailand, and United States 161 Table 9.10 Cashmere in Mongolia 162 Table 10.1 Average Import Tariffs (Applied Duty Rates) and Percentage Change in Tariffs of Indonesia, Republic of Korea, and Thailand between 1980s and 1990s 173 Table 10.2 Core Nontariff Barrier Measures: Indonesia, Republic of Korea, and Thailand, 1989­99 175 Table 10.3 Decade Average Shares of Import Tax Revenue in Total Government Revenue and Total Trading: Indonesia, Republic of Korea, and Thailand, 1980­2000 175 x East Asia Integrates Table 10.4 Volatility of Annual Growth Rates in GDP and Value Added in Manufacturing, Measured by Coefficient of Variation: Indonesia, Republic of Korea, and Thailand, 1970­96 176 Table 10.5 Level and Fluctuation of Wage Growth Rates (Males) by Trade Exposure Groups: Republic of Korea, 1976­1997 176 Table 10.6 Estimated 2 Correlation of Variance of Residual Earnings by Education, Gender, and Sector: Indonesia 183 Table 10.7 Vulnerability by Education, Gender, and Sector: Indonesia, 2001 184 BOXES Box 3.1 Forms of Regional Preferential Trading Arrangements 40 Box 3.2 "Early Harvest" 52 Box 4.1 Global Trade in Rice 65 Box 5.1 Successful Integration of Ports and Land Transport Networks: Republic of Korea 88 Box 6.1 Protection of Indigenous Knowledge in the Philippines 111 Box 9.1 Value Added from Value Chains 152 FOREWORD This collection of studies focuses on one of the · The third force is the determination that East most important economic issues facing East Asian Asia should develop strategies to narrow the gap and Pacific nations today. We have titled this collec- between richer and poorer countries and tion East Asia Integrates: A Trade Policy Agenda for between urban and rural households to promote Shared Growth to capture three forces operating in regional stability and reduce poverty. the region. This is a very broad agenda, and, of course, no · First, a strong force for policy integration is single volume can tackle all the problems in depth. sweeping through the region, and it has gath- But in launching this work we felt it was important ered strength since the 1997­98 Asian financial to broaden the approach beyond the technical per- crisis. This new regionalism in East Asia is based spective of trade policy to emphasize development on a strategic re-rethinking of what needs to be outcomes and the links to social stability. Increas- done to bolster growth and stability across the ingly, the development agenda in the region--with region. its focus on growth, jobs, and social stability--and · Second, trade policy has gained prominence as a the trade policy agenda--with its focus on market central force, in part because trade within the access and competitiveness--have become inter- region has been one of the dynamos of growth twined. Economic integration is a growing force, in recent years, accounting for almost 90 percent and we want to ensure it is managed carefully so of East Asian export growth in 2002, and in part that its benefits can be shared broadly by poorer because of the historic accession of China to the countries and the poor within countries. World Trade Organization in November 2001. To be sure, East Asia is a very diverse region, and Other countries in the region will have to adapt our analysis shows that there is no one-size-fits-all their own national development strategies as approach. Yet the studies included here suggest China adjusts to its new commitments. several policy priorities for East Asia as well as xi xii East Asia Integrates directions for further empirical analysis that would efit everyone. Thus this volume contains studies on carry policymaking to a conclusive stage. It is agriculture, market access for the goods produced imperative that all countries have the capacity to by the poor, logistics to connect far-flung producers undertake detailed analysis on these priority with markets, environmental and labor standards, issues, because today there is an urgency to the income and employment vulnerability, and the policymaking process. Negotiating positions are effects on households of opening trade. A compre- now being developed for an array of regional and hensive approach is needed to link trade to devel- bilateral agreements and for the Doha Develop- opment and poverty reduction, and we urge policy- ment Agenda after Cancun. The stakes are high, so makers to develop such a comprehensive approach we must move from debate and academic discus- in their national development strategies, in their sions to the real world of negotiations and concrete regional and bilateral agreements, and in their policies. global negotiating positions. The World Bank has a long and proud tradition We hope you, the reader, will find the studies in of supporting the expansion of trade in East Asia this volume interesting and that in turn you will and around the world as a crucial tool in promot- contribute your voice to this critical debate. ing growth and development. But this support does not mean a blind adherence to lower tariffs and Jemal-ed-din Kassum reduced nontariff barriers. It does mean support- Vice President ing measures to ensure that the opportunities East Asia and Pacific Region afforded by trade are available to everyone and ben- World Bank ACKNOWLEDGMENTS We have benefited from the input and guidance of Narongchai Akrasanee, Ataman Aksoy, Myrna Austria, numerous colleagues and officials who have supported Choong Yong Ahn, Florian Alburo, Kanemi Ban, Dou- the analysis underpinning this volume. We would like glas Brooks, Siow Yue Chia, Lu Ding, Carsten Fink, to express our particular appreciation to the authors of Yujiro Hajami, Masahiro Kawai, Jeffrey Lewis, Mai Lu, the chapters and the background papers. Earlier drafts Aaditya Matoo, Deunden Nikomborirak, Hadi Soesas- of the chapters and the main findings were presented tro, Jomo Sundaram, Somkiat Tangkitvanich, Jose at seminars and workshops throughout East Asia, Tongzan, Shujiro Urata, Dominique Van der Mens- including Bangkok, Beijing, Hanoi, Jakarta, Phnom brugghe, Sangui Wang, John Wilson, Lan Xue, and Penh, Seoul, Singapore, Tokyo, Ulaanbaatar, and Vien- Ippei Yamazawa. tiane. We would like to thank the organizers, including The views expressed in this book are those of the the Institute for Southeast Asian Studies at the authors and should not be attributed to any particular National University of Singapore, the Development institution, including institutions with which individ- Research Center and Tsinghua University in China, the ual authors may be associated, including the World Thailand Development and Research Institute, and the Bank and the World Trade Organization. World Bank Institute, as well as the staff of the country offices of the World Bank. We also benefited from interaction with scholars and policymakers at the Fourth Asia Development Forum on Trade and Poverty held in Seoul and organized jointly with the Korean Institute for International Economic Policy, the Korea Development Institute, and the Asia Develop- ment Bank. Valuable comments and contributions were provided, and we would like to thank all of those who participated in these seminars as moderators, dis- cussants, and presenters. Special thanks go to Overview Kathie Krumm Homi Kharas Emerging East Asian economies1 have seen their Against this background, some analysts estimate share of world exports more than triple during the that East Asia would still benefit more than any past quarter-century, from 5.4 percent in 1975 to other region from global liberalization (World 19.8 percent in 2002. Their trade with one another Bank 2002c)--because of the potential shown by has grown faster than their trade with any other its dynamic exporters--and that it could achieve market and now makes up 7.2 percent of global much of the benefits of liberalization through trade.2 Broad measures of development in East Asia regional integration--because of the wide scope of have improved at the same headlong pace; since intraregional trade among what are very diverse 1990 more than 300 million people have seen their economies. The potential gains to the region from incomes rise above a poverty threshold of US$2.00 global liberalization are estimated at hundreds of a day.3 billions of U.S. dollars by 2015. Increasingly, the But performance has varied sharply over time development agenda in the region, with its focus on and across the region. Most recent trends show growth, jobs, and social stability, and the trade pol- mildly rising inequality within countries, including icy agenda, with its focus on market access and China, Vietnam, and the Philippines, and a widen- competitiveness, have become intertwined. ing income gap between richer economies such as Pursuit of a "trade for development" strategy in Singapore and Hong Kong (China) and the poor- East Asia will not be easy. It involves economic est, Lao People's Democratic Republic (Lao PDR) change at a time when policymakers are trying to and Cambodia.4 Both trends are of concern to pol- manage major financial and corporate restructur- icymakers interested in a stable, prosperous region, ing, repair overstretched social safety nets, respond and both trends have been linked with economic to the challenges posed by China, listen to the liberalization. greater plurality of voices and interest groups com- xv xvi East Asia Integrates peting for political power, and adapt to sometimes The region is now shaping its responses to these vicious global economic cycles. Given that stability two events. The financial crisis has forced policy- is seen as a key to investor confidence and a return makers to rethink the financial, trade, and invest- to high growth, it is not surprising that there is ment linkages that connect regional economies. At resistance to still further change. the same time, it has given new impetus to ideas for regional institutions to help shape common responses. East Asian countries had historically Why Push Integration Now? pursued unilateral and nonpreferential approaches Two economic events of historic proportions pro- to liberalization, sealed through commitments vide the context for East Asian development today. made under WTO agreements. Indeed, through the The first event, the currency and financial crisis crisis market access was largely preserved. But of 1997­98, still affects every aspect of economic equally, significant gaps in the international archi- policymaking. The crisis shattered a deeply held tecture were exposed, in finance, investment flows, and broadly shared view of the policy keys to devel- and macroeconomic (especially exchange rate) opment success, and called for a review of what coordination. now needs to be done to build on the basic tenets of On the financial front of this new regionalism, macroeconomic stability, high savings and invest- the Chiang Mai initiative has created a web of swap ment, rapid expansion of education, and a strong agreements between regional central banks to be export orientation.5 Many in the region consider deployed in case of future liquidity problems. The the crisis to have been a "defining moment" (World Asian Bond Fund concept has been endorsed by Bank 2002b). several countries. Informal economic cooperation The second major event was China's accession to forums have discussed regional currency coopera- the World Trade Organization (WTO) in November tion and a host of other economic and noneco- 2001--the culmination of a 15-year-long bid for nomic issues. membership. China's economy is already one of the In this context, the trade policy agenda for inte- largest in the world, and China's prominence as an gration--the subject of this volume--is only one economic force in the region has clearly risen since element of a broader set of policy and institutional the crisis.6 China's strong currency and strong changes taking place in the region. Although growth, set against the weakening currencies and regional trade negotiations have not advanced as far weak growth of other developing countries in the as those governing financial arrangements, many region, have meant that China's gross domestic analysts consider them to be more fundamental and product (GDP) now surpasses that of the rest of the foundation for regional economic cooperation. emerging East Asia combined.7 China was the Success will ultimately depend on the consistency of world's largest recipient of foreign direct investment approaches among all elements of integration. To in 2002. Moreover, with its trade growing rapidly, achieve such consistency, the scope of discussion China has become an important destination for must be broadened. To this end, the integration exports from the rest of the region, and it is a fierce agenda discussed in this volume covers three areas: competitor in third-country markets.8 Because of its size and diversity, China resists easy definitions of its 1. Widening opportunities, globally and regionally, comparative advantage in terms of high-tech or low- in trade arrangements wage manufacturing, and now that it has acceded to 2. A behind-the-border agenda to maximize the the WTO its comparative advantage may change development impact of international commit- appreciably. Thus it is no surprise that many policy- ments makers in the region feel the need to understand the 3. Broad sharing of benefits to reinforce social sta- trends at work in China before they can formulate bility and a long-term political commitment to their own development strategies (World Bank further reforms. 2002b). Typically, too, they recognize that overcom- ing the formidable challenges of implementing These areas are discussed in sections of this China's WTO commitments will be as important for overview, and the concluding section outlines pri- the rest of East Asia as it is for China itself. ority areas for action. Overview xvii Context: Recent Trends in Trade reflected in the growth of trade in components or partly assembled goods. China is increasingly a Rapid and sustained growth in international trade central player in such production networks, both as has long been a hallmark of successful growth and a final assembler of products and as an efficient development strategies in East Asia. Some success provider of components. Although Japan remains stories are well known: those of the newly industri- an important center of production-sharing opera- alizing economies (NIEs) such as the Republic of tions in East Asia and is the origin of about one- Korea, as well as middle-income economies such as third of all regional exports of components for Malaysia and the transition economy of China. assembly, China is finding niches; its exports of More recent entrants to world markets that have parts and components grew by almost $20 billion seen rapid export growth include low-income during 1996­2001. At the same time, other East economies such as Cambodia and Vietnam. Trade Asian economies have been finding their own has been an important factor in growth in the niches within China's markets and increasing their region, enabling progress in poverty reduction. shares of China's imports. Although the 1997­98 financial crisis interrupted Meanwhile, East Asian trade profiles have this progress, recovery since then has brought become more mutually complementary, and the poverty rates in every emerging economy in the degree of complementarity for emerging East Asia region to record lows, and in economies like that of is now comparable to that of the EU and NAFTA.10 Vietnam, trade growth has brought with it a rapid The exceptions are low-income countries such reduction in poverty. as Cambodia, Lao PDR, and, to a lesser extent, Viet- Intraregional trade in East Asia has grown faster nam. This finding is perhaps not surprising, given than trade with any other market, and while the that a country's trade structure becomes more like largest economies account for the bulk of this the world's trade structure as the country's income trade, the regional trade of most smaller economies rises. The wide gap between rich and poor coun- has also grown. Trade integration has been market- tries in East Asia is also largely responsible for a rise led, stemming from a combination of unilateral in regional income inequality (see Milanovic 2003 reforms, fulfillment of multilateral commitments, and the discussion later in this chapter). Policies and a pattern of relocation of production processes and institutions that contribute to fuller integra- (see Kawai and Urata 2002). Intraregional trade has tion of the low-income countries and regions of been driven not only by growing demand but East Asia will be critical to addressing income increasingly by improved competitiveness in inequality in East Asia and contributing to a stable regional markets, as reflected in increased market region. shares (Figure 1). China has been particularly dynamic, but almost all countries increased their Widening Opportunities in Trade competitiveness in regional markets during Arrangements 1995­2001.9 This increase was accomplished with- out loss of competitiveness in other markets; East Today, the region's policymakers see trade and Asia continued to expand its market shares in the investment integration as key elements of strategy. European Union (EU) and North American Free Although a few trade barriers were erected in Trade Agreement (NAFTA) markets in the same response to the Asian economic crisis, they were period. minor and temporary.11 Indeed, instead of more The product composition of intraregional trade protection, the crisis produced a new impetus for has become highly concentrated. The top 30 openness. A multitrack approach is being followed, exports now account for more than half of total with support for a new global round of trade nego- trade, with 38 percent in just four subsectors: office tiations as well as for regional and bilateral arrange- machinery, telecommunications equipment, elec- ments. tronics, and textiles and clothing (Figure 2). Liberalization of trade and investment policies is One of the major factors in the concentration of recognized as a way to spur gains in efficiency. As intraregional trade has been the remarkable such, it fits well with the new emphasis in Asia on increase in international production networks, as innovation in firms and on economy-wide produc- xviii East Asia Integrates FIGURE 1 Trends in Regional Trade Emerging East Asia economies' share of world exports Intraregional trade is driven increasingly by grew...and trade among these economies grew even competitiveness and growing market share. faster. Intraregional trade: share of export Emerging East Asia: share of world trade change accounted for by competitiveness and demand changes Percent Percent 20 100 15 80 10 60 40 5 20 0 0 1980 1982 1984 1986 1988 1990 1992 1994 1996 1998 2000 85­95 95­01 85­95 95­01 85­95 95­01 Emerging East Asia intraregional trade With Within Emerging East emerging emerging Asia--Japan East Asia East Asia Competitive factora Demand factorb Note: Emerging East Asia is ASEAN plus other newly Note: Diversification, or change in exports due to new industrializing economies plus China plus Mongolia. products, accounts for a small share. Source: IMF, Direction of Trade Statistics (various years); a. Competitive factor = change in market share. data reported from exporter country accounts. b. Demand factor = change in market size. Source: Ng and Yeats (2003: Table 9.1). Exports to China are particularly dynamic, with East Asia finding niches in China's markets. East Asia NIEs: export markets Developing East Asia: export markets (as % of total exports) (as % of total exports) Percent Percent 40 40 35 35 30 30 25 25 20 20 15 15 10 10 5 5 0 0 1980 1982 1984 1986 1988 1990 1992 1994 1996 1998 2000 2002 1980 1982 1984 1986 1988 1990 1992 1994 1996 1998 2000 2002 China Other East Asia China Other East Asia Japan United States Japan United States Note: East Asia NIEs refers to Hong Kong (China), Repub- Note: Developing East Asia refers to Indonesia, Malaysia, lic of Korea, Singapore, Taiwan (China). Philippines, Thailand, Cambodia, Lao PDR, Mongolia, Sources: IMF, Direction of Trade Statistics (various years); Myanmar, Vietnam. data reported from exporter country accounts. Sources: IMF, Direction of Trade Statistics (various years); data reported from exporter country accounts. tivity growth to replace the old model of growth domestic-oriented firms, with productivity differ- through physical and human capital accumulation entials of 40 percent in Indonesia and the Philip- (Yusuf and others 2003). There is ample evidence pines and 15­20 percent in Thailand and Korea that export-oriented firms and those with foreign (Hallward-Driemeier, Iarossi, and Sokoloff 2002). equity participation are far more productive than Moreover, China's restructuring provides an impe- Overview xix FIGURE 2 Intraregional Trade Intraregional trade is increasingly concentrated...and production sharing is a major determinant. Share of 2001 intraregional trade accounted for Share of 2001 five-digit SITC parts and by 30 largest four-digit SITC exports components in top 30 exports Office machinery 6% Parts and . components Telecom Others in top 30 equipment 49% 28% 5% Electrical machinery 21% Textiles and Other top Others clothing 30 72% 6% 13% Note: SITC = Standard International Trade Classification. Note: In addition, parts and components in the non­top Source: Ng and Yeats (2003: Table 12.1). 30 exports account for an additional US$8.3 billion or 4 percent of total non­top 30 exports. Source: Ng and Yeats (2003: Table 17.1). China is an increasingly central player in production networks. Share of components in China's imports Share of components in China's exports relative to share of components in world trade relative to share of components in world trade RCA RCA 2.5 2.0 2.0 1.5 1.5 1.0 1.0 0.5 0.5 0 Office Telecom Electrical Office Telecom Electrical machinery equipment machinery machinery equipment machinery 1996 2001 Note: Revealed comparative advantage (RCA) measures a country's relative export performance in a good j. It is defined as the country's share of world exports of good j (xij/xwj) divided by its share of total world exports (Xi/Xw), where xij and xwj represent the value of good j exported by country i and the world, respectively, and Xi and Xw are country i's and the world's total exports, respectively. If the value of the index exceeds unity, the country is said to have a comparative advantage in the production of good j. Source: Ng and Yeats (2003). tus for establishing new regional production net- institutional arrangements. The willingness of the works, initially in electronics, that would promote world's richer countries to target the needs of highly productive firms. For now, these production developing countries in the context of the Doha networks are oriented toward exports to developed Development Agenda is helpful, but the outcome of countries, but the growing markets within East Asia these negotiations is still uncertain. And regional provide a potential complementary source of and bilateral agreements fit in well with other polit- demand.12 ical aims of East Asian countries, giving these Policies can support these natural economic agreements a momentum that the global discus- forces, and the drive to liberalize is shaping new sions have yet to achieve. ASEAN has moved to a xx East Asia Integrates FIGURE 3 China's Trade in Third-Country Markets This trade is expanding...but not at East Asia's expense. Share of U.S. imports Share of Japanese imports Percent Percent 20 10 15 10 5 5 0 0 1969 1972 1975 1978 1981 1984 1987 1990 1993 1996 1999 2002 1969 1972 1975 1978 1981 1984 1987 1990 1993 1996 1999 2002 China Other developing East Asia Source: IMF, Direction of Trade Statistics (various years.) two-tier system to permit its more advanced mem- What WTO Accession Means for China: Sectoral bers to accelerate trade reform, while launching Shifts. As explained in Chapter 1, assessing the new initiatives to integrate its less-developed mem- implications of major economy-wide reform such bers--Cambodia, Lao PDR, Myanmar, and Viet- as that involved in China's WTO accession is inher- nam. At the same time, the speed with which the ently difficult. Much of the impact will consist of ASEAN-China Framework Agreement has been reinforcing the structural changes already under launched and the ideas for a new ASEAN-Japan are way. The experience with other trade events such as creating new alliances in favor of liberalization the formation of NAFTA suggests that the actual within the region. As Philippines president Gloria overall impact will be much greater than suggested Arroyo noted in a speech of May 2002, combining by ex ante modeling, largely as a result of produc- the Association of Southeast Asian Nations tivity and other changes that are hard for models to (ASEAN) and Chinese economies "would give capture. birth to a market of 1.8 billion consumers or almost In agriculture, the initial levels of effective pro- one third of humanity." tection are in fact likely to be far lower than others have estimated, implying that sectoral shifts are likely to be less dramatic than others are predict- China's Role in the Region ing.13 Nonetheless, China's agricultural trade liber- In line with its own reforms, China has been a alization plan is the expected source of about half growing economic force in East Asia. Exports to of the efficiency gains from the WTO accession China have been dynamic over the last decade, par- package. Reduced protection and greater scope for ticularly for the newly industrializing economies imports are likely in a range of products, including but also for the rest of developing East Asia (Figure oilseeds, sugar, and dairy products (where the prin- 1), while shares to the United States and Japan have cipal form of protection has been tariffs, which are been flat or declining. China's trade with major being reduced substantially) and cotton (where markets also has been expanding, but not at the export subsidies are now ruled out) (Figure 4). And expense of developing East Asia (Figure 3). demand will sharply increase for other raw materi- Membership in the WTO and the reforms asso- als and natural resources such as wood and energy ciated with membership give China an opportunity products. to play an even larger and growing role in the world China's decision to contain agricultural protec- economy, and in the regional economy, amplifying tion has more far-reaching implications.14 This use preexisting trends. of WTO rules as a commitment device will focus Overview xxi FIGURE 4 Reduced Agricultural Protection in China China is following a strategy of reducing agricultural protection...albeit from a lower starting point than commonly measured. Measures of import protection Percent 160 120 80 40 0 ­40 Wheat Oilseeds Maize Rice Dairy Livestock Sugar Vegetables Cotton and meat and fruits 1998 Statutory tariffs 2001 Actual protection Postaccession protection Sources: Chapter 1. Average statutory rates taken from Schmidhuber (2001) and www.chinavista.com. See Huang and Rozelle (2002) for estimates of protection. policymakers on other types of policies that are favored-nation (MFN) tariffs on China. Analysis needed to deal with the very real problems of rural indicates that China's clothing production could poverty--such as improving rural education and more than double and its textile production could reducing barriers to labor mobility. At both the expand by nearly 50 percent as a result of WTO regional and global levels, China's commitment to a accession combined with the proposed phase-out low-protection agricultural regime sets the stage of quotas as part of the Agreement on Textiles and for its interests to coincide with those of countries Clothing (ATC).16 that are calling for opening the large and currently These estimates, however, omit some important highly protected markets for labor-intensive agri- potential elements of trade policy--namely, the cultural exports. application of antidumping and safeguards mea- In industrial products, the accession agreement sures against China and the possibility that China builds on the substantial liberalization already itself will increasingly apply such measures. One undertaken. The 6 percentage point reduction in feature of the accession agreement is the product- average tariffs that remains to be implemented is specific transitional safeguard provisions, lasting small relative to the 33 percentage points achieved over a 12-year period, which may be applied to over the last decade; the number of products sub- China by any WTO Member and may then trigger ject to import licenses has already fallen from about actions against the diversion of Chinese exports to two-thirds of tariff lines to less than one-twentieth other markets, with special textile safeguards for (see Ianchovichina and Martin 2002 and Lardy three years. These provisions are particularly trou- 2002, respectively.) blesome: no such measure targeted specifically at In the case of automobiles, a massive sectoral China existed before China's accession.17 Similarly, rationalization is expected to lead to scale in the area of antidumping provisions China could economies that could more than offset the impact remain vulnerable for up to 15 years to nonmarket of the reduction in protection, making China's economy provisions that dramatically increase the automobile industry internationally competitive.15 probability of dumping being found and the rele- Another major impact of accession will be a vant antidumping duties being applied. China's strong expansion in China's textile and clothing own increased use of antidumping and safeguard sectors. Importing countries have committed to measures would be legal and consistent with the abolishing, by 2005, the quotas on textiles and rules-focused approach to WTO implementation. clothing originally imposed under the Multi-fiber But one hopes such scenarios can be avoided Arrangement (MFA), and the United States and because the scenarios would be inconsistent with other countries have agreed to impose most- the emphasis on development that has character- xxii East Asia Integrates ized China's trade reform agenda since the begin- productivity in the services sector from expanded ning of the reform era. competition and foreign entry, for example, implies China's trade in services was the subject of one of welfare gains of $10 billion and a real GDP increase the most radical services reforms ever negotiated in of 2.2 percent, equivalent to the total estimated the WTO. The potential for rapidly increasing trade static gain from China's accession to the WTO (Ian- in services is significant. The important feature of chovichina and Walmsley 2002). China's commitments is that they focus on market access and do not discriminate between domestic What WTO Accession Means for the Rest of East and foreign suppliers. With China expected to be Asia: On Balance, More Opportunity. The fore- the largest market for telecommunications in the most opportunity for the rest of East Asia arising world by 2010, its commitments in that sector are from China's WTO accession lies in expanding profound in that they allow foreign entry for a wide exports to China's growing and more open mar- range of activities. One area of services with an kets; China's imports are expected to grow from 3 important bearing on trade is the activities that percent of global GDP in 2000 to more than 6 per- make up the logistical chain. China's WTO com- cent by 2005 across an array of commodities (Fig- mitments promise increased competition and ure 5). China receives 55 percent of its imports reduced costs in several areas, including through from East Asia and 37 percent from emerging East the development of third-party logistical firms. Yet Asia. All groups of countries in East Asia will bene- restrictions remain in place on the form of estab- fit from China's more open markets, but evidence lishments, such as requirements for joint ventures, suggests the scope is especially large for China's and on geographic scope.18 major trading partners of Japan and the newly industrializing economies, but will still be signifi- What WTO Accession Means for China: Ways of cant for the middle-income countries and the Doing Business. The implications for China of lowest-income countries, as explained in Chapter 2. accession to the WTO go beyond the specific com- Indeed, the sectoral landscape just described mitments in its accession package. Changes in the suggests significant export opportunities for devel- ways of doing business--elimination of dual pric- oping East Asian economies. For countries such as ing, phasing out of restrictions on trading, more Indonesia and Thailand, the weighted average tariff uniform administrative arrangements, and judicial on their top 100 export products falls by a half as review--are likely to result in dynamic gains that the result of China's accession, and some of their will outweigh those based on comparative-static top agricultural and raw material exports--rubber, estimates of efficiency gains. Every 1 percent gain in sugar, urea, and to a lesser extent rice--will benefit FIGURE 5 Chinese Imports, 1995 and 2005 What will China import more of? China: Percent Change in Imports Between 1995 and 2005 Foodgrains Feedgrains Oilseeds Meat and livestock Dairy Other food Beverages/tobacco times 10 Extractive industries Textiles Wearing apparel, net Wood and paper Metals Electronics Other manufactures 0 100 200 300 400 500 600 700 Percent Without accession With accession Source: Ianchovichina and Martin (2001). Overview xxiii from a reduction in nontariff barriers. These coun- the expansion of China's presence in major mar- tries also may see increased interindustry trade and kets will not necessarily crowd out other develop- production sharing as China's electronics and other ing East Asian exports, although it may constrain manufacturing industries, for example, grow. their growth prospects to levels below those of the For the lowest-income countries such as Cam- late 1980s. bodia and Lao PDR, there also will be opportuni- Nonetheless, certain exports appear at risk. ties. The decline in these countries' weighted aver- Clearly, the apparel and textile sectors in other East age tariff for the top 100 exports to China is quite Asian countries will face additional competition in small (from 4.3 to 2.6 percent for Cambodia and markets that are currently constrained by quota from 7.3 to 6.0 percent for Lao PDR), and only restrictions. The market-by-market and product-by- about 10 percent of Cambodia's exports and less product analysis in Chapter 2 indicates that for Thai- than 1 percent of Lao PDR's exports, including land, for example, only about 15 percent of exports wood and rubber, will face fewer quantitative to the United States are at risk from increased com- restrictions. However, given the trade intensity with petition from China, but that 25 percent of exports China, they will benefit from China as a powerful to Japan are at risk. For Indonesia, the second and source of external demand.19 third most important exports to the United States-- China's accession also offers the rest of the footwear and video-recording and -reproducing region the opportunity to import cheaper pro- apparatuses--face direct competition with China, as duced goods. Most obviously, consumers of Chi- do about a quarter of exports to Japan. nese finished good imports will benefit. But many Among lower-income countries, Cambodia is producers will also gain by access to cheaper inter- particularly vulnerable because its exports are mediate inputs into production processes. For highly concentrated in apparel. Roughly 30 percent example, chemicals already represent nearly 10 per- of Cambodia's exports to the United States have cent of imports from China in Indonesia and Thai- unit values similar to those of Chinese exports, land. By 2001 China was exporting more than $20 indicating direct competition, or are in categories billion in parts and components to other parts of in which China is currently quota-constrained, emerging East Asia, representing up to 20 percent indicating that competition from China is likely to of those countries' parts and components trade. increase. This prospect highlights the need for Continued growth in imports of parts and compo- urgency; such countries must speed up their gover- nents from China will represent an opportunity for nance and other reforms to ensure the develop- the rest of emerging East Asia to develop competi- ment of alternative products. tive firms. Nonetheless, on balance, China represents more To benefit from these opportunities, it will be of an opportunity than a threat to the rest of East important that East Asian countries resist the grow- Asia. Certain challenges will need to be met, and ing protectionist pressures for imposition of exces- adjustments made. However, the scope for gains sive safeguard measures in a futile effort to protect from a dynamic China is large across the range of all domestic producers. In Thailand, local manufac- countries in East Asia. turers are complaining about low-cost imports of electrical appliances and motorbikes from China. Regional Interdependence In 2002 Indonesia imposed temporary safeguard measures for garment imports, and Vietnam did so Recent trends in the magnitude and composition of for motorbikes. Such decisions can only prolong intra­East Asian trade bode well for further inte- the shifts in production that are necessary to realize gration (also see Ng and Yeats 2003). Comparisons regional comparative advantages, and distract poli- show that conditions within East Asia are now very cymakers from facilitating the adjustment of work- similar to those in countries that were previously ers through appropriate labor market and safety able to implement such successful regional net policies and programs. arrangements as the European Union and NAFTA. China's WTO entry also implies increased com- East Asia has achieved a high degree of market- petition in third-country markets. If recent Chi- driven regional integration, but regionalism--more nese export performance (Figure 3) is an indicator, formal economic cooperation and economic inte- xxiv East Asia Integrates gration arrangements and agreements between include trade facilitation measures, such as con- countries--has traditionally been quite limited, as formity of standards and procedures across explained in Chapter 3.20The major exceptions have national boundaries, and trade in services. The been the ASEAN Free Trade Area (AFTA), estab- New-Age Partnership between Singapore and Japan lished in 1993, and dialogue under the Asia-Pacific announced in January 2002 is notable in this Economic Cooperation (APEC) forum. However, regard. the past several years have seen a plethora of pro- The countries in the region could reap worth- posals for new bilateral and regional trade arrange- while economic benefits from increased regional- ments: the Framework Agreement on ASEAN- ism aimed at removing border barriers, according China Comprehensive Economic Cooperation, to analysis using the standard models for evaluating signed in November 2002; an East Asia­wide free benefits from regional trading arrangements (Table trade agreement; a more recent proposal by Japan 1). As is typical in these models, the gains are for an ASEAN-plus-Japan free trade agreement; and shown to be greater the wider the country coverage bilateral agreements under negotiation, in particu- and the wider the sectoral coverage of the regional lar by Japan, Singapore, Korea, and Thailand. arrangements. Notably, the gains for ASEAN coun- tries are shown to be greater than those for the Scope for Gains from Regionalism. Addressing other countries in the region. Given that the lower- intraregional barriers can ensure continued and middle-income countries of ASEAN are likely dynamism in regional trade and investment to benefit less from China's WTO accession than flows--including in response to China's growing the newly industrializing economies, greater role. The stated motivation of several initiatives, regionalism in East Asia can help spread the gains including the ASEAN-China Framework Agree- across a wider set of economies.21 ment, is to take advantage of complementarities One danger with the current regional arrange- and build on existing strengths in order to make the ments is that they have extensive provisions for region collectively more efficient and competitive excluding sensitive sectors. Such exclusions could and thereby attract investment. The regional agree- substantially reduce the potential for welfare gains ments under consideration are increasingly com- from an agreement. Within AFTA, for example, the prehensive in scope, going beyond the removal of lack of progress in agriculture and other sensitive tariffs and nontariff barriers on trade in goods to sectors has been disappointing. To give some indi- TABLE 1 Effects on Economic Welfare of Various Regional Trade Proposals % of GDP (% of GDP excluding agricultural liberalization) Proposal ASEAN China Korea, Rep. of Japan USA China + Korea + Japan ­0.26 (­0.16) +0.1 (­0.2) +1.0 (+0.6) +0.1 (+0.2) +0.0 (+0.0) ASEAN ­ China +0.9 (+0.5) +0.0 (+0.1) ­0.1 (­0.1) +0.0 (+0.0) +0.0 (+0.0) ASEAN ­ Japan +1.1 (+0.2) ­0.1 (­0.1) ­0.2 (­0.1) +0.0 (+0.1) +0.0 (+0.0) ASEAN + 3 +1.5 (+0.6) +0.1 (­0.2) +1.1 (+0.8) +0.2 (+0.2) ­0.1 (+0.0) ASEAN + 3 + CER +1.3 (+0.6) +0.0 (­0.1) +1.1 (+0.9) +0.2 (+0.2) ­0.1 (+0.0) APEC liberalization (MFN) +0.7 +0.5 +0.7 +0.4 ­0.0 APEC preferential liberalization +0.8 +0.6 +0.9 +0.4 +0.0 Note: Calculations for ASEAN include only Indonesia, Malaysia, the Philippines, Singapore, Thailand, and Viet- nam. CER (Australia­New Zealand Closer Economics Relations Trade Agreement) includes Australia and New Zealand. Figures in parentheses refer to net welfare effects when agriculture is excluded. Sources: Scollay and Gilbert (2003) for free trade agreement proposals and Scollay and Gilbert (2001) for APEC liberalization proposals. Overview xxv cation of the importance of this point, some eco- tively low-tariff bloc already exists, and because nomic models suggest that if agriculture is included many of the region's economies, especially the more in the ASEAN-China or ASEAN + 3 proposals, the developed, are moving toward lower average tariffs. estimated welfare gains for ASEAN countries To generate momentum, measures can be taken roughly doubles (Table 1). before the parties agree on formal arrangements. The greatest welfare gains, however, are likely to The "Early Harvest" aspect of the ASEAN-China come from addressing the deeper integration Framework Agreement exemplifies this approach. agenda being proposed here. In particular, the ben- As part of the Early Harvest, the agreement provides efits from integrating trade in services and over- accelerated access for key products to China and coming technical barriers to trade are likely to be ASEAN markets (ahead of the MFN schedule that is many times those from reducing border barriers part of the WTO accession agreement). Also, China alone.22 Gains from liberalizing services trade are has extended most-favored-nation status to the low- felt not only within the services sector itself, but income non­WTO Member countries Lao PDR and also within other sectors because services are vital Cambodia. inputs in their production processes. As manufac- The challenge in pursuing regional integration turing in East Asia moves into a phase where partic- arrangements will be to avoid the "spaghetti bowl" ipating in global markets will become more effect of different rules and regulations associated dependent on services, it increasingly will be held with different agreements, which can add unneces- back by poor efficiency in the services sector, which sary administrative costs to firms doing business has been protected from competition. Services lib- across the region and result in a bloc-ed up world eralization in East Asia lags relative to that of other (Figure 6). Such costs can be especially high when a regions, and action to integrate trade in services host of bilateral agreements are put in place, with could help restore its global competitiveness. Simu- large numbers of side agreements that each permit lation results from China, consistent with findings some discriminatory treatment. Inconsistencies from other parts of the world, indicate that even between agreements with overlapping membership small productivity gains of 1 percent in services also can become a problem. The exchange of ideas outweigh those based on static border trade effi- on a regional basis can help avoid such difficulties. ciency gains alone.23 While, in theory, there are some reasons to expect The policy scope for enhancing efficiency in that uncoordinated bilateral and regional agree- services, as well as in other sensitive sectors such as ments build in incentives toward global free trade, agriculture, may well be wider initially in a regional this outcome depends crucially on maintaining a context than in a global one. In fact, expanding common set of principles for each agreement-- integration into these sectors is one of the main that is, focusing on trade facilitation measures; potential advantages of regional, as opposed to allowing the most liberal rules of origin; and gener- global, liberalization. Most regional partnerships ally contributing positively to multilateral trade lib- already in place or under discussion include the eralization. For example, allowing for accumula- services and agriculture sectors, whereas these sec- tion under rules of origin will be important to the tors have proven much harder to tackle in global facilitation of production networks in the region, talks. Because the liberalization of services may be and allowing for multiple rules for conferring ori- necessary for industries to benefit fully from the gin would make it easier for more firms to partici- removal of regional barriers to goods trade, it may pate in regional trade (see Brenton 2003b). be preferable to give the services sector a high pri- Unless countries tackle the sensitive sectors in ority from the start of any negotiations. regional trade arrangements--and ultimately at the global level--many of the potential economic ben- Making East Asian Arrangements a Stepping- efits alluded to earlier will not be realized. China's stone to Global Integration. Liberalization goals binding to low tariff levels as part of its agricultural can be well served by moving on multiple fronts. In policy therefore offers scope for the liberalization of the East Asian goods trade, traditional concerns that agriculture in a regional setting, which could set in a regional arrangement will merely divert, rather place a favorable political dynamic for more open than create, trade are less worrisome because a rela- agricultural sectors throughout East Asia. xxvi East Asia Integrates FIGURE 6 Regional Trade Arrangements Low-tariff bloc reduces trade diversion. Multidimensional nature of agreements compounds "spaghetti bowl" and impedes merging. Average tariff rates: comparison Number of articles of agreement: of MFN with preferential rates recent East Asia initiatives 12 200 10 150 8 6 100 4 2 50 0 0 AFTA AFTA Global Global AFTA ASEAN­ JSEPA MFN CEPT LDCs INDs China (2002) MFN MFN Note: AFTA = ASEAN Free Trade Area; MFN = most- Note: JSEPA = Japan-Singapore Economic Partnership favored nation; CEPT = common effective preferential tar- Agreement. iff; LDCs = less developed countries; INDs = industrial Source: www. aseansec.org and www.mofa.go.jp. countries. Source: World Bank (2001a) and www. aseansec.org. Another way to proceed would be to unbundle As discussed in Chapter 4, a consensus appears to proposed agreements into components for deeper be growing about the negative impact on develop- integration that could realistically be agreed on and ing countries of trade barriers on products that are would yield immediate results in terms of increased exported by the poor--particularly agricultural efficiency and development of intraindustry trade products and labor-intensive manufactures such as and production networks. In the case of China and textiles and clothing, which are the sectors that ASEAN, for example, two-way business transac- confront the greatest entry barriers in high-income tions are already proliferating, and--long before countries. One of the disappointments of the government policymakers began to act--busi- Uruguay Round was its failure to achieve greater nesses had recognized and responded to the China market access in these products. The Doha Devel- challenge. Such transactions could be facilitated opment Agenda launched in November 2001 has and enhanced by an agreement that focuses on the potential to make the world trading system trade and investment facilitation, such as customs. more conducive to development. It includes a large It also would be preferable to focus attention number of policy reforms that can and should be and the resources of policymakers on as few undertaken by high-income countries and that will regional arrangements as possible but to make generate significant benefits for ordinary people in them broad in scope and country coverage, espe- developing East Asia. cially for small economies that have limited capac- ity to evaluate the development implications of Trade in Textiles and Clothing. As noted earlier, these arrangements. the textiles and clothing sector poses a special chal- Finally, a further benefit from stronger regional lenge for East Asia. Many countries of the region cooperation could be to link issues of common have large garment export businesses employing interest to an effective multilateral stance, whether sizable numbers of low-skilled workers, many of in services liberalization or discipline on whom are female. These businesses will be threat- antidumping measures. ened by competition from China in the form of both a macroeconomic threat to the balance of pay- ments (from lost exports) and a social threat of lost Expanding Market Access Globally jobs and lower wage incomes for many poor house- One of the motivating factors behind the move to holds. regional integration in East Asia is disappointment Implementation of agreements for further trade with the pace of multilateral trade liberalization in liberalization in textiles and garments is critical to products of most interest to developing countries. social stability in East Asia. Under the Agreement Overview xxvii on Textiles and Clothing as part of the Uruguay port programs in Europe and the United States on Round, quota restrictions were to be gradually farmers in East Asia is limited; such domestic pro- abolished over a 10-year period (that is, no later grams are heavily concentrated on meat, dairy that January 2005) in three stages, and the remain- products, and cereals.26 Thus with the notable ing quotas were to be subject to a progressive exception of rice, East Asian farmers are not increase in their rates of expansion. However, directly engaged in these distorted world markets. implementation to date has been disappointing. But other barriers in international agricultural Acceleration of the quota phase-out under the ATC markets are important for East Asia, and four prac- would have a significant impact on poverty by sus- tices of the developed countries are especially taining or increasing employment in these indus- harmful. First, their use of specific rather than ad tries. It would help maintain export growth in low- valorem tariffs often leads to very high effective income East Asia, for example, permitting 20 protection, particularly on low-quality goods pro- percent more exports from Vietnam. It also would duced by the poorest countries. Products such as help protect workers in middle-income economies palm oil, rice, and sugar, as well as fish, crustaceans, such as Thailand, where it would affect an esti- and fruits and vegetables are the ones most mated 50,000 jobs for garment workers and 10,000 affected. Second, developed countries use tariffs jobs for textile workers, about 20,000 of whom are that cascade upward on goods such as coffee and from poor households. Finally, acceleration would vegetables as protective devices for their agropro- benefit workers in the textile and clothing sector in cessing industries. Third, complex rules of origin China--workers who are mainly women drawn make it hard for countries to avail themselves of all from poor, remote regions--and could result in 5 the incentives theoretically made available to them. million fewer poor Chinese overall.24 Fourth, conformance with health and safety regula- At the same time, some economies in East Asia tions, notably maximum pesticide residue levels, will lose. Garment industries in several countries and difficulties with understanding and adminis- have depended on quotas, benefiting from some of tering standards are costly for many exporting the rents accruing from preferential access to EU countries. and U.S. markets. Garment exports from higher- Agricultural trade barriers are not just an issue income economies such as Korea are likely to for developed countries; the emerging economies decline. And in Cambodia, in particular, ATC of East Asia also maintain high barriers. Regional implementation is likely to put pressure on agricultural markets already are large. China's com- employment and wages in the garment sector. mitment to a regime of low protection in agricul- Cambodia will have to pursue a broader agenda to ture is fortunate for other countries that are well open up other trade opportunities. placed to supply it with imports, providing a direct boost to their foreign exchange earnings and to the Agricultural and Rural-Based Trade. Reduction of incomes of poor households. The ASEAN-China trade barriers in agriculture is particularly impor- trade agreement, with its "Early Harvest" provi- tant for poverty reduction. Agriculture is the main sions, provides another avenue to ensure that poor livelihood for poor households: for more than 90 agricultural households have every opportunity to percent in Cambodia and Vietnam, more than 75 benefit from trade opportunities in the region. percent in Indonesia, and roughly 70 percent in There may be further scope to address the sensitive Thailand and the Philippines (World Bank 2002a). food security and safety issues initially within the Gains from agricultural liberalization by higher- context of East Asian arrangements. income countries are estimated at more than $10 billion for developing East Asia, and that figure Development Orientation for a triples once dynamic productivity gains are taken Behind-the-Border Agenda into account.25 In contrast to the move toward increased transparency in trade policy for manu- As WTO director general Supachai Panitchpakdi facturers, agriculture remains protected in myriad eloquently stated in November 2002, "Trade poli- ways, as laid out in Chapter 4. However, the direct cies do not stand alone. Mutually supportive com- impact of agricultural subsidies and domestic sup- panion policies are also necessary." Many of the xxviii East Asia Integrates mechanisms required to maximize the develop- The WTO rules in this area aim at ensuring that mental benefits of trade and investment liberaliza- technical regulations, voluntary standards, and tion are "behind-the-border" institutional and reg- testing and certification of products do not consti- ulatory reforms. Each of the new regional and tute unnecessary barriers to trade. Under the Sani- bilateral partnerships reflects this business reality. tary and Phytosanitary Standards (SPS) Agree- Technical barriers and product standards, logistics, ment, WTO Members are encouraged to adopt services sectors, intellectual property rights, com- internationally recognized standards, but also are petition policy, and environmental and labor stan- free to apply stricter ones. Although the agreement dards have become as important a part of the pri- recognizes importing countries' right to implement vate sector environment as trade policy itself. These SPS measures that diverge from international "behind-the-border" issues also are featured norms, it does require them to provide scientific prominently on the Doha Development Agenda. justification for such measures and applies risk The thrust is to place trade policy within a broader assessment mechanisms. framework of development and poverty reduction. In many cases, East Asian economies will want to adopt unilaterally the international standards being set.28 China is further strengthening its food Technical Barriers and Standards: Balancing Trade safety regulations and capacity to implement these with Consumer Interests regulations as part of its WTO accession and is As the traditional barriers to market access decline, adopting various authentications and labeling sys- measures aimed at trade facilitation become more tems. Compliance with SPS in export markets can critical. Conformity with health and safety stan- induce changes in production systems and supply dards can be one of the most decisive determinants channels, but it is likely to require modernization of of access.27 Many of the standards in use represent standards infrastructure as well as information and legitimate consumer interests, but the use of tech- training. Countries may need to seek technical nical regulations, such as standards, has risen to assistance and support for additional projects in further commercial policy in multilateral, regional, this area. A Standards and Trade Development and global trade. An important problem for Facility to address this need is being established, to exporters is that different national governments be administered by WTO in collaboration with the often apply different technical specifications to World Bank, the World Health Organization meet the same goal. (WHO), the UN Food and Agriculture Organiza- For East Asian developing countries, importing tion (FAO), Codex Alimentarius, and others. countries' sanitary and phytosanitary (SPS) stan- At the same time, East Asian countries can bene- dards can impose significant additional costs that fit from rationalizing standards in a regional and impede exports. For example, it has been estimated global context. Cooperation on standards is one of that application of the strictest EU standard rather the important issues generally encompassed in than the international Codex Alimentarius standard regional arrangements. Recognizing that foreign for aflatoxin has cost Thailand $350 million in standards can achieve the same level of social or export losses in cereals and dried fruit, China $380 consumer protection as domestic standards, million, and Vietnam $15 million. The adoption of mutual recognition agreements (MRAs) are one a Codex standard on tetracycline could double option. Such agreements have mainly been used Thailand and China's beef exports (Otsuki and Wil- between developed countries, but ASEAN currently son 2001, 2002). Exporting countries also have to has signed two MRAs, with plans for an additional contend with importers' use of precautionary mea- 20 products, thus avoiding duplication of testing sures, such as the EU import ban on genetically and conformity assessment. modified organisms (GMOs), which hurts Thai- land's tuna exports as well as exports from China, Transport and Logistics which is the third largest GMO producer in the world. Only some of these standards may be justi- The commodity mix in East Asia is changing rap- fied on grounds of consumer safety and preferences; idly from resource-based commodities to low- and others may simply be protectionism in disguise. medium-technology goods to high-technology Overview xxix goods (Table 2). Logistics improvements, essential ers in many parts of East Asia can reach the U.S. for moving up the value chain, have very high pay- market more cheaply than inland areas or neigh- offs in East Asia, as discussed in Chapter 5. High- boring countries. For example, the inland transport value agriculture (flowers, fruits, seafood) and costs of moving goods from some remote regions manufacturing (electronics) demand sophistica- of China to external markets are roughly 10 times tion not only in production but also in logistics the inland transport costs at the other end (Figure handling. Timeliness matters, and a fast, reliable 7). Port logistics have been identified as a high pri- supply chain is essential. ority (Figure 8). Considerable scope exists for In East Asia, the key logistical bottlenecks seem reduction of transport cost margins (Figure 9). to be high internal land transport costs and port Modest improvements could lead to more than $50 logistics. This situation is in sharp contrast to exter- billion in additional exports from emerging East nal transport costs: with the sharp decline in trans- Asia to the rest of APEC, with an even larger impact Pacific shipping costs over the last decade, produc- on efficiency of imports (Wilson and others 2002). TABLE 2 Structure of Manufactured Exports by Country, 1985 and 1996 (% by value) Low-/medium- High- Resource-based technology technology 1985 1996 1985 1996 1985 1996 Hong Kong (China) 2.1 4.4 78.5 66.7 19.4 28.9 Singapore 42.3 12.7 25.4 21.9 32.3 65.4 Korea, Rep. of 7.8 9.4 72.1 55.0 20.1 35.6 Taiwan (China) 8.7 5.1 70.6 54.1 20.7 40.8 Indonesia 72.2 34.9 25.1 50.4 2.7 14.7 Malaysia 53.7 17.8 15.2 21.8 31.1 60.4 Thailand 42.1 14.5 44.8 49.1 13.1 36.4 China 11.7 9.8 78.9 69.7 9.4 20.5 Source: Lall (1998). FIGURE 7 Inland Transport Costs for Remote Regions Logistics costs of container transport from Chongqing, China to the U.S. West Coast: breakdown of total cost of US$3,650 per 20-foot equivalent unit (TEU) Port to final Port handling destination (receiving) 7% 4% Maritime transport 21% Land access to port 63% Port handling (sending) 5% Source: Carruthers and Bajpai (2002). xxx East Asia Integrates FIGURE 8 Expanded Exports to APEC from Improvements in Trade Facilitation Better logistics and harmonized standards are central to expanding trade. Expansion in exports to APEC from improvements in trade facilitation (% change in manufactured, agriculture and raw material exports from base) 20 15 10 5 0 China Indonesia Korea, Malaysia Philippines Singapore Thailand Vietnam Rep. of Port logistics Standards harmonization E-business Transparency and professionalism Source: Wilson and others (2002). FIGURE 9 Costs of Producer Services Significant scope exists for reduction of transport costs... and telecommunication costs. Estimated transport cost margins for International calling price comparison exports of textiles to United States, for selected Asian countries (US$ per using Singapore as a benchmark, 2000 three minutes), 2002 Percent 20 3.5 3.0 15 2.5 10 2.0 1.5 5 1.0 0.5 0 0.0 Indonesia Malaysia China China Indonesia Korea, Malaysia Singapore Rep. of Source: Bureau of Economic Statistics, U.S. Department of Sources: International Telecommunications Union and Commerce. operative Web sites. High inland transport and logistics costs are a impede the shipment of food commodities from particular impediment for the poor in remote areas surplus to deficit areas, worsening food security. or landlocked countries. In Lao PDR, poor inland Logistics requirements become more onerous transport contributes to a 25 percent differential in for trade in differentiated products than for trade paddy prices for farmers. On the island of Mindanao in homogenous products. Manufacturing firms, in the Philippines, agricultural producers are especially those integrated into global production plagued by poor access roads, lack of storage facili- chains, seek not only low transport costs but also a ties, and inadequate trading areas, leading to small host of sophisticated logistical needs: short transit volumes and high costs, compounded by interisland times, reliable delivery schedules, careful handling shipping fraught with high tariffs, product deterio- of goods in cold storage chains, certification of ration, and cargo losses. In Mongolia, poor transport product quality, and security from theft. This situa- between Ulaanbaatar and the central regions con- tion suggests that, by strengthening infrastructure tributes to a 40­85 percent price difference for the services, East Asian economies can influence their flour staple. Constraints on internal logistics also comparative advantage and, particularly, their Overview xxxi prospects for expanding production of more suggests that the productivity gains associated with sophisticated products (see Chapter 5 and Fink, more efficient services are particularly high, and Mattoo, and Neagu forthcoming). that competitiveness in high-value, differentiated Improving logistics involves investing in trans- agriculture and in manufacturing depends on effi- port infrastructure, but also introducing comple- cient business services. Static gains for developing mentary policies outside the transport sector. East Asia and Korea from services liberalization Transport regulatory agencies can improve logistics throughout the developing world are estimated at performance by allowing paperwork clearance for about $270 billion, or 10 percent higher income, by inland travel of containers and removing distor- 2015 (World Bank 2002c). tionary queuing pricing for commodities accessing To realize gains, reforms in the services sector the transport system. Multimodal coordination is need to be designed with broader development essential to promote seamless freight movement. In objectives in mind and with an eye toward introduc- Korea, for example, inland container terminals and ing competition, ensuring effective regulation to the development of intermodal exchanges have remedy market failure, and providing essential ser- made it easy for containers to reach cities far from vices to the poor. Pro-competitive regulation is par- the ports. Coordination is also critical for urban ticularly important in network-based services such as land use, standards, licensing, and security. Exam- transport, telecommunications, and energy services. ples of helpful urban land use policies include In East Asia, the importance of regulatory and pro- diverting traffic around congested areas and pro- competitive reforms to complement privatization has viding space for landside container storage. not always been fully appreciated (Figure 10). Better cross-border trade facilitation--customs, In this context, international trade agreements in e-commerce, paperless clearance, electronic process- services offer East Asia three main benefits: ing, and safety measures--can stimulate trade improved access to markets abroad and greater throughout the region. Some studies show that for openness at home through reciprocal liberalization; emerging East Asia, modest improvements in trade credibility of reforms as the result of binding inter- facilitation could expand trade by more than $200 national commitments; and regulatory cooperation. billion (Wilson and others 2002; also see Figure 8). In some services in which East Asian countries have Customs is the most obvious example. In some East a stake in cross-border trade--for example, in data Asian countries such as Singapore, trade documen- processing and other information technology- tation takes only 15 minutes to clear; two to three enabled services--they could secure access to other days is required for customs documentation and countries' markets through legally binding commit- clearance of sea cargo. By contrast, customs clear- ments from their trading partners. Some countries ance takes a minimum of five days in Manila and in East Asia have the potential to export significant seven days in Jakarta. In China, trade documentation transport and logistics, tourism, and business ser- clearance takes a day, whereas sea containers can take vices to China. East Asia also has a stake in tempo- 30­35 days to clear (Janet Tay Consultants 2002). rary labor mobility, as, for example, when its con- Several countries, including Cambodia and Viet- struction workers or accountants work abroad. nam, are addressing corruption in customs as critical A nonpreferential approach to market opening, to private sector trade (World Bank 2001b, 2003a). negotiated within a multilateral context, may be Certain logistics needs are best served by the pri- important for overall efficiency. It ensures that vate sector, particularly in ports, freight forward- access is provided by the most competitive ing, third-party logistics, warehousing, trucking, providers and avoids conferring a first-mover and the bus industry. Foreign entry may play an advantage on inferior regional suppliers. important role in increasing competition and the Nonetheless, regional arrangements may pro- quality of logistics. vide the advantages of greater policy space for regu- latory cooperation and learning-by-doing. East Asia is actively pursuing services liberalization in Services Sector Liberalization the context of the regional arrangements just out- East Asia has lagged behind other developing lined. Even preferential liberalization can improve regions in liberalizing trade in services.29 Evidence the status quo, given that many existing barriers are xxxii East Asia Integrates FIGURE 10 Regulatory and Competitive Reforms, Selected Countries Many regulatory and competitive reforms have not kept pace with privatization. Sequence of fixed line telecom reform in selected Asian countries 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 China Privatization Competition LD Regulation Indonesia Privatization 19% 23% Competition Regulation Korea Privatization 10% 20% 29% Competition ILD LD Local Regulation Malaysia Privatization 25% Competition ILD Local LD Regulation Philippines Privatization 100% Competition ILD Local LD Regulation Singapore Privatization 11% 17% Competition Local Regulation 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 Note: The percentage figures indicate the share of private equity ownership in the incumbent operator. Local, LD, and ILD refer to the local long-distance and international fixed-line segments, respectively. Regulation only captures the existence of a separate regulatory authority. Source: Fink, Mattoo, and Rathindran (2002). frictional (such as unnecessary qualification domestic reforms to remove policy barriers to com- requirements for foreign professionals and border petition. International agreements on investment hurdles for foreign transporters). Initial opening and competition can provide benefits through reci- among regional partners may provide opportuni- procity. However, most of the remaining restric- ties for learning-by-doing and lessen concerns that tions on reciprocal market liberalization and foreign competition would prematurely drive out nondiscrimination are in the services sector. Vehi- potentially competitive domestic providers. Regu- cles already exist at both the regional and multilat- latory cooperation, such as the harmonization and eral (General Agreement on Trade in Services, or mutual recognition of domestic regulations in GATS) levels for realizing gains potentially arising financial, professional, and a range of other ser- from service investment commitments. vices, may be more feasible in a regional context. International agreements that focus on protect- ing investors cannot be expected to expand markedly the flow of investment to new signatory Investment and Competition Policies countries. The bilateral investment treaties of East Raising the productivity of investment in East Asia Asian countries (Figure 11) already contain many requires harnessing the full force of competition protections, and, even though they are relatively inherent in global markets.30 Lower barriers to strong, these protections do not seem to have international trade and investment can be a power- increased investment flows to their signatories. ful pro-competitive force, like in the services sector Policy barriers to trade, such as high tariffs or where investment is one of the major modes for quotas, are the most important restraints on com- trade in services. The greatest potential for making petition. Another restraint on competition is fre- investment more productive lies in unilateral quent recourse to antidumping and other types of Overview xxxiii FIGURE 11 Geographic Distribution of Bilateral Trade Treaties Many protections for investors are already contained. Western Europe and other developed Other developing Asia Pacific countries countries China Hong Kong Indonesia Japan Korea Malaysia Philippines Singapore Taiwan Thailand Vietnam Covers 75% or more of countries in the partner group Covers 50­74% of countries Covers 25­49% of countries Covers less than 25% of countries Source: Bora (2001). contingent protection. Antidumping laws were rights (IPRs) have become a mainstream topic in originally created to counteract the predatory prac- global trade discussions. As the subject of the 1995 tices of foreign sellers into a home market. In prac- Trade-Related Aspects of Intellectual Property tice, more than 90 percent of the antidumping cases Rights (TRIPS) agreement, IPRs have in fact investigated would never have been launched if a become a contentious issue in the WTO, with competition standard--potential threat of injury developing countries concerned that across-the- to competition--had been used as a criterion board upgrading is premature, given their posi- (Messerlin 2000). Since 1995, more than 1,800 tions as users rather than producers of technologi- antidumping investigations have been initiated, cal innovations. Much of the debate has focused on many of them against East Asian exporters. In 2001 the impact of more stringent rights on East Asia as the countries most often subject to investigations a user of technology created elsewhere, but included China (53), Korea (19), and Indonesia and strengthened IPR regimes also play a role in local Thailand (16 each). technology generation by compensating inventors Large international companies with market and creators. power can form cartels that fix prices, allocate mar- Within the more advanced emerging economies kets, and restrain competition. The uncovering of and parts of China, stronger IPR regimes, success- several international cartels in the late 1990s has fully enforced, could stimulate innovation in tech- encouraged prosecutions, including by Korea. And nology, as outlined in Chapter 6, provided other there may be further scope for initiatives to disci- conditions are favorable. Korea's dramatic success pline international cartels, ranging from extending in patenting may be particularly relevant: The the reach of industrial nations' anticartel laws to number of Korean patents registered in the United notification and information exchanges by national States grew quickly in the late 1990s, propelling enforcement authorities to a multilateral agreement. Korea to sixth in the U.S. patent ranking, overtak- ing India, Brazil, and Singapore. Korea's strength- ened IPR regime played a role, but so did industrial Intellectual Property Rights: Nurturing Domestic upgrading, a big push in research and development Innovation from the chaebols (big company groups), and the Given the recent surge in knowledge-intensive government's selective targeting of the semicon- trade and investment flows, intellectual property ductor/electronics industry, where Korea has xxxiv East Asia Integrates emerged as a leading innovator worldwide. Given some difficult conceptual and practical obstacles that impending TRIPS-related obligations mandate that will need to be overcome before substantive stronger patent protection, it will be worthwhile for progress can be made in protecting or compensat- countries to identify complementary policies to ing the ownership of traditional knowledge. In East boost innovative activity among private firms in Asia, the Philippines is experimenting in this field their economies. with existing and pending legislation. For a broader range of middle- to low-income East Asian economies, copyrights might offer Environment and Labor Standards more scope for gains, given the considerable tal- ents of software developers, musicians, artists, and In general, environmental standards and workers' authors. Indonesia is one example of a country in rights are weaker in East Asia than in other parts of which there is potential for expansion in copy- the world with similar income levels; indices of right-sensitive industries as rights are improved environmental sustainability and regulation place and successfully enforced, particularly for the East Asia near the bottom, as do indices of labor software industry, small film industry, and invest- standards (Figure 12). Many observers have ment in artist development by music recording assumed that the region's weak standards play a companies. role in its strong export competitiveness and ability Traditional knowledge happens to be concen- to attract foreign investment. And within the trated in lower-income nations, and its protection region there is great suspicion that efforts to intro- is generally expected to have direct benefits for duce higher standards in global trade talks are reducing poverty. Often overlooked is the fact that backdoor ways for rich countries to deny new protecting traditional knowledge and genetic export opportunities to developing countries. resources also promotes efficient innovation in But, in fact, the available evidence does not sug- agriculture and biosciences. Yet traditional knowl- gest that stricter environmental and core labor edge needs to be combined with research and standards would hurt trade. As documented in development (R&D) activity, which is heavily con- Chapter 7, econometric evidence contradicts the centrated in industrial countries, for mutual bene- "pollution haven hypothesis," showing no signifi- fits from derived products. Chapter 6 identifies cant relationship between environmental regula- FIGURE 12 Environmental and Labor Standards, East Asia These standards are generally lower in East Asia than predicted by level of income. Environmental sustainability index, late 1990s Labor standards index, late 1990s 70 90 80 60 PHL KOR MYS 70 THA 50 60 IDN THA 40 50 CHN KOR PHL MYS 30 40 IDN CH 30 20 20 10 10 0 0 0 5,000 10,000 15,000 20,000 0 5,000 10,000 15,000 20,000 PPP per capita GDP, 1999 PPP per capita GDP, 1999 Note: PPP = purchasing power parity. Sources: Verite (2002) and World Bank (2001a). Source: Esty and Cornelius (2002) and World Bank (2001a). Overview xxxv tion and exports of pollution-intensive manufac- core labor standards, based on social concerns, tures for East Asia (see Table 7.5 in Chapter 7). without undue concern that their export competi- Analysis shows that exports and investments in East tiveness will suffer. Asia in the most sensitive sectors (pollution-inten- sive industries and labor-intensive industries) Reinforcing Social Stability appear to benefit when standards are raised. The through Broad Sharing of Benefits argument is that standards help set transparent "rules of the game," which then promote techno- To pursue these trade arrangements and behind- logical upgrading and skills development from the-border agendas, East Asian policymakers will which all parties benefit. Thus the evidence sug- need a supportive authorizing environment. To be gests that East Asian countries can raise their envi- effective, trade policy must be widely perceived as ronmental standards without adversely affecting reinforcing social stability and not contributing to their exports and investment inflows. further inequalities. Since the economic crisis of As for the impact of trade on the environment, 1997­98, many Asian countries have become more evidence suggests that, while trade-induced growth democratic and their authorities more broadly adds to pollution and the output of dirty industries, accountable for economic policy decisions. Impor- this effect is outweighed by the demand for a cleaner tant constituencies question the distribution of environment as a country raises its income and benefits that are produced by complex trade nego- acquires better technologies. For example, an analy- tiations. Evidence on how trade policies affect dif- sis of the effects of trade liberalization on water pol- ferent groups in society and different countries in lution in Chinese provinces during 1987­95--a the region must be brought into the policy debate period in which there was both an extensive pollu- more squarely in East Asia. Here it is important not tion levy system and significant opening to trade-- simply to generalize from other countries' experi- illustrates that trade opening was beneficial to the ences; East Asia has specific characteristics that environment overall (Dean 2002). determine the distribution of benefits from trade. The links between trade competitiveness and This understanding is critical for designing a set of core labor standards are equally complex. As with measures that enable the poor and the population trade and the environment, causality operates in more broadly to take fuller advantage of the growth both directions: Weak labor standards and poor opportunities from trade. As Oxfam noted, working conditions may influence trade flows and "[W]hen trade is harnessed to effective economic patterns, but labor standards themselves are poli- policies and positive poverty reduction strategies, it cies determined by many factors, including open- can act as a powerful force for change" (Oxfam ness to trade. Within East Asia, labor standards 2002). appear to be strengthening. The number of Inter- While progress in poverty reduction has been national Labour Organization (ILO) fundamental significant in the region, more needs to be done. ratifications has been increasing, and there is no Well ahead of schedule, the region has attained the evidence of backsliding on regulations or legisla- Millennium Development Goal of reducing tion. While better labor standards and improved extreme poverty incidence by half.31 Even so, working conditions could raise the costs of labor, roughly 13 percent of the people of emerging East workers who are treated better may also respond Asia were living on less than $1 a day in 2001. Inter- with more effort and invest in more skills, which country differences are wide, with the proportion could lower costs overall. ranging from virtually zero in Korea and Malaysia Econometric evidence suggests that stronger to more than 30 percent in Cambodia and Lao labor rights are empirically associated with higher PDR. Using the $2.00 a day poverty line, estimates export performance in East Asia (see Table 7.3 in for 2001 range from virtually zero in Korea to more Chapter 7). This association is particularly the case than 75 percent of the population in Cambodia and for core labor standards, including rights of free Lao PDR (World Bank 2003c). association and a ban on child labor. It is therefore The wide gap between rich and poor countries in likely that East Asian countries could introduce East Asia has been responsible for a rise in regional xxxvi East Asia Integrates income inequality.32 The most recent estimate of the One of the more promising approaches is a Gini coefficient for interpersonal inequality within micro simulation­cum­computable general equi- emerging East Asia is 46, higher than that in high- librium modeling approach that draws on the rich- income countries (40), though lower than that in ness of detail available from a modern integrated Latin America (58). Part of the increase has been household survey to provide a reasonably detailed driven by mildly rising inequality within countries, "map" of predicted welfare impacts by location and including China, Vietnam, and the Philippines. But socioeconomic characteristics. The authors of the bulk of the interpersonal inequality within Chapter 8 apply this approach to assess who within emerging East Asia--70 percent--is driven by China will gain from accession to the WTO. They inequality across location and across countries (Fig- find that, overall, the reduction in tariffs, quantita- ure 13).33 Policies and institutions that contribute to tive restrictions, and export subsidies associated trade and the fuller integration of the low-income with accession will have only a small immediate countries and regions of East Asia will be critical to impact on mean household income, inequality, and addressing income inequality in East Asia and con- the incidence of poverty. These findings are per- tributing to a stable region. haps not surprising in the case of China, given that the changes in agricultural effective protection are likely to be far lower than others have estimated Impact of Trade Reforms on Households and that significant changes already have been tak- At the household level, the impacts of the changes ing place in the manufacturing sector. Nonetheless, induced by specific trade policy reforms are com- the approach allows policymakers to go beyond plex. Policymakers need tools to better understand averages and aggregates that may hide offsetting how households will be affected in order to guide impacts across various households and to identify their policy sequencing and to help them identify those segments of the population for which the concrete risk-mitigation measures where needed. impact may well be more significant. For China, FIGURE 13 Poverty and Inequality Although poverty has declined steadily... ...inequality, driven by gaps between countries and locations, is cause for concern. Poverty: headcount index Gini coefficient for interpersonal Percent (US$2-a-day poverty line) income equality 90 80 Gini coefficient 70 50.0 60 40.0 30.0 50 20.0 40 10.0 30 0.0 1990 1996 1999 2000 2001 2002 2003 1988 1993 1998 East Asia Inequality within countries/locations Southeast Asia Inequality between countries/locations China Other transition (3) Source: World Bank (2003a). Note: "Country" is defined such that both China and Indonesia are divided into rural and urban parts. Source: Milanovic (2003). Based on household survey data. Overview xxxvii importantly, all gains accrue to urban households. Sectors Important to Social Stability: Specific Micro Meanwhile, rural families in some provinces will Interventions suffer income losses (Figure 14). The severest impacts will be felt in the northeast region covering In any economy, a handful of sectors tend to be Heilongjiang, Jilin, Liaoning, and Inner Mongo- critically important to social stability and the poor. lia--a region in which rural households depend The extent to which poor people, especially poor more heavily on feed grain production (for which producers, benefit from expanded trade opportuni- falling prices are expected from WTO accession) ties depends not only on trade policy but also on than those elsewhere in China. The most vulnera- complementary measures designed to tackle ble households are those dependent on agriculture, behind-the-border constraints to efficient produc- with relatively fewer workers and weak economic tion and exports. These measures, among other links to the outside economy through migration.34 things, foster the development of competitive mar- These impact analyses suggest that efforts to kets and public action to provide information, encourage labor flexibility are central to protecting reduce transaction costs (often linked to corrup- households adversely affected by reforms. Workers tion), call for delivery of public services, and need skills to be flexible, and they must be allowed address market and collective action failures. Poor to move occupationally and geographically. In producers are not only economically disadvantaged some cases, labor market reforms must be carried but often politically powerless, and when their out in parallel with trade reforms. In China, for interests are pitted against those of more powerful example, while there is a considerable degree of actors, they frequently lose. It is crucial that anyone labor mobility (as evidenced by the large migrant designing a set of measures that enables the poor to population), restrictions and impediments remain. take fuller advantage of greater access to markets With increased labor flexibility, all households-- understand the institutional and political economy not only urban--can potentially benefit from WTO underpinnings of the organizational structure of a accession. In Vietnam, expanding the labor market particular commodity or sector. options for vulnerable rural households and urban Chapter 9 applies this approach to rice in Cam- workers affected by enterprise restructuring will be bodia and Vietnam and cashmere in Mongolia. In critical. Impact "mapping" can also be used to Cambodia and Vietnam, as in much of East Asia, guide targeted safety net programs. most of the poor earn a living by growing rice. In FIGURE 14 Effects of China's Accession to WTO on Household Incomes All gains from China WTO accession accrue to urban Some regions suffer rural income losses. households. Percentage of gainers by Net gain or loss as a percentage of provinces, 2001­07 income by provinces, 2001­07 % of gainers % net gain/loss per capita 100 2 90 Urban 80 1 70 Urban 60 Total Total 0 50 40 ­1 30 20 ­2 Rural 10 Rural 0 ­3 1 6 11 16 21 26 31 1 6 11 16 21 26 31 Provinces ranked by provincial per capita income Provinces ranked by provincial per capita income Source: Chapter 8. xxxviii East Asia Integrates Mongolia, the livestock herding tradition, which is agricultural staple crops such as rice. Efforts to centuries old, provides a livelihood to the bulk of relieve transport and logistics constraints in remote the rural population. Domestic resource cost esti- areas are an important component of realizing effi- mates confirm that these two commodities offer ciencies along the value chain. Many of the people scope for efficient and expanding international who have yet to benefit from East Asia's successful trade opportunities. In these economies, the exter- trade strategies are located in these more remote nal trade policies for rice and cashmere, by and areas. As illustrated by the research findings on the large, no longer represent important impediments. effects of China's WTO accession, changes in trade This situation suggests that policymakers look policy alone may have a relatively small impact. more deeply into domestic institutions and the Other complementary policies will be essential for value chain domestically to identify the sectoral benefits to reach the poor more broadly. institutional and policy constraints between the two ends of the supply chains for rice and cashmere Trade and Stability that reduce the ability of poor producers to benefit from these expanding trade opportunities. Employment and earnings stability is another As presented in Chapter 9, analysis of institu- important element in efforts to ensure a supportive tional arrangements along the value chain for authorizing environment for trade. In most of the these two sectors highlights the considerable region's economies, labor is flexible; the adjustment impediments that exist both to upgrading the to demand shocks takes place through wages, pro- value of rice and cashmere production and to tecting employment, and spreading the gains or increasing the share of poor producers in the value losses over the broad labor force.35 The East Asia chains. In all three countries, powerful con- financial crisis and more recent global slowdown stituents appropriate rents through explicit prefer- have brought concerns about volatility to the fore- ences (credit for state-owned enterprises in Viet- front. Trade liberalization, it is argued, may have nam, export tax in Mongolia), industrial structure increased the exposure of the economy to more and (near monopsony for a few millers in Cambodia), larger shocks in the tradable goods sectors. If a sec- and corrupt practices (illegal fees) that are toler- tor faces larger or more frequent shocks, workers in ated while poor producers remain trapped in low- that sector may experience greater fluctuation in productivity states in the absence of improvements employment or earnings and therefore greater job in public service delivery (poor road network, and earnings insecurity. Adjustment can be costly, inadequate research and extension, and so forth). with loss in efficiency and displacement of workers. Similar forces may be present in many economies, There is particular concern that volatility transmit- but they are compounded in these three transition ted via trade liberalization may make workers more economies by the absence of strong market institu- vulnerable and that poorer workers might be hurt tions--institutions that are still developing as part disproportionately. of the transition from planned to market The analysis presented here for East Asia sug- economies. According to the analysis, the poor can gests that trade liberalization has not resulted in benefit from expanded opportunities presented by increased volatility or vulnerability for wage work- global integration, provided that these constraints ers. As outlined in Chapter 10, research in three are addressed. Giving voice to poor producers' countries (Korea, Thailand, and Indonesia) exam- interests by placing these issues on the policy ined the relationship among trade, idiosyncratic agenda is crucial to fostering reforms that unleash shocks, and fluctuations in workers' earnings and the productivity potential of poor people and employment. The first comparison is between peri- increase their bargaining power. ods of different degrees of openness to trade; the The trading arrangements and behind-the-bor- second is among industries characterized by differ- der agendas just discussed are also critical to ensur- ent degrees of trade exposure (see Korea in Figure ing a broad sharing of the benefits from trade in 15). Although the analysis is constrained by data key commodities. Reducing market access barriers availability, it finds no correlation between greater in agriculture in developed countries is critical for exposure to trade and greater variability in earnings sectors that are central to social stability, including or employment. Indeed, worker cohorts experi- Overview xxxix enced lower year-to-year fluctuations in their earn- portion of unskilled workers falling from high-pay- ings during the 1990s--characterized by fewer ing jobs into low-paying jobs was two to three times trade barriers--than in the previous decade. There that of skilled workers. Thus, while trade openness is some evidence that employment is more volatile makes an important contribution to growth and in trade-intensive industries; at the onset of the does not worsen vulnerability, policies to reduce the East Asia financial crisis, employment fluctuated vulnerability of workers must also focus on broad- more widely in the more trade-intensive industries ening education and developing skills. than in the less trade-intensive industries. However, employment in these industries recovered rapidly, Conclusions with few long-term adverse impacts. The analysis does find a clear relationship East Asia is emerging from the financial and cur- between vulnerability and workers' educational rency crises of 1997­98 with a new perspective on level and gender. Workers who have little education development. East Asian policymakers are turning and those who are female are much more likely to to the region as well as to the rest of the world to fall into poverty. In Indonesia, for example, the pro- develop a coherent set of economic policies that FIGURE 15 Trade and Stability in Labor Markets, Republic of Korea No higher earnings volatility is evident in periods of greater openness. Annual wage growth by birth-year cohorts, males Standard deviation of wage growth by birth-year 0.14 cohorts, males 0.12 0.1 0.1 0.09 0.08 0.08 0.06 0.07 0.04 0.06 0.02 0.05 0 0.04 ­0.02 0.03 ­0.04 0.02 ­0.06 0.01 ­0.08 0 1926 1929 1932 1935 1938 1941 1944 1947 1950 1953 1956 1959 1926 1929 1932 1935 1938 1941 1944 1947 1950 1953 1956 1959 1976­87 1987­97 Nor is higher earning volatility evident in sectors more exposed to trade. Standard deviation of log of real wages of sample Standard deviation of log of real wages of sample cohort by low, medium, and high trade exposure cohort of Korean males by broad sectors, 1986­2000 groups, 1976­2000 Males, secondary education, age 40 in 1990 Males, secondary education, age 40 in 1990 0.8 0.8 0.6 0.6 0.4 0.4 0.2 0.2 0 0 1980 1985 1990 1995 2000 1980 1985 1990 1995 2000 Low exposure Medium exposure High exposure Manufacturing Services and construction Others Source: Kim (2002). xl East Asia Integrates can deliver stability, growth, and regional integra- · Broadening agreements to include agriculture and tion. This volume does not try to be comprehen- services. China's approach to agricultural trade sive, but focuses on fundamental strategies that liberalization offers scope for the region, espe- promote cross-border flows of trade, along with cially Southeast Asia, to benefit significantly.36 domestic policies to maximize the impact of these China will emerge as a major importer of agri- flows on development and distribute the gains cultural and natural resource­intensive com- from trade widely. modities. If agriculture can be liberalized in a Several bilateral, regional, and multilateral ini- regional setting, as in the proposed ASEAN- tiatives have been advanced to pursue this strategic China free trade agreement, then a favorable agenda. But to succeed, these initiatives must have political dynamic can be set in place that will two features. First, they must provide a compelling lead to more open agricultural sectors overall. vision of how integration can deliver broadly As for the region's agricultural exports to the shared growth and prosperity. There are concerns developed world, all countries need to improve in Asia that a narrowly focused approach to global- their ability to meet sanitary and phytosanitary ization, like that found in the 1980s and early standards through better laboratories, stan- 1990s, has worsened the income distribution dards, and negotiation of mutual recognition within countries and widened the gap between the agreements, and rich countries need to reduce richer and poorer countries of the region. Even the degree of cascading tariffs in agricultural though the evidence may not support such con- processing and move toward the imposition of cerns, the new initiatives must move beyond the transparent ad valorem tariffs. For ASEAN perspective of trade policy to broader approaches countries in particular, the inclusion of agricul- that emphasize the links between trade and social ture in global and regional trade agreements is stability and the coherence with overall develop- important--potential welfare gains can be ment, stability, and growth. For this reason, the roughly doubled. chapters in this volume contain several studies of Several international studies37 suggest that the impacts of integration on poorer countries and the gains from liberalizing services could be very on vulnerable groups within countries. substantial--the same order of magnitude as Second, as initiatives for integration move that for liberalizing trade in manufactured beyond the narrow confines of trade policy, they goods and agriculture. This is an area in which must deal increasingly with second-best issues and East Asia has lagged behind other developing trade-offs, including in areas once treated as sensi- regions. The imperative for moving ahead with tive. For example, policymakers must balance user services liberalization is stronger because of the rights and incentives for innovation in deciding on emergence of China as a low-wage, efficient an intellectual property rights regime; decide where manufacturer. Other countries in the region will to allocate scarce resources to reduce logistics costs find it very difficult to compete in international and improve competitiveness; introduce appropri- markets on the basis of wages alone. Instead, ate environmental and labor standards; and expand they must rely on better producer services to liberalization into services and agriculture. enhance their competitiveness. Improved ser- An empirical approach is required to guide vices will permit them to participate in regional countries on the priorities in these areas; after all, production networks that offer prospects for East Asian countries are diverse and each has a dif- maximum efficiency. ferent set of priorities. The empirical estimates of · Improving logistics and trade facilitation. East the potential gains are often substantial, amount- Asia's progress on logistics has failed to keep ing to hundreds of billions of dollars for the prior- pace with its growth in trade. Logistics costs in ity areas identified in the rest of this section. It is many parts of East Asia are high, and logistics clearly worthwhile for policymakers to devote industries are underdeveloped. Logistics plays a major effort to get the reforms right. The chapters critical role both in determining aggregate levels in this volume suggest some broadly applicable of trade and in ensuring that development bene- priority areas for action, as described in the follow- fits spread beyond coastal regions. For East ing list. Asia's less open and accessible countries and Overview xli regions, the development of more tightly inte- · Framing policies on intellectual property rights. grated domestic markets and logistics systems is Intellectual property rights are a part of the a high priority. Complementary institutional multilateral framework for trade. Strong protec- actions are needed not only to promote an tion of intellectual property rights encourages appropriate mix of transport modes, but also to innovation, but it is not a magic bullet; other extend better transport services to remote areas conditions, including investments in skills and and establish better conditions for market devel- provision of a competitive business environ- opment. Beyond strengthening physical infra- ment, are also needed. Several of the region's structure, governments also need to undertake more advanced countries, and parts of China, improvements in the regulatory environment can hope to stimulate technological advances if and in the conditions for cross-border facilita- they better protect intellectual property and pay tion. Customs clearances deserve special atten- more attention to competition policy (Yusuf tion; addressing corruption has been high- and others 2003). Stronger copyright laws, pro- lighted as an issue in several countries. vided they can be enforced, may yield some sig- · Institutional strengthening to safeguard and bene- nificant gains for domestic software industries fit the poor. The fear that globalization has as well as the arts and music sectors. As for tradi- increased volatility seems to be misplaced. Avail- tional knowledge, some important conceptual able evidence for East Asia suggests that greater and practical problems need to be sorted out openness to trade does not make workers more before progress can be made in protecting its vulnerable; indeed, greater openness has in fact ownership and linking traditional knowledge stabilized wage incomes and employment across issues to WTO agreements or other mechanisms all skill categories. That said, women workers for managing research and development. and workers with little education are noticeably · Reconsidering environmental and labor stan- more vulnerable than others to falling into dards. There is no evidence from East Asia to poverty, emphasizing the need for policies to support the argument that improved environ- broaden education and develop skills. To ensure mental and core labor standards would unfairly that the poor within countries benefit from affect manufacturing competitiveness. Indeed, openness, policies are also needed to support the evidence suggests that East Asian countries workers' geographic and occupational mobility, can raise their environmental and core labor as are specific actions to reduce transport and standards without adversely affecting their marketing costs in order to improve farm gate exports and investment inflows. The fierce prices for the products on which the poor rely. resistance of many in the region to considering these issues in line with broader development Beyond these immediate priorities, several other strategies may be misplaced. findings are highlighted in this volume: This rich agenda suggests scope for East Asian · Easing adjustment in labor-intensive industries. countries to build on the international commit- Certain sectors in East Asia, especially the gar- ment to a new development round of trade negoti- ment sector, are likely to be seriously affected by ations based on the Doha Development Agenda, expanded output and exports from China. The and on their own commitment to deepening region's more developed economies will be able regional economic ties. Countries must operate on to shift out of garment-making into higher three levels. Internationally, they can influence the value added industries, and today's presence of negotiations, and, in doing so, they should pay par- current account surpluses implies that the exter- ticular attention to standards and to liberalization nal adjustment will be manageable. For other in agroindustries and services. Regionally and bilat- countries, supply chain analysis suggests that erally, they can achieve deeper integration, which there is scope for greatly improving competitive- offers prospects for gains in sensitive sectors ness in the garment and other manufacturing (although with due care to use these agreements to industries by improving logistics and reducing further multilateral liberalization). Nationally, the costs of corruption. countries need to take a more comprehensive xlii East Asia Integrates approach to policy than before, moving beyond misleading. Recent research reveals that overall protection for narrow trade and industry interests focused on crops such as rice, wheat, and maize is far less than indicated by statutory tariff rates (Huang and Rozelle 2002). protecting specific manufacturing subsectors to 14. The decision contrasts with the costly and wasteful path to adopt a strategy with sectoral components that are ever higher protection followed by previous high-growth aligned to encourage integration and competition East Asian economies. 15. The positive impact on the automobile sector is a relatively as means of delivering growth, stability, and new finding from François and Spinanger (2002), as dis- poverty reduction. cussed in Chapters 1 and 2. 16. See François and Spinanger (2002). The agreement requires the elimination of quotas by January 2005. Endnotes 17. The situation also highlights the importance of improving this area of WTO rules. 1. Emerging East Asia is defined here as the Association of 18. The number of sectors with guaranteed unrestricted access Southeast Asian Nations (ASEAN), plus other newly indus- is still lower in China than in most other countries for trializing economies, China, and Mongolia. ASEAN mem- cross-border trade and consumption abroad (modes 1 and bers are Brunei Darussalam, Cambodia, Indonesia, Lao 2) and essentially zero for establishment trade (mode 3). People's Democratic Republic, Malaysia, Myanmar, the 19. Trade intensity is the degree to which the value of trade Philippines, Singapore, Thailand, and Vietnam. between two countries conforms to expectations that are 2. This figure is not much less than the 7.7 percent share of global based on the partners' relative importance in world trade. It trade among the North American economies when the North is computed as the ratio of the share of country i's exports American Free Trade Agreement went into force in 1994. going to country j (xij/Xit, where xij and Xit are the values of 3. $2.15 per day in 1993 PPP$. See World Bank (2003c). All i's exports to j's and i's total exports) relative to the share of other dollar amounts are current U.S. dollars. world trade destined for country j (xwj/Xwt, where xwj and 4. East Asia is perhaps the most diverse region in the world. Xw are the values of the world's exports to j and the world's The income per capita, in purchasing power terms, of Sin- total exports). An index of more than unity indicates that gapore, for example, is 25 times that of Lao PDR, and the bilateral trade is larger than expected given the partner gap between the two in institutional capacity to manage country's importance in world trade. development may be even greater. 20. See Schiff and Winters (2003) for a discussion of regional- 5. See World Bank (2000) for a first assessment of the new ism from the perspective of developing countries more policy and institutional agenda. broadly. 6. China is widely seen as having contributed to stability in 21. Nonmembers may be net losers, though Table 1 suggests 1997­98 by maintaining a stable exchange rate in the face these losses may be minimal. Nonetheless, consistency with of major devaluations of other currencies. Few scholars the WTO could be questioned, and such arrangements give credence to the notion that competition from China could be subject to challenge in the WTO. was the cause of the Asian crisis. 22. See results in Hoekman and Konan (1999) for the Euro- 7. The ratio of China's GDP to that of the rest of emerging pean Union and Egypt and in Brenton and Manchin (2002) East Asia was 1:2 in 1996 and 1:1 in 2002. for the European Union and Russia. 8. China's share of East Asian exports has grown significantly 23. See Ianchovichina and Walmsley (2002) for China. over the last decade--from 10 to 20 percent for Singapore, 24. See Kalima and Krumm (2002), which draws on a combi- Korea, and Taiwan (China), and from 2 to 5 percent for nation of labor market surveys, World Bank income distri- ASEAN countries. bution data, and computable general equilibrium (CGE) 9. Exceptions were Singapore, Cambodia, and Brunei. modeling results. 10. The trade complementarity index measures how well the 25. World Bank (2002c). Real income gains for Korea and export profile of one country matches the import profile of developing East Asia are estimated for 2015 at $11 billion, another. The index is zero when no good exported by one assuming fixed productivity, and $32 billion with endoge- country is imported by the other, and 100 when the export- nous productivity, excluding production subsidies, com- import profiles exactly match. Higher index values indicate pared with baseline income. Static gains are slightly less, more favorable prospects for a successful trade arrange- including production subsidies, but dynamic gains would ment between countries. The index stands at more than 60 increase to $76 billion. China would capture one-third of for emerging East Asia. By contrast, indices for previous static gains and two-thirds of dynamic gains. failed trade arrangements ranged from values of 7 to 22. 26. The direct impact of domestic support programs is of less See Ng and Yeats (2003). relevance to East Asia because this support is concentrated 11. During the crisis, real exchange rate depreciation also in products that are not of crucial importance to East Asia played a role in reducing protectionist tensions. Nonethe- developing countries, with the exception of rice. Products less, the dangers of protectionism and of vested interests such as fruits, vegetables, and spices together account for capturing political support in specific cases should not be only 7 percent of the developed countries' total domestic underestimated. support payments. 12. China's domestic market is the largest of these, but recent 27. Many nonreciprocal preference schemes have had a low trends show Korea, Thailand, and other economies grow- rate of uptake by developing countries, including those in ing on the strength of domestic consumption demand. East Asia, in large part because of stringent rules of origin. 13. Assessments of the impact of WTO accession on China's agri- For example, Cambodia made use of only 36 percent of its cultural sector based on limited comparisons of statutory tar- entitlement under the EU's Everything but Arms initiative iffs and the rates agreed on in the accession process would be in 2001 (Brenton 2003a). Overview xliii 28. For example, by Codex Alimentarius, the Organization for Brenton, Paul. 2003a. "Integrating the Least Developed Coun- Animal Health, the International Organization for Epi- tries into the World Trading System: The Current Impact of zootics, or the Hazard Analysis Critical Control Point EU Preferences under Everything But Arms." Policy Research (HACCP). In Malaysia, the Ministry of Health operates a Working Paper. World Bank, Washington, D.C. Forthcom- voluntary certification program for private firms in the ing, Journal of World Trade. food processing sector. This program was introduced in ------. 2003b. "Notes on Rules of Origin with Implications for 1996­97 in response to EU requirements for HACCP in Regional Integration in South East Asia."World Bank, Wash- fish processing plants. ington, D.C. Processed. 29. See Mattoo, Rathindran, and Subramanian (2001). Brenton, P., and M. Manchin. 2002. "Trade in Services, Foreign 30. The discussion in this section is based on World Bank Direct Investment and Technology Transfer: Implications of (2003b). an EU-Russia Free Trade Agreement for Economic Effi- 31. On the basis of a higher poverty line of $2.00 per day, two- ciency and Growth." Report prepared for the European thirds of the population was poor in 1990; this has declined Commission. Processed. to about two-fifths today (Figure 14). Carruthers, Robin, and Jitendra N. Bajpai. 2002. "Trends in 32. The Gini coefficient for interpersonal inequality in emerg- Trade and Logistics: An East Asian Perspective." Working ing East Asia was estimated at 42.0 in 1988 and 46.8 in Paper No. 2. World Bank, Transport Sector Unit, Washing- 1993. ton, D.C. 33. The Gini coefficient of income inequality (G) can be Centre for International Economics. 2002. "Integration and decomposed into a within-country inequality component Poverty: An Economy-Wide Analysis." Draft prepared for and a between-country inequality component (G = S Gi pi the World Bank. Canberra and Sydney, November. pi + S S [(yj ­ yi )/ yi] pi pj + residual). The within-country Dean, Judith M. 2002. "Does Trade Liberalization Harm the component refers to the weighted sum of within-country Environment? A New Test." Canadian Journal of Economics inequalities, where each country's (i-th) inequality is repre- 35 (4): 819­42. sented by its own Gini coefficient (Gi), and the weight is Esty, Daniel C., and Peter K. Cornelius. 2002. Environmental Per- given by the product of the country's share in the world's formance Measurement: The Global Report 2001­2002. population (pi) and the country's share in world income Oxford: Oxford University Press. (pi). The between-country component is derived such that Fink, Carsten, Aaditya Mattoo, and Ileana Cristina Neagu. all countries are ranked by their mean income (from poor- Forthcoming. "Assessing the Impact of Communication on est to the richest) and the relative distance between coun- Trade." Policy Research Working Paper. World Bank, Wash- tries' mean incomes [(yj ­ yi )/ yi] is weighted by the prod- ington, D.C. uct of the poorer country's share in world population (pj) Fink, Carsten, Aaditya Mattoo, and Randeep Rathindran. 2002. and the richer country's share in world income (pi). "Liberalizing Basic Telecommunications: The Asian Experi- This situation contrasts with the one in Latin America, ence." HWWA-Institut fur Wirtschaftsforschung Discussion where the bulk of inequality reflects differences within Paper 63. countries, and only 19 Gini points are attributable to François, Joseph F., and Den Spinanger. 2002. "Market Access in between-country differences. For East Asia, the definition Textiles and Clothing." Prepared for Conference on Inform- of country is such that both China and Indonesia are ing the Doha Process: New Trade Research for Developing divided into rural and urban areas. Countries, Cairo, May. 34. A similar approach assessed the impact of the ASEAN Free Hallward-Driemeier, Mary, Giuseppe Iarossi, and Kenneth L. Trade Area and the U.S. bilateral agreement on aggregative Sokoloff. 2002. "Export and Manufacturing Productivity in Vietnam household groupings--see Centre for Interna- East Asia: A Comparative Analysis with Firm-Level Data." tional Economics (2002). That analysis suggests that the NBER Working Paper No. W8894. National Bureau of Eco- benefits are spread across the income deciles and across nomic Research, Cambridge, Mass. rural and urban areas. However, the poorest rural deciles Hoekman, Bernard, and Denise Eby Konan. 1999. "Deep Inte- benefit the least, and individual households could experi- gration, Nondiscrimination, and Euro-Mediterranean Free ence vulnerabilities. Trade CEPR." Policy Research Working Paper 2130. World 35. The most severe test of this occurred during the Asian eco- Bank, Washington, D.C. nomic crisis. Except in Korea, unemployment in the crisis Huang, J., and S. Rozelle. 2002. "The Nature of Distortions to countries was kept to manageable levels; most of the Agricultural Incentives in China and Implications of WTO adjustment fell on wages. Accession." Paper presented at Seminar on WTO Accession, 36. As noted earlier, the adjustments in China's agricultural sec- Policy Reform and Poverty Reduction in China, World Bank, tor resulting from WTO accession may be smaller than Beijing, June 28­29. some other analysts suggest, because China's current prac- Ianchovichina, Elena, and William Martin. 2001."Trade Liberal- tices are already more liberal than implied by statutory rates. ization in China's Accession to the World Trade Organiza- 37. See Yusuf and others (2003) on producer services. tion." Journal of Economic Integrations 16 (4): 421­45. ------. 2002. "Economic Impacts of China's Accession to the References WTO." Paper presented at Seminar on WTO Accession, Pol- icy Reform and Poverty Reduction in China, World Bank, The word processed describes informally reproduced works that Beijing, June 28­29. may not be commonly available through libraries. Ianchovichina, Elena, and Terrie Walmsley. 2002. "Regional Impact of China's Accession."World Bank, Washington, D.C. Bora, Bijit. 2001. "Prospects for Harmonizing Investment Poli- Processed. cies in the East Asian Region." Draft prepared for the World IMF (International Monetary Fund). Various years. Direction of Bank. UNCTAD, Geneva. Trade Statistics. Washington, D.C. xliv East Asia Integrates Janet Tay Consultants. 2002. "Facilitating Trade: The East Asian sumption, Prices, and Trade."In OECD, China's Agriculture in Experience in a Comparative Context." World Bank, Wash- the International Trading System, OECD Proceedings. April. ington, D.C., May. Processed. Scollay, Robert, and John Gilbert. 2001. New Subregional Trading Kalima, Blandina, and Kathie Krumm. 2002. "Delays in Imple- Arrangements in the Asia-Pacific. Washington, D.C.: Institute mentation of Agreement on Textiles and Clothing: Poverty for International Economics. Impact in East Asia."World Bank, Washington, D.C., August. ------. 2003. "Impact of East Asian Regional or Subregional Processed. FTAs." Report for Australian Department of Foreign Affairs Kawai, Masahiro, and Shujiro Urata. 2002. "Trade and Foreign and Trade. May. Direct Investment in East Asia." Paper presented to Confer- Verite. 2002."Report to California Public Employees' Retirement ence on Linkages in East Asia: Implications for Currency System (CalPERS): Emerging Markets Research Project." Regimes and Policy Dialogue, Seoul, September. Available at verite@verite.org. Kim, Dae II. 2002. "Openness and Worker Vulnerability in Wilson, John, Catherine Mann, Yuen Pau Woo, Nizar Assanie, Korea: A Descriptive Study." World Bank, Washington, D.C. and Inbom Choi. 2002. "Trade Facilitation: A Development Processed. Perspective in the Asia Pacific Region." World Bank working Lall, Sanjaya. 1998. "Exports of Manufactures by Developing paper presented to APEC. World Bank, Washington, D.C. Countries: Emerging Patterns of Trade and Location." Processed. Oxford Review of Economic Policy 14: 54­74. World Bank. 2000. East Asia Regional Overview: Recovery and Lardy, Nicholas. 2002. Integrating China into Global Economy. Beyond. Washington, D.C. Washington, D.C.: Brookings Institution Press. ------. 2001a. Indonesia: "Environment and Natural Resource Mattoo, Aaditya, Randeep Rathindran, and Arvind Subraman- Management in a Time of Transition." Washington, D.C. ian. 2001."Measuring Trade Liberalization and Its Impact on ------. 2001b. Cambodia Integrated Framework. Washington, Economic Growth: An Illustration." World Bank, Washing- D.C. ton, D.C. Processed. ------. 2002a. East Asia Regional Overview: Making Progress in Messerlin, Patrick. 2000. Measuring the Costs of Protection in Uncertain Times. Washington, D.C. November. Europe. Washington, D.C.: Institute for International Eco- ------. 2002b: "Trends, Issues and a Glimpse of the Future." nomics. East Asia Regional Office Mimeo. Washington, D.C. Milanovic, Branko. 2003. "Worlds Apart: International and ------. 2002c. Global Economic Prospects, and Developing Coun- World Inequality, 1950­2000." World Bank, Washington, tries: Making Trade Work for the Poor. Washington, D.C. D.C. Processed. ------. 2003a. Vietnam's Exports: Challenges and Opportunities. Ng, F., and A. Yeats. 2003. "Major Trade Trends in East Asia: Washington, D.C. What Are Their Implications for Regional Cooperation and ------. 2003b. Global Economic Prospects and Developing Coun- Growth?" Policy Research Working Paper 3084. World Bank, tries: Investing to Unlock Global Opportunities. Washington, Washington, D.C. D.C. Otsuki, Tsunehiro, and John S. Wilson. 2001. "Global Trade and ------. 2003c. East Asia Regional Overview. East Asia Navigates Food Safety: Winners and Losers in a Fragmented System." Short-Term Shocks for a Stronger Future. Washington, D.C. Working Paper 2689. World Bank, Washington, D.C. Available at: www.worldbank.org/eap/eapnsf. ------. 2002. "Beef Trade and Veterinary Drug Standards." Yusuf, Shahid, with M. Anjum Altaf, Barry Eichengreen, Sudar- World Bank, Washington, D.C. Processed. shan Gooptu, Kaoru Nabeshima, Charles Kenny, Dwight H. Oxfam. 2002. Rigged Rules and Double Standards: Trade, Global- Perkins, and Marc Shotten. 2003. Innovative East Asia: The ization, and the Fight against Poverty. Oxford: Oxfam. Future of Growth. Washington, D.C: World Bank and Oxford Schiff, Maurice, and L. Alan Winters. 2003. Regional Integration University Press. and Development. Oxford: Oxford University Press. Schmidhuber, J. 2001. "Changes in China's Agricultural Trade Policy Regime: Impacts on Agricultural Production, Con- Part I WIDENING OPPORTUNITIES IN TRADE ARRANGEMENTS 1 China's Accession to the WTO: Impacts on China William J. Martin Deepak Bhattasali Shantong Li One of the great milestones for economic reform in studies have been conducted of aspects of the China, and for the world trading system, was the impact of WTO accession on China, including accession of China to the World Trade Organiza- symposia in the China Quarterly (see Fewsmith tion (WTO) in December 2001. China's accession 2001) and China Economic Review (see Chun, will have an enormous impact on both China and Fleisher, and Parker 2001); surveys of the estimated the world trading system (Martin and Ian- impacts of trade liberalization (see McKibbin and chovichina 2001). Tang 2000; Gilbert and Wahl 2002); and studies of Other developing countries will feel the impact the impact of WTO on China and vice versa (Mar- of China's WTO accession through four main tin and Ianchovichina 2001). This chapter builds channels: on a large integrated study--undertaken by China's Development Research Centre of the State Council 1. Expansion of markets in China for exports and the World Bank--that examines the legal 2. Increases in the supply of exports into others' aspects of accession; estimates the impacts of the markets resulting policy changes on the overall economy; 3. Competition in third-country markets and then assesses impacts on poverty and policy 4. Expansion of investment in China and, poten- options for dealing with these problems.1 tially, outward foreign investment from China. There are many perspectives on China's acces- sion to the WTO. According to one view, which Understanding how these linkages will play out focuses on legal rights and responsibilities, China's requires a good understanding of how WTO acces- key challenges are to meet its legal commitments to sion will affect China itself and, particularly, of how implement particular policies and to ensure that its policy might develop in the many areas of reform rights are maintained through WTO mechanisms that go beyond the specific commitments in such as its dispute settlement mechanism. Another, China's WTO accession package. Many excellent purely economic, view is that accession involves a This chapter draws on a program of research presented in further detail in Bhattasali, Li, and Martin (forthcoming). The research benefited from the support of UK Department for International Development. Particular thanks are due to Ippei Yamazawa and Shujiro Urata for their comments at a seminar held at the World Bank office in Tokyo. 3 4 East Asia Integrates set of economic policy changes that will open up China's case, the application of this general princi- China's economy and make it much more engaged ple has involved some additional commitments, in the global economy. The most compelling per- including eliminating dual pricing systems, phasing spective is the one that sees WTO accession as a out restrictions on trading, and introducing more component of the broader set of reforms that uniform administrative arrangements and judicial China has undertaken since 1978--that is, as a set review. These agreements are of crucial importance of policies that can contribute to development and not just for the central authorities but also for the poverty reduction. lower tiers of government, which are often involved WTO rules are important ingredients in formu- in internal trade and regulation. lating good policies for development, but they do not constitute a complete recipe. Like most other Market Opening laws, their intent is to reduce the adverse impacts of the actions of one individual or group on another The market opening principle is reflected in com- rather than to guide a country on the best way to mitments by China to abolish nontariff barriers, achieve its goals. Many WTO agreements allow a reduce tariffs, and open its service sectors; in com- great deal of choice within the range of legally per- mitments by countries importing from China to missible policy options. Within the framework of abolish the quotas on textiles and clothing that its WTO commitments, China may choose to pur- were originally imposed under the Multi-fiber sue its development goals either aggressively or Arrangement (MFA); and in commitments by the quite tentatively.2 United States and other countries to impose MFN In seeking to understand the implications of tariffs on China. China's accession to the WTO, we examine the The cuts in import tariffs that China has offered nature of the policy changes associated with acces- are very substantial, and will result in a reduction in sion; the implications of these policy changes for the the weighted average tariff from 12 percent in 2001 Chinese economy at large and for individual house- to 6.8 percent at the end of China's WTO imple- holds; and the complementary policies needed to mentation period.4 But these reductions are small take advantage of the opportunities created by compared with the reduction of 29 percentage accession and to minimize the adjustment costs. points that China achieved between 1992 and 2001. China's commitments to liberalize trade in services are extremely large relative to those of almost all Policy Reforms Associated with other countries, although they are more often sub- Accession ject to qualifications or reservations than those of The five basic principles of the General Agreement other countries. on Tariffs and Trade (GATT) and the WTO provide China's trading partners, for their part, have a useful framework for analyzing the implications of made an important "concession" in abolishing the China's accession. These principles are: (1) nondis- quotas that were imposed under the MFA. The crimination (the most-favored-nation [MFN] prin- abolishment of these quotas is likely to provide sig- ciple, under which the best market access given to nificant opportunities for China to increase its any one Member is extended to all other Members); exports of textiles and clothing, given its strong (2) market opening; (3) transparency and pre- comparative advantage in these goods, although dictability; (4) undistorted trade; and (5) preferen- the effects will be tempered if the importers exer- tial treatment for developing countries.3 cise their right to impose special textile and cloth- ing safeguards for a year at a time during a transi- tion period up to 2007. Nondiscrimination Another important "concession" by almost all The general principle of nondiscrimination existing members of the WTO is to refrain from requires WTO Members to give equal treatment to invoking nonapplication provisions of the type that competing suppliers and not to discriminate were widely invoked against Japan when it joined between domestically produced and imported the GATT. Even though this concession does not goods or services in their internal markets. In involve much actual market opening, it means that China's Accession to the WTO: Impacts on China 5 China now receives permanent MFN status in vir- industrial goods, disciplines on some forms of tually all markets. Having this status frees China export subsidies generally allowed in developing from onerous one-sided review procedures, such as countries. the former annual review of China's MFN status in The existing regime of antidumping and safe- the United States, and it assures investors in China's guard measures has troubling implications for export industries that foreign markets will be avail- China's access to export markets. The WTO rules able to them on a continuing basis. against dumping are biased toward finding dump- ing even where no economically meaningful dumping exists (Messerlin 2002). The situation is Transparency and Predictability worse for China than for other WTO Members, The transparency and predictability of trade policy because 70 percent of China's exports are in prod- are enhanced both through general WTO policy ucts that are most vulnerable to antidumping rules, such as the need to publish trade rules and measures. Furthermore, China could remain vul- regulations, and through specific commitments nerable, for up to 15 years, to highly discriminatory China has made, including provisions for uniform provisions that are applied to nonmarket application of the trade regime and for indepen- economies and dramatically increase the probabil- dent judicial review. China has also put in place a ity of dumping being found. When antidumping mechanism whereby concerned parties can bring duties are applied under these provisions, they are problems of local protectionism to the attention of generally much higher than the duties applied to the central government. Another important con- market economies. For example, the average 40 tributing factor is China's binding of its entire tariff percent duty applied by the United States against schedule for goods, almost always at tariff levels nonmarket economies was more than 10 times below previous applied rates. This binding not only higher than that applied where the margin was cal- reduces tariffs and their variance, but also increases culated based on actual costs. predictability by ruling out tariff increases in the A particularly worrying feature of China's acces- future. The annual transitional reviews to be held sion agreement is the product-specific transitional for eight years after China's accession will provide safeguard provisions. These provisions may be additional information about China's regime and applied by any WTO Member, and may then trigger its reforms during that period. Also important for actions against the diversion of Chinese exports to increasing transparency are China's commitments other markets (Panitchpakdi and Clifford 2002). to phase out restrictions on trading rights for all They are, in a sense, worse than the provisions on products, except for a short list of commodities that nonmarket economy treatment in that they intro- may remain subject to state trading, and to allow duce an entirely new form of protection, targeted the entry of foreign, and frequently domestic, sup- specifically against China; they are more readily pliers into distribution and wholesale services. triggered than regular safeguards; and they are avail- The emphasis on transparency, and the specific able to China's trading partners for up to 12 years requirements to this effect, may help to avoid costly from the date of China's accession (Anderson and and acrimonious disputes of the kind that marred Lau 2001). The trade diversion measures allowed trade relations between Japan and (in particular) under these provisions are particularly troubling, the United States during Japan's era of high export because they provide even less procedural protec- growth. tion than is available under regular safeguards. Proposals to use the product-specific safeguards against China's exports of textiles and clothing have Undistorted Trade already surfaced in the United States, despite the The WTO principle of undistorted trade involves continuing presence of quotas originally imposed general disciplines in areas such as subsidies and under the Multi-fiber Arrangement. If the product- countervailing measures, antidumping, and safe- specific safeguards are invoked and other countries guards. China has made more stringent commit- do not resist the temptation to use the trade diver- ments than those normally required, including one sion measures, there is a risk of a domino effect: not to subsidize its agricultural exports, and, for China's exports would be diverted to fewer and 6 East Asia Integrates fewer markets, and China would increasingly be Preferential Treatment for Developing Countries tempted to retaliate against what it would almost certainly see as unfair barriers against its exports. Preferential treatment was a particularly vexing For example, China might increase its use of issue throughout the negotiations. Although China antidumping actions and contest the safeguard has a much lower per capita income than many actions through the WTO's dispute settlement economies in the WTO that are classified as devel- mechanism, or it might mount a concerted cam- oping, its size and growth performance made exist- paign for reform of the rules in the Doha negotia- ing WTO Members reluctant to accord it full devel- tions, particularly in the area of antidumping oping country treatment. In many areas of the (Messerlin 2002). Whatever the case, the result agreement, China is likely to have full access to the could be serious damage to both China and the developing country provisions, but in particular trading system. cases it faces tighter restrictions than other devel- Clearly, China will need to contest unjustified oping countries.5 At the same time, China has actions and seek settlement of disputes. But retalia- obtained specific transitional arrangements in tion that involves launching antidumping actions is areas such as the phasing out of quotas and licenses likely to be extremely costly to its economy, both by and phased entry of foreign enterprises--areas that reversing the liberalization process and by increas- are not generally available to developing country ing the uncertainty about trade policy. Even though members. Because special treatment in the form of retaliation is likely to be politically attractive (as preferential access to industrial country markets is confirmed by the recent upsurge in antidumping not important for China, it has a strong interest in actions in China--and the dramatic upsurge in reducing the trade barriers in industrial countries these actions by developing countries such as in the only way it is able--through multilateral Argentina, India, Mexico, and South Africa), the trade reform that lowers protection in the indus- economic costs to China in particular suggest that trial countries, particularly on labor-intensive it should be avoided as much as possible. products such as textiles, clothing, and footwear in If China chooses instead to lead a push for which China has a strong comparative advantage. reform of the antidumping and safeguard rules to reduce the abuses of these protectionist measures, it Intellectual Property Rights could greatly improve the performance of its own economy in the short run and that of the global The Agreement on Trade-Related Aspects of Intel- trading system in the longer run. Messerlin (2002) lectual Property Rights, or TRIPS, which is an inte- suggests two courses of action. First, related to its gral part of the WTO, involves a number of GATT treatment as a nonmarket economy, China could principles such as nondiscrimination and seeks to press for new rules on the automatic granting of achieve a balance between offering incentives for market economy status in a particular commodity, innovation and allowing broad access to informa- applicable as long as a country meets basic condi- tion. An intellectual property regime appropriate to tions such as low rates of protection, an absence of a developed country may be much too rigorous for serious nontariff barriers, and an absence of state a developing country; such a regime may inhibit monopoly in the distribution of that commodity. growth by limiting innovation and diffusion and Second, as for antidumping measures more gener- result in excessive transfers to foreign producers of ally, China could put forward, or strongly support, intellectual property. But all countries require regu- proposals to narrow the use of antidumping mea- lations to ensure that markets remain competitive sures and to reduce their severity. China might also without excessively reducing the incentive to inno- seek similar relief on the product-specific safe- vate. The TRIPS agreement is generally seen as pro- guards. The abuse of antidumping policies by the viding the flexibility needed to design such a major trading countries and by a growing number regime, but implementation will not be easy. of developing countries is a problem for most of For China, the TRIPS agreement is a key aspect the other economies of East Asia, and China could of WTO accession. Recognizing the need to stimu- surely strengthen the coalitions formed at the WTO late innovation domestically and gain access to for- to push for stronger rules against this abuse. eign technology and responding to pressure from China's Accession to the WTO: Impacts on China 7 its trading partners, China has strengthened its many poor people engaged in this sector. Much of intellectual property rights (IPR) regime. Since the concern has arisen from comparisons of China's 1990, China has updated its laws on copyrights, statutory tariffs on agriculture in the 1990s with the trademarks, patents, and trade secrets and adopted rates agreed on in the accession process (see, for protection for new plant varieties and integrated example, Schmidhuber 2001). But other authors circuits. These changes are particularly important have pointed out that the statutory tariff rates bore for China's East Asian neighbors, many of whom little relationship to the actual protection (or taxa- are engaged in intellectual property­intensive tion) that China's agricultural sector experienced activities in China. (see, for example, Johnson 2000 and Lin 2000). China's intellectual property rights regime is The evidence on the actual rates of protection broadly appropriate to China's situation.6 In partic- applying to agriculture is still extremely limited and ular, Maskus (2002) believes that China's policy of often contradictory. Agricultural trade in China has public procurement of pharmaceuticals at negoti- been influenced by a bewildering array of policies ated prices is appropriate for providing public on imports and exports, including state trading, health services. He also concludes that, with current designated trading, quotas, licenses, tariffs, and tar- reforms, the regime will be fully consistent with the iff-rate quotas. Many studies have tried to deal with TRIPS requirements. this problem by summarizing the protective impact However, Maskus raises some important issues of agricultural trade policies in terms of the price about TRIPS policies and their implementation in distortions created by these measures. The more China. One concerns proposals to extend patent restrictive the trade measure, in general, the larger protection to computer software, giving a level of will be the distortion--that is, the gap between the protection currently provided only in the United domestic price and the international price. States, Japan, and Australia that is perhaps excessive Several studies have estimated the size of the for a young industry such as China's. Serious prob- agricultural distortions using the available series on lems in enforcing trademarks, patents, and trade domestic and international prices. Unfortunately, secrets in particular could inhibit the transfer of the results obtained have varied widely.7 And while technology in China and the development of inno- Carter (2000) and Martin (2001) felt that WTO vative domestic businesses. For China, with its cur- accession would require relatively little liberaliza- rent low allocation of resources to research and tion in China, Schmidhuber (2001) and many oth- development, laws protecting domestic innovations ers believed that dramatic changes would be are of limited benefit. Key issues for the future needed. Clearly, a new approach was required, include enhancing pricing regulations on pharma- especially because the policy consequences are so ceuticals as patent protection becomes stronger and large. developing a broader competition policy regime to Huang and Rozelle (2002) adopted a new deal with abuses of IPRs such as monopoly pricing approach by basing their analysis of policy impacts and restrictive licensing arrangements. on detailed interviews with participants in China's agricultural markets rather than on available price series. Their approach provides a much clearer Sectoral Impacts of Accession indication of the implications of agricultural trade To evaluate the impacts of changes in China's trade policies for product prices and of the real-world policy arising from WTO accession, we first assessed impacts of policies. They show, for example, that a the policy stance prior to accession and then traced major source of the discrepancies in earlier research the implications of the policy measures being intro- is differences in quality between domestic products duced. This section describes the evidence on agri- and those traded internationally. They also identify culture, manufacturing, and services in turn. features of the trade regime, such as export subsi- dies on maize and cotton, that have important impacts on product markets. Agriculture What then do China's accession commitments Many authors have raised concerns about the impact imply for agricultural markets? For those products of WTO accession on China's agriculture and the protected by ad valorem tariffs, the implications are 8 East Asia Integrates straightforward. A reduction in the tariff indicates way in which rice exports were taxed slightly in 2001. directly the reduction in the domestic price of the After accession, the rate of protection is expected to good, and this change, together with information remain the same, because accession to WTO does on the slope of the import demand curve, can be not require reductions in negative protection that is used to estimate the cost of protection. For those administered by state trading or an export tax. For products protected by both a tariff and an export wheat, protection averaged an estimated 12 percent subsidy, it may be necessary to consider changes in before accession--much lower than would be sug- both variables. For products that are being pro- gested by simple price comparisons. After accession, tected, or are to be protected, using tariff-rate quo- this rate of protection need not be greatly reduced tas, the analysis becomes much more complex. For on average, because it seems likely that wheat them, the impact of a tariff reduction depends imports will exceed the tariff-rate quota reasonably greatly on whether it is the within-quota or the out- often (Martin 2001), allowing the imposition of a of-quota tariff that determines the price of the tariff of up to 65 percent. For maize, the rate of pro- good. And where the quota will be filled in some tection before accession was higher, at 32 percent, years but not in others, the average rate of protec- because of an export subsidy. After accession, the tion may be a combination of the two tariff rates. level of import protection need not change greatly Table 1.1 shows some key assessments of the on average, because there is a significant probability implications of the level of protection and the that the tariff-rate quota will bind by the end of the changes associated with WTO accession. The statu- decade (Martin 2001). But the maize export subsidy tory tariff rates for 1998 used by Schmidhuber must be abolished, implying a potentially substantial (2001) and others are given in the first column; the reduction in the price support given to maize. estimates by Huang and Rozelle (2002) of protec- Oilseeds present a different case. Here, the prin- tion in 2001 are given in the second. The third col- cipal form of protection has been a tariff, and the umn shows the anticipated average rates of protec- tariff is being reduced substantially. For sugar, the tion after accession, taking into account the reforms protection provided must be halved to meet required by accession and likely market outcomes. China's commitments to a bound tariff of 20 per- For rice, wheat, and maize, the adjustments cent. On cotton, import protection will not change required by WTO accession will be much smaller greatly, but export subsidies such as the 10 percent than those suggested by analyses based on the statu- export subsidy observed in 2001 are ruled out in tory rates of protection. For rice, Huang and Rozelle the future. For livestock and meat, protection could (2002) estimate that the average rate of protection remain negative as a consequence of export restric- was slightly negative before accession, implying that tions to markets such as Hong Kong (China). Pro- China's system of state trading for rice operated in a tection of dairy products can be expected to decline TABLE 1.1 Some Measures of Import Protection in China's Agriculture (percent) 1998 statutory 2001 actual Postaccession tariffs protection protection Rice 127 ­3.3 ­3.3 Wheat 133 12.0 12.0 Maize 130 32.0 32.0 Vegetables and fruits 15 ­4.0 ­4.0 Oilseeds 132 20.0 3.0 Sugar 30 40.0 20.0 Cotton 3 17.0 20.0 Livestock and meat 35 ­15.0 ­15.0 Dairy 46 30.0 11.0 Source: Average statutory rates taken from Schmidhuber (2001) and www.chinavista.com. See Huang and Rozelle (2002) for estimates of protection. China's Accession to the WTO: Impacts on China 9 to meet China's tariff-binding commitments. increase the pressure for their reform. Looking The reductions in protection shown in Table 1.1 ahead, China is unlikely to be satisfied if traders (also Figure 4 of the Overview) are just one set of such as Japan and the Republic of Korea are allowed possible outcomes in a situation in which rates of to maintain agricultural protection rates much agricultural protection can vary substantially, par- higher than its own. The increasing openness of ticularly if import levels exceed the tariff-rate quo- China's market for agricultural products will create tas. However, the reductions show that China's pol- much bigger markets for land-intensive products icymakers will still be able to exercise significant such as rice from some East Asian countries, and for discretion after accession. specialty, high-quality products from almost all Removing the negative protection from labor- countries. But strong policy efforts by China and intensive products would be consistent with WTO other agricultural exporters in the current WTO rules and is likely to be particularly beneficial for negotiations will be required to expand market employment in rural areas, as well as for economic access in the current highly protected agricultural efficiency. By taking part in the WTO agricultural markets. negotiations being conducted under the Doha China will clearly have a strong interest in Development Agenda, China could potentially expanding its access to export markets for its labor- reinforce these benefits by opening large, and cur- intensive agricultural products. Particularly for rently highly protected, markets for its labor-inten- perishable products such as fruits and vegetables, sive agricultural exports. Unfortunately, the high many of the logical markets are in East Asia. It is rates of agricultural protection that arose when vitally important that China's trading partners pro- GATT rules on agriculture were extremely weak vide opportunities for China to expand its exports mean that China faces barriers to its agricultural of these products, which are so important for creat- exports that are four times as high as those it faces ing employment for relatively poor rural people. on its other merchandise exports (Martin 2001). Analysis by Yu and Frandsen (2002) suggests that China's accession to the WTO with relatively low agricultural liberalization by member countries of tariff bindings on agricultural products prevents it the Organisation for Economic Co-operation and from following the path of ever-increasing agricul- Development (OECD) would benefit China and tural protection that other high-growth East Asian improve its agricultural trade balance. China's economies have followed (Anderson and others rights as a WTO Member at least give it an oppor- 1986). Even China's highest tariff bindings, of 65 tunity to insist that adequate procedures be fol- percent, provide only one-tenth of the protection lowed and to limit the duration of measures such as currently observed for wheat and rice in Japan safeguards. It would be particularly unfortunate if (Martin 2002). Given China's size, this situation is China's trading partners resorted to the use of stan- extremely important for world agricultural mar- dards as a means of restricting access. Such mea- kets; had China gone down the road of Japan, the sures are not transparent and are divisive in nature. world market for rice and other agricultural com- If China elects to remove its negative protection modities would be permanently depressed. from key commodities such as rice, vegetables, and China's commitment to a low-protection agri- meats, Anderson, Huang, and Ianchovichina cultural regime will greatly reduce the costs of (2002) estimate that the returns to unskilled rural achieving successful economic development. More- labor and to farmland would rise slightly, so that over, it will force future policymakers to focus on the overall impact of the accession on rural wages policies such as improvements in rural education would be ­0.5 percent instead of ­0.7 percent. If, on and reductions in barriers to labor mobility that the other hand, the in-quota tariff rates were uni- will deal effectively with the problems of rural formly applied, then the returns to farm factors poverty in China--rather than resort to commod- would deteriorate by about the same amount. ity price distortions, which serve, at best, as a short- term, palliative measure. Industrial Products China's agricultural trade commitments throw into stark contrast the costly and inefficient regimes Most of the adjustment needed in industrial tariffs in neighboring economies (OECD 2002a) and has already occurred, and what remains are an 10 East Asia Integrates expansion in both imports and exports and a likely ing trade in both final goods and production inputs painful restructuring of some key industries. between China and its regional neighbors. China's China substantially reduced its tariffs on manu- imports will grow because of the reductions in pro- factures during the 1990s: weighted average tariffs tection, and exports will expand because of on manufactures fell from 46.5 percent in 1992 to induced drops in production costs in China. 25 percent in 1995 and to around 13 percent by the Industry in China will face substantial adjustment time of accession in 2001. With full implementation pressures in key products such as automobiles, bever- of China's accession commitments, these tariffs will ages, and tobacco, where external protection is being fall to 6.9 percent. The 6 percent reduction in aver- substantially reduced. Restructuring of scale-inten- age tariffs that remains to be implemented is impor- sive industries such as the motor vehicle will be essen- tant, but small relative to the 33 percent reduction tial and can generate substantial productivity gains. since 1992 or the 12 percent reduction since 1995 Ianchovichina and Martin (2002) estimate that (Ianchovichina and Martin 2002). China's imports of merchandise will rise by 17 per- The largest reductions in industrial tariffs are cent on average as a result of the WTO-induced tar- now required in beverages and tobacco (a 28 per- iff reductions after 2001, with particularly large cent reduction from 2001 levels) and in automobiles increases in products such as beverages and tobacco (a 15 percent reduction from 2001 levels)--see (112 percent). Much of the benefit is likely to Table 1.2. While large, these reductions are much accrue to China's neighbors in East Asia. smaller than those already undertaken since 1995. The reduction in protection to the automobile sec- Motor Vehicles. In the absence of reform and tor is particularly important given the high profile restructuring, China's output of motor vehicles of this industry and its linkages throughout the would likely drop sharply in the postaccession years, economy (see the next section on motor vehicles). notwithstanding strong increases in domestic Other industries in which substantial reductions demand and the shift in China's comparative advan- in tariffs will be needed include textiles, clothing, tage to more capital- and skill-intensive products electronics, and light manufactures. Many of these such as motor vehicles (Francois 2002). China's auto industries are relatively labor-intensive ones in industry has been shaped by protective policies that which China has a comparative advantage and in have encouraged inefficient production and allowed which liberalization will help to maintain efficiency for market segmentation; indeed, most plants are and competitiveness. operating well below global standards for efficient The new tariff concessions in manufacturing production. Development of a healthy and interna- will provide enormous opportunities for expand- tionally competitive industry will require consider- TABLE 1.2 Protection of Industrial Sectors in China (percent) 1995 2001 Postaccession Processed food 20.1 26.2 9.9 Beverages and tobacco 137.2 43.2 15.6 Extractive industries 3.4 1.0 0.6 Textiles 56.0 21.6 8.9 Apparel 76.1 23.7 14.9 Light manufactures 32.3 12.3 8.4 Petrochemicals 20.2 12.8 7.1 Metals 17.4 8.9 5.7 Automobiles 123.1 28.9 13.8 Electronics 24.4 10.3 2.3 Other manufactures 22.0 12.9 6.6 Total manufactures 25.3 13.5 6.9 Source: Ianchovichina and Martin (2002). China's Accession to the WTO: Impacts on China 11 able restructuring of this industry (Harwit 2001). benefits of the liberalization associated with acces- We estimate that restructuring in the motor vehicle sion, and they downplay some important actual industry to achieve scale economies in final assem- and potential elements of trade policy. bly could reduce production costs by about 20 per- A key omission of the analysis is the possibility cent (Francois 2002). Such restructuring would that antidumping and safeguard measures might be more than reverse the negative impact on output of applied against China. Another is the increasing use the reduction in protection from 1997 levels, and of measures of this type by China. The introduc- would allow the industry to expand dramatically as tion of the product-specific safeguards against China's growth and shifting comparative advantage China is particularly important in this respect shift resources into sectors such as motor vehicles. If because no such measure specifically targeted the industry is restructured successfully, exports of China before China's accession. The risk that China finished motor vehicles would increase rapidly, will increase its use of antidumping and safeguard resulting in an increase in total exports of vehicles measures beyond the currently high levels is also of and parts of more than US$4 billion a year.8 concern for development policy. Such an action Increases in the efficiency of the final assembly would be a triumph of a rules-focused approach to industry relative to the production of intermediate WTO implementation--"it is legal therefore, we parts are likely to increase the demand for imported should do it!"--over the sharp focus on develop- parts substantially, with their share of total parts ris- ment that has characterized China's trade reform ing from 39 percent to 52 percent. agenda since the beginning of the reform era. Clearly, profound adjustments are in store throughout the auto industry. Considerable painful Services restructuring, such as the closure of inefficient plants in many cities, will be required. However, Trade in services was a key area in China's WTO there seems to be no alternative if China is to move accession negotiations, and China's commitments to an efficient and internationally competitive represent perhaps the most thoroughgoing liberal- motor vehicle sector. ization of the services trade ever undertaken in the WTO (Mattoo 2002). Its range of offers is Policy Concerns. What are the likely effects of the extremely broad, although some commitments large reductions in the barriers facing China's involve restrictions on ownership, business scope, exports of textiles and clothing to Europe, the United or region. Critical sectors such as telecommunica- States, Canada, and Norway? Because these barriers tions, logistics, and finance are going to face are implemented through export quotas--that is, an renewed competition and are likely to see a burst of exporter is required to purchase an export quota, or innovation and productivity growth as they are to forgo the opportunity to sell quotas it has been restructured. Many middle-income developing allocated by the government--they impose a cost on countries, such as Thailand, that have experience in exports that is analogous to an export tax. Based on dealing with these problems are likely to find sub- detailed information on quota prices, this tax is esti- stantial market opportunities. mated to be around 15 percent for clothing and 10 An important feature of China's commitments percent for textiles (see www.chinaquota.com). For is that they focus on market access and do not dis- particular products, the export tax equivalent of criminate between domestic and foreign suppliers. these measures is much higher. However, China's commitments are carefully The analysis by Ianchovichina and Martin crafted: in cross-border trade and in services con- (2002) focuses largely on tariffs (given the great sumed abroad, the number of sectors with guaran- uncertainty about the protective impacts of nontar- teed unrestricted access is smaller than in most iff barriers such as designated trading, quotas, and other countries, and in establishment trade it is licenses in China, and their limited remaining cov- essentially zero. erage9), although it also considers the export quotas on textiles and clothing. Omission of the effects of Restrictions on Service Activities. Restrictions on removing nontariff barriers means that the results the form of business establishments, such as give something of a lower-bound estimate of the requirements for joint ventures, business scope, 12 East Asia Integrates and geographic scope, have a long history in China, that all parts of China have access to telecommuni- and have often been justified as a means to acquire cations services. technology or obtain a share of monopoly rents. As Mattoo (2002) points out, rules requiring busi- Logistics. China's WTO commitments on logistics nesses to form joint ventures may in fact inhibit the involve a range of General Agreement on Trade in transfer of technology. A more thoroughgoing Services (GATS) service sectors, including packaging approach to the problem of monopoly rents would and courier services, maritime and rail transport, ensure that competition between firms, whether freight forwarding, and storage and warehousing domestic or foreign, would eliminate these rents. services. Logistics costs are disproportionately high Many restrictions on the geographic scope of in China, and service quality is lower than desirable, service suppliers, such as restrictions on the cities in part an enduring legacy of the planned economy that can be served by insurance companies, date (Findlay and Luo 2002). Logistics-related costs are from an era when it was believed that experimenta- widely seen as accounting for up to 30­40 percent of tion with market-oriented approaches needed to be the wholesale costs of manufactured goods in China, isolated because of the inconsistencies between, for as against 5­20 percent in the United States (see example, planned and market prices. There seems Findlay and Luo 2002 and Tanzer 2001). High logis- to be much less need for such policies now that the tics costs are a particularly important problem for operation of market economies is so much better people in the poorer parts of China. Their ability to understood in China. Restrictions such as those in trade and, consequently, their real incomes are sig- the WTO agreement that confine foreign ventures nificantly reduced by these excessive costs. China has to five cities for five years--insurance is one exam- committed itself to increasing competition in some ple--might encourage agglomeration of these key areas, including road transport, rail transport, activities in the favored cities, and that agglomera- warehousing, and freight forwarding. The breadth of tion will not be reversed when the restrictions are these commitments also provides a much stronger subsequently lifted. This situation may reduce the basis for development of integrated third-party10 opportunities for other parts of China, such as inte- logistics firms able to reduce the costs and increase rior cities with a potential comparative advantage the quality of logistics services in China. in these activities, to get started in these activities. China has made substantial progress toward There appear to be good developmental reasons meeting its commitments on logistics services. for China to phase out its geographic restrictions Findlay and Luo (2002) believe that the costs of a more quickly than required by the WTO commit- wide range of goods and services might be reduced ments, given the risks of exacerbating the already by about 10 percent from current levels--a huge substantial inequalities between coastal and inte- savings that would raise incomes substantially in rior provinces. China's commitments do not pre- China, increase opportunities for exports to China, vent the authorities from moving ahead faster than and strengthen China's competitiveness in export required, as they have done frequently in the past. markets. Regulations on services may have various goals, Achieving the full potential of logistics in China including making competition work, improving will require not just freer trade but also regulatory the availability of information to consumers, and reforms--to remove discrimination against partic- ensuring universal service. Making competition ular types of enterprises, to separate local adminis- work is particularly important in network indus- trations from enterprises, and to eliminate local tries such as telecommunications, where the domi- protectionism. In addition, substantial investments nant incumbent firms frequently do not find it in in infrastructure will be needed to improve the their interests to allow new firms to interconnect to timeliness, and reduce the cost, of providing logis- their networks. Improving the availability of infor- tics services. mation is particularly important in financial ser- vices, where lenders often do not have enough Telecommunications. China's commitments in information on the prospects and repayment telecommunications take on particular importance capacity of borrowers. Developing efficient provi- given that China is expected to be the largest market sions on universal service is important for ensuring for telecommunications in the world by 2010 China's Accession to the WTO: Impacts on China 13 (Pangestu and Mrongowius 2002). These commit- cent of the gross domestic product (GDP) in 2001, ments allow foreign entry to a wide range of activities and China's securities market was the second largest currently closed to foreign investment. Furthermore, in Asia, with a market capitalization approaching 45 this entry takes place in a sector that was monopo- percent of GDP. However, four large state banks lized by China Telecom until 1994 and is currently accounted for 67 percent of deposits and 56 percent dominated by a small number of state-owned firms. of total financial assets. A key problem in the finan- In basic telecommunications, China has com- cial sector was insolvency and a lack of profitability mitted to the disciplines of the WTO reference in the four big state banks, with more than a quarter paper on the regulatory framework for telecommu- of outstanding loans nonperforming. Although nications (WTO 1996). These disciplines aim to there have been improvements, these banks were ensure a competitive environment that allows widely regarded as having limited ability to make interconnection between systems under reasonable lending decisions based on repayment ability and and nondiscriminatory conditions and that allows credit risk. Furthermore, the administered spreads for universal service provisions. They also require between deposit and lending rates were very nar- the existence of a regulator independent of the tele- row, reducing the profitability of the banks and their com provider and set criteria for licensing of entry incentives to lend to many dynamic areas of the and allocation of scarce commodities such as the economy. Many foreign banks were active in China, mobile telephone spectrum. but they accounted for less than 3 percent of assets In the context of a basic telecommunications and only 0.1 percent of deposits. system governed by the WTO regulatory frame- China's accession commitments allowed foreign work, China's other GATS commitments cover banks to enter the foreign currency business imme- value added services such as voice mail and online diately, the local currency business for enterprises information services; mobile voice and data ser- after two years, and the local currency business for vices; and domestic and international services such all clients after five years. In the securities markets, as private leased circuit services. Most of these ser- foreign securities were allowed immediate entry vices are initially subject to a combination of own- into trading in foreign currency­denominated ership restrictions and geographic restrictions securities and entry within three years to joint ven- within China. Although the geographic restrictions tures trading in domestic shares. Insurance firms will be phased out over several years, China has not were allowed immediate entry into reinsurance and committed to allowing more than 49 percent for- life insurance and entry after two years into health eign ownership in important areas such as mobile insurance, pension insurance, and annuities. telephone service. Allowing higher levels of foreign The objective of these radical reform commit- ownership would, of course, be consistent with ments appears to have been the vital one of bring- China's GATS obligations. ing about technical innovation that will reduce the Allowing the telecommunications sector to cost of financial services and help to stimulate make its maximum contribution to China's devel- growth. It seems clear that considerable further opment will require further reforms of the regula- reform, and probably some re-capitalization of the tory framework. Important issues will include state banks, will be required if this goal is to be ensuring the independence of the regulator, ensur- achieved without the collapse of the state banks or a ing that interconnection works adequately, and broader financial crisis. making pricing regulations more flexible. Impacts of Reforms on the Financial Services. China's WTO commitments on Economy financial services call for radical changes in the structure of the system. This was a remarkable out- Because the reforms undertaken by China in come given that, prior to accession, China's financial response to WTO accession are broad-ranging and sector was widely regarded as being far from ready their economy-wide interactions extensive, it is for a major increase in competition from abroad. important to evaluate these reforms on an econ- The size of the Chinese financial market was sig- omy-wide basis. They need to be seen in the context nificant, with deposits at banks reaching 150 per- of the dramatic changes already under way in 14 East Asia Integrates China's industrial structure, output, and trade pat- accession into two components: those associated terns related to more capital- and skill-intensive with the liberalization undertaken between 1995 and goods--changes made in response to changes in 2001 in preparation for accession and those stem- demand for China's exports, high rates of invest- ming from the liberalization to be undertaken after ment, and rapid growth in educational levels. 2001 to meet China's accession commitments.13 Ianchovichina and Martin (2002) have analyzed the impacts of liberalization associated with WTO Exports accession in agriculture, manufactures, and ser- vices.11 For this analysis, they use a special variant of The liberalization associated with WTO accession the Global Trade Analysis Project (GTAP) model will speed the growth in China's trade relative to (Hertel 1997), developed to take into account the output. The total volume of exports is projected to importance to China of duty exemptions for inputs rise by 17 percent as a consequence of the liberal- into export production (see Ianchovichina 2003). In ization after 2001 (Table 1.3). this model, liberalization of trade reduces the costs The fastest-growing exports are clothing, which of imports and the costs of production for the are projected to double after 2001 in response to domestic market by reducing the costs of inputs and the abolishment of the export quotas on clothing. the costs of production of all types through reduc- Exports of most agricultural products will also rise, tions in the prices of nontraded goods and factors. reflecting the decline in input costs to agriculture The second type of cost reduction is frequently and in the persistent difficulties workers face in termed a real exchange rate depreciation.12 Ian- migrating out of agriculture. Exports of plant- chovichina and Martin (2002) divide the impacts of based fibers (predominantly cotton) are projected TABLE 1.3 Impacts of Reduction in Protection Required by WTO Accession from 2001 Tariff Levels, China Trade balance Output Employment Exports Imports (millions (%) (%) (%) (%) of US$) Rice ­2.1 ­2.3 6.1 ­7.1 64 Wheat ­2.0 ­2.3 18.9 ­10.1 174 Feed grains ­2.3 ­2.6 ­77.8 ­2.4 ­596 Vegetables and fruits ­3.4 ­3.7 14.6 ­6.3 214 Oilseeds ­7.9 ­8.4 29.8 20.9 ­789 Sugar ­6.5 ­7.4 13.9 24.1 ­73 Plant-based fibers 15.8 16.4 ­51.8 7.7 ­189 Livestock and meat 1.3 1.1 15.5 ­8.9 837 Dairy ­2.0 ­2.4 13.5 23.8 ­143 Other food ­5.9 ­6.4 11.4 62.6 ­3,460 Beverages and tobacco ­33.0 ­33.1 9.7 112.4 ­14,222 Extractive industries ­1.0 ­1.3 7.5 ­4.4 2,088 Textiles 15.6 15.5 32.7 38.5 ­10,366 Apparel 57.3 56.1 105.8 30.9 49,690 Light manufacturing 3.7 3.7 5.9 6.8 1,786 Petrochemical industry ­2.3 ­2.3 3.1 11.8 ­8,810 Metals ­2.1 ­2.1 3.7 6.8 ­1,893 Autos 1.4 ­2.2 27.7 24.0 516 Electronics 0.6 0.4 6.7 6.8 453 Other manufactures ­2.1 ­2.2 4.1 18.9 ­11,291 Total 1.0 0 16.8 17.3 717 Source: Ianchovichina and Martin (2002). China's Accession to the WTO: Impacts on China 15 to fall, reflecting the increased demand for cotton taken between 1995 and 2001 is estimated to yield a in the export production of textiles and clothing continuing gain of $30 billion a year. The smaller and the abolishment of export subsidies. Feed grain reduction in protection between 2001 and the end exports will also fall, because of abolishment of the of the implementation period will generate a export subsidy on exports of maize. Exports of smaller gain of $10 billion a year.14 automobiles will rise substantially, because auto As for the distribution of the projected gains, the production becomes more efficient as it exploits wages of skilled and unskilled urban workers will economies of scale and becomes more exposed to rise modestly, and the wages of unskilled farm international competition workers will decline by 0.7 percent in real terms. None of these projected increases in exports The distribution of gains is examined further in the takes into account the possible benefits to China of next section. being able to expand its market access through par- ticipation in WTO market accession negotiations, Effects of WTO Accession at the such as those currently under way as part of the Household Level Doha Development Agenda. The expansion in tex- tile and clothing exports is, in fact, a delayed benefit The simplest approach to capturing the effects of from the Uruguay Round, previously denied to WTO accession at the household level requires an China as a nonmember of the WTO. assessment of the changes in the prices consumers pay; the changes in the prices that owners of labor, capital, and other factors receive for their resources; Imports and the effects of accession on the government's China's imports are projected to rise in a range of ability to provide transfers or public goods. In addi- sectors in which trade barriers have been substan- tion, it is useful to assess the ability of households tially reduced--including beverages and tobacco, to adjust to the changes resulting from accession, processed food, textiles, clothing, oilseeds, dairy perhaps by changing their activities. products, and sugar. Imports of beverages and tobacco will rise the most because of the sharp Effects on Rural and Urban Households reductions in tariffs on these commodities. Imports of services are also projected to rise substantially, Because ongoing work on trade and poverty has because trade liberalization in services is expected found that impacts felt through factor markets are to reduce import barriers. consistently more important than impacts felt through consumer prices, it is certainly worthwhile to examine the impacts of trade liberalization on fac- Employment tor markets. This is particularly the case in China, The projected movements of labor between sectors with its large income differences between urban and are generally quite small relative to the changes in rural workers and explicit policy barriers to the trade patterns. The biggest postaccession change in movement of labor between urban and rural sectors. employment is likely to be in clothing, with a rise of China's labor markets are affected adversely by a more than 50 percent after 2001. Employment in range of regulations that restrict the movement of the textile sector and in plant fibers used in textile workers from rural to urban areas. These include production will also rise to meet the demand from the hukou system of residence permits, which regu- the clothing sector. Trade reform will lead to small lates movement between urban and rural employ- reductions in employment in most agricultural sec- ment. Another inhibiting feature is the restrictions tors and in manufacturing sectors such as petro- on the sale of farmland usage rights. The effect of chemicals, metals, and automobiles. these rights is that farm families that move perma- nently out of agriculture may have to relinquish their land rights without compensation. Welfare Gains The earnings of rural and urban households dif- Trade liberalization is a source of substantial overall fer substantially, even for households with labor of welfare gains for China. The liberalization under- the same skill level (Sicular and Zhao 2002). The 16 East Asia Integrates large barriers that exist between urban and rural reflect a combination of falling rural wages and labor markets, particularly for poorer households, increases in the prices of goods consumed by these depress income levels and make it more difficult for households. workers to respond to changes in economic oppor- Hertel, Zhai, and Wang (2002) also provide tunities. Even so, the barriers are not absolute: for important insights into the impacts of trade reform labor supplied to nonagricultural activities by rural on poverty.16 Because of limitations on the avail- households, elasticities of supply with respect to ability of household data, they focus on Liaoning, wage rate differentials are on the order of two for Sichuan, and Guangdong, three relatively diverse unskilled workers and closer to three for skilled provinces. Their scenario for agricultural tariffs is workers (Sicular and Zhao 2002). similar to that in the Ianchovichina and Martin Shi Xinzheng (2002) has examined the substan- (2002) study, but focuses on an agricultural policy tial differences between the earnings of urban and determined by the tariff-rate quotas. In aggregate, rural households and the extent to which these dif- their conclusions are much more optimistic than ferences are caused by the formal barriers between those of Chen and Ravallion in Chapter 8, showing China's urban and rural labor markets. Only virtually all households benefiting from trade between 30 and 40 percent of the total difference reform. However, their results suggest that inequal- between rural and urban wage rates can be explained ity worsens after accession: urban households ben- by the hukou policy. However, this policy component efit substantially more than rural households, and of the gap, which is a fundamentally important rural households with diverse income sources ben- building block for assessments of the implications of efit more than those that depend only on farming. liberalization, is difficult to estimate precisely. The results of these quantitative analyses are A key political concern throughout the imple- highly stylized, because they assume that enterprises mentation period is likely to be the impacts of and households adjust successfully to the changes in accession on the formerly favored workers in state- incentives that are created by WTO accession. As owned enterprises. These enterprises are experienc- OECD (2002b) has pointed out, to make these ing greater competition from imports as a result of changes successfully China will have to strengthen abolishment of the barriers that formerly discour- its economic system in areas such as enterprise gov- aged competition from foreign firms in the domes- ernance and reform of the banking system. tic market. The vulnerability of the rural sector to trade reform arises from restrictions on the movement of labor out of agriculture when returns fall. These Effects on Poverty restrictions inhibit the adjustment needed after From a poverty perspective, it seems likely that the accession and increase the vulnerability of poor peo- central issues of concern will be in rural areas. ple to downturns in agricultural prices. Reform of Poverty is most prevalent in the rural areas of China, these restrictions has now become urgent, because and rural workers face barriers in moving into the the partial liberalization of agriculture is increasing sectors that are likely to expand as a result of China's the pressures on workers to leave agriculture. ongoing growth and liberalization. Two studies eval- Hertel, Zhai, and Wang (2002) consider the uate the impacts of China's trade liberalization on implications of two key complementary policies poverty, and both find that the rural sector is more that might be used to deal with the problems of vulnerable than the urban sector to this reform. poverty in rural areas: (1) reducing the barriers to In Chapter 8 of this volume, Chen and Ravallion mobility of rural labor into nonagricultural consider the impact of WTO accession on income employment, and (2) improving the availability of distribution and poverty.15 They find a sharp con- education in rural areas. They conclude that reduc- trast between the experience of urban and rural tions in the barriers to mobility of labor out of agri- households. Most urban households, and particu- culture would be an important antidote to the larly those who are relatively poor, gain from WTO increases in inequality between rural and urban accession. This is not the case for rural households, sectors experienced after accession, with diversified the poorest of which experience noticeable reduc- rural households becoming the largest gainers from tions in their living standards. These reductions a policy package combining WTO accession with China's Accession to the WTO: Impacts on China 17 reductions in the barriers to rural labor mobility. nonmarket economy provisions that countries are And over the longer term there is another poten- permitted to invoke against China for up to 15 tially important source of gains to the rural sector: years are likely to result in the imposition of China will have an opportunity to press for greater antidumping duties substantially higher than those market access for its labor-intensive exports. invoked against other countries--and in a situation Expansion of educational opportunities could in which China faces seven times as many also have powerful beneficial effects in helping antidumping actions per dollar of exports as the unskilled workers in both urban and rural areas United States. The product-specific transitional (Hertel, Zhai, and Wang 2002; Ianchovichina and safeguard provisions, applicable for the next 12 Martin 2002). And improvements in agricultural years against China alone, are a new form of pro- technology would potentially yield large benefits to tection. The associated provisions on trade diver- poor rural households that are able to adopt the sion lack even basic procedural restraints and pose new production techniques (Ianchovichina and a potentially serious threat to China's export devel- Martin 2002). opment. China may be tempted to retaliate, partic- ularly with antidumping actions of its own, but retaliation would damage China more than its Social Protection trading partners. A better option would be to seek China's network of social protection measures is reform of the WTO rules in these areas. still quite underdeveloped, with a stark dichotomy Our analysis suggests that agriculture is being between urban and rural systems, a focus on reduc- liberalized less than was suggested by some earlier ing absolute poverty, and a high degree of decen- studies that began with the assumption that agri- tralization in financing. The system in urban areas cultural tariffs were being reduced from their statu- is relatively comprehensive, but that in rural areas is tory levels. However, significant liberalization has seriously deficient. In many areas "the social safety occurred in areas such as maize, cotton, and sugar, net is full of holes" (Hussain 2002). and it appears that there will be significant adjust- China's social welfare measures are likely to ment pressures in these industries. Yet opportuni- remain partial in coverage and are not a viable tool ties will emerge to expand exports of some labor- for large-scale poverty reduction. Even so, particu- intensive exports as part of a broader policy reform lar attention needs to be given to strengthening and by seeking increases in agricultural market basic safety nets for rural residents. Here, a feasible access in the Doha negotiations. first step might be to extend a mechanism like the The industrial sector will face substantial adjust- urban unemployment insurance schemes to wage ment pressures in key sectors such as automobiles, employees in township and village enterprises. beverages, and tobacco, where external protection is being substantially reduced. Restructuring of scale-intensive sectors such as the automobile sec- Conclusions tor will be essential, and it can generate substantial China's WTO accession agreement provides a com- productivity gains. Overall, however, most of the prehensive road map for many aspects of reform in adjustment in this industry has already occurred, China's trade regime. In particular, it calls for sub- and what remains involves an expansion in both stantially reducing protection, for strengthening imports and exports. the protection of intellectual property rights, and China's GATS commitments represent perhaps for adopting a framework of trade rules at home the most thorough-going liberalization of services and abroad. However, the agreement is not all- trade ever undertaken in the GATT. Its range of inclusive, and China needs to keep a strong focus offers is extremely broad, although some commit- on its development needs, maintaining the per- ments involve restrictions on ownership, business spective that has guided the steady transformation scope, or region. Critical sectors such as telecom- of its trade regime and economy, generally, from munications, logistics, and the financial sector will planned to market. confront renewed competition, and they are likely A central concern in the accession agreement is to see a burst of innovation and productivity the provisions on antidumping and safeguards. The growth as they are restructured. 18 East Asia Integrates China's labor markets are affected adversely by a 2. For example, tariffs may be set at any level below China's range of regulations, such as the hukou system of tariff bindings. Or the Agreement on Trade-Related Aspects of Intellectual Property Rights (TRIPS) may be used in a residence permits regulating movement between way that stimulates the development of a knowledge-based urban and rural employment. This system, and and more productive economy, or simply in a way that related labor market policies, inhibit the adjust- results in transfers to industrial countries. Or, similarly, WTO rules on contingent protection may be used in a way ment needed after accession and increase the vul- that is extremely damaging to open trade and economic nerability of poor people to downturns in agricul- development, or they may be used (or not used) in a way tural prices. Other features of the labor market, that is more consistent with economic efficiency and equity. 3. See Gertler (2002) for an outline of these principles. such as the "tie" to the land where households have 4. Although the implementation period extends to 2010, use rights to land but cannot sell it because prop- almost all of the reduction will have been completed by erty rights are not sufficiently well defined, restrict 2005. the mobility of labor out of agriculture. 5. For example, in agriculture, China had to accept a limit of 8.5 percent on de minimis domestic support, as opposed to The two studies described in this chapter of the the usual 10 percent limit for developing countries. impacts of trade reform on poverty both find that 6. See Maskus (2002). His assessment of China's intellectual the rural sector is more vulnerable to this reform property regime considers the provisions for patents, trade- marks, trade secrets, and copyrights. He compares China's than is the urban population. One study concludes rules on intellectual property rights with international that WTO accession is generally beneficial to urban benchmarks for middle-income developing countries, households but that most rural households lose in using data from interviews with market participants. 7. Huang, Chen, and Rozelle (1999), for example, estimated the short run. The other study provides a more the protection applying to rice, wheat, and maize in the upbeat picture, but one in which rural households mid-1990s at 4, 20, and 25 percent, respectively. By con- are more subject to negative shocks such as those trast, Tuan and Cheng (1999) estimated these protection resulting from reductions in protection on some rates to be ­29, 62, and 15 percent, respectively. Carter (2000: 80) relied on producer price data and found gener- agricultural commodities. A major conclusion is ally negative price distortions. that the key to mitigating these problems is 8. All dollar amounts are current U.S. dollars. reforms such as reduction of barriers to the move- 9. The frequency of import licenses, in particular, has fallen from about two-thirds of tariff lines in the late 1980s to less ment of labor out of agriculture and investment in than one-twentieth in 2001 (Lardy 2002). rural education. 10. Service suppliers that are independent of both the sender The network of social protection measures in and the recipient of the goods. 11. Ianchovichina and Martin (2002) approximate very China is quite underdeveloped and a constraint on crudely the estimated liberalization of services by halving China's ability to grow while dealing with widely held the barriers to trade in these activities estimated by concerns about the need to compensate the potential François (2002). They also analyze the opportunities that "losers" from the policy reforms. Policymakers arise from the elimination of the quotas against China's (and other countries') exports of textiles and clothing. should pay particular attention to strengthening the Their analysis takes into account China's important export- social welfare systems available to rural residents. processing arrangements and builds on the labor market The dramatic increases in China's imports and studies undertaken by Sicular and Zhao (2002) and by Shi (2002) and on the analysis of automobile industry restruc- exports associated with its accession create enor- turing undertaken by Francois (2002). The resulting mous opportunities for China's trading partners to changes in the specification of their model greatly increase benefit, both as suppliers of exports to this rapidly the realism of their analysis and have important implica- tions for their results. growing market and as beneficiaries of lower- 12. The reduction in the price of nontraded goods is often priced and higher-quality imports. In addition, called a real exchange rate depreciation. The analysis many countries with similar ranges of exports will described in this chapter considers only the impact of liber- face greater competition in third-country markets alization on goods markets, and ignores possible impacts on real exchange rates operating through induced increases for many products, particularly textiles and cloth- in investment, an issue addressed in Chapter 2. McKibbin ing, where the removal of quotas in 2005 will lift an and Tang (2000) conclude that this stimulus to investment enormous burden from China's exporters. associated with reductions in the cost of capital and expan- sion of exports may be sufficient to reverse the long-run tendency toward real exchange rate depreciation, and may Endnotes actually require real exchange rate appreciation, during the first years after accession. 1. See www.worldbank.org/trade for those studies and more 13. The choice of 1995 as a starting period is somewhat arbi- detailed references to the literature. trary, given that China reformed its trade regime through- China's Accession to the WTO: Impacts on China 19 out the 1990s. Nevertheless, 1995 was an important turn- Fewsmith, J. 2001. "The Political and Social Implications of ing point, when China had to forgo its hopes of resuming China's Accession to the WTO." China Quarterly 167 (Sep- its seat in the GATT and apply as a newcomer to the tember): 573­91. World Trade Organization, under a process much more Findlay, Christopher, and W. Luo. 2002. "Logistics in China: focused on the commercial implications of the accession Accession to the WTO and Its Implications." Paper presented package. to Conference on China's Accession to WTO, Policy Reform 14. These projections of export and income growth are very and Poverty Reduction, Beijing, June 28­29. lower-bound estimates, because they ignore the benefits Francois, Joseph F. 2002. "The Motor Vehicle Sector in China from abolishing nontariff barriers, because they involve and WTO Accession." Paper presented to Conference on serious aggregation biases, and because models of this type China's Accession to WTO, Policy Reform and Poverty appear to understate greatly the implications of major Reduction, Beijing, June 28­29. trade liberalizations (Kehoe 2002). Gertler, J. 2002. "What China's WTO Accession Is All About." 15. Their study draws on the simulation model results provided by Paper presented to Conference on China's Accession to Ianchovichina and Martin (2002) and uses a sample of WTO, Policy Reform and Poverty Reduction, Beijing, June 84,000 households--17,000 urban and 67,000 rural--from 28­29. National Bureau of Statistics surveys. The price impacts from Gilbert, J., and T. Wahl. 2002. "Applied General Equilibrium the GTAP analysis are applied to the households' initial Assessments of Trade Liberalization in China." World Econ- income and expenditure shares, taking into account the omy 25(5): 697­731. impacts of these price changes on the prices households Harwit, E. 2001. "The Impact of WTO Membership on the must pay for their consumption goods and purchases of Automobile Industry in China." China Quarterly 167: inputs, and the prices they receive for their sales of goods and 655­70. of labor and other factors. The loss of government revenue Hertel, T. W., ed. 1997. Global Trade Analysis: Modeling and from falling tariffs is restored very simply in the model Applications. Cambridge: Cambridge University Press. experiment by increasing the price of all consumption goods Hertel, T. W., F. Zhai, and Z. Wang. 2002. "Implications of WTO through a consumption tax used to maintain government Accession for Poverty in China." Paper presented to Confer- revenues. This approach to measuring the impact likely over- ence on China's Accession to WTO, Policy Reform and states the extent to which tax rates would need to rise, Poverty Reduction, Beijing, June 28­29. because government revenues from tariffs as a percentage of Huang, J., and S. Rozelle. 2002. "The Nature of Distortions to the value of imports were only about half of the 8 percent Agricultural Incentives in China and Implications of WTO that would have been implied by the statutory tariff rates. Accession." Paper presented at Conference on WTO Acces- 16. Their model of the Chinese economy takes into account sion, Policy Reform and Poverty Reduction in China, World important features such as the duty exemptions for inter- Bank, Beijing, June 28­29. mediate goods used in the production of exports. Huang, J., C. Chen, and S. Rozelle. 1999."Reform, Trade Liberal- ization and Their Impacts on China's Agriculture." Paper presented at International Agricultural Trade Research Con- References sortium Symposium on China's Agricultural Trade Policy: Issues, Analysis and Global Consequences, San Francisco, The word processed describes informally reproduced works that June 25­26. may not be commonly available through libraries. Hussain, A. 2002. "Coping and Adapting to Job Losses and Fall in Farm Earnings." Paper presented at Conference on WTO Anderson, K., and others. 1986. The Political Economy of Agricul- Accession, Policy Reform and Poverty Reduction in China, tural Protection: East Asia in International Perspective. World Bank, Beijing, June 28­29. Boston, London, and Sydney: Allen and Unwin. Ianchovichina, Elena. 2003. "GTAP-DD: A Model for Analyzing Anderson, K., J. Huang, and E. Ianchovichina. 2002. "Impact of Trade Reforms in the Presence of Duty Drawbacks." GTAP China's WTO Accession on Agriculture and Rural-Urban Technical Paper No. 21. Available at Income Inequality." Paper presented to conference on www.agecon.purdue.edu/GTAP/techpapr/index.htm. China's Accession to WTO, Policy Reform and Poverty Ianchovichina, Elena, and William J. Martin 2002. "Economic Reduction, Beijing, June 28­29. Impacts of China's Accession to the WTO." Paper presented Anderson, S., and C. Lau. 2001. "Hedging Hopes with Fears in at Conference on WTO Accession, Policy Reform and China's Accession to the World Trade Organization: The Poverty Reduction in China, World Bank, Beijing, June Transitional Special Product Safeguard for Chinese Exports." 28­29. Powell, Goldstein, Frazer, and Murphy, LLP, Geneva. Johnson, D. G. 2000. "The WTO and Agriculture in China." Processed. China Economic Review 11: 402­4. Bhattasali, D., Li Shantong, and W. Martin, eds. Forthcoming. Kehoe, T. 2002. "How Well Did Computable General Equilib- WTO Accession, Policy Reform and Poverty Reduction in rium Models Predict the Effects of NAFTA?" Paper pre- China. Washington, D.C.: World Bank and Oxford Univer- sented to Conference on General Equilibrium Modeling, sity Press. Cowles Foundation, Yale University, April. Carter, C. 2000. "China's Trade Integration and Impacts on Fac- Lardy, Nicholas R. 2002. Integrating China into the Global Econ- tor Markets." In China's Agriculture in the International Trad- omy. Washington, D.C.: Brookings Institution Press. ing System. OECD (Organisation for Economic Co-opera- Lin, J. Y. 2000. "WTO Accession and China's Agriculture." China tion and Development) Proceedings, April. Economic Review 11: 405­8. Chun, Chang, B. Fleisher, and E. Parker. 2001. "The Impact of Martin, William J. 2001. "Implications of Reform and WTO China's Entry into the WTO: Overview." China Economic Accession for China's Agricultural Policies." Economics of Review 11 (4): 319­22. Transition 9 (3): 717­42. 20 East Asia Integrates ------. 2002. "Reforming Japan's Rice Policies: Comment." In Panitchpakdi, Supachai, and M. Clifford. 2002. China and the R. Stern, ed., Issues and Options for Japan-US Trade Policies. WTO: Changing China, Changing World Trade. Singapore: Ann Arbor: University of Michigan Press. John Wiley (Asia). Martin, William J., and Elena Ianchovichina. 2001."Implications Schmidhuber, J. 2001. "Changes in China's Agricultural Trade of China's Accession to the World Trade Organization for Policy Regime: Impacts on Agricultural Production, Con- China and the WTO." World Economy 24 (9). sumption, Prices, and Trade." In China's Agriculture in the Maskus, Keith E. 2002."Intellectual Property Rights in the WTO International Trading System. OECD (Organisation for Accession Package: Assessing China's Reforms." Paper pre- Economic Co-operation and Development) Proceedings, sented to Conference on China's Accession to WTO, Policy April. Reform and Poverty Reduction, Beijing, June 28­29. Shi Xinzheng. 2002. "Empirical Research on Urban-Rural Mattoo, Aaditya. 2002. "China's Accession to the WTO: The Ser- Income Differentials: A Case of China." Peking University. vices Dimension." Paper presented to Conference on China's Processed. Accession to WTO, Policy Reform and Poverty Reduction, Sicular, T., and Y. Zhao. 2002. "Employment, Earnings and Beijing, June 28­29. Poverty in Rural China: A Microeconomic Analysis." Paper McKibbin, W., and K. Tang. 2000. "Trade and Financial Reform presented to Conference on China's Accession to WTO, Pol- in China: Impacts on the World Economy." World Economy icy Reform and Poverty Reduction, Beijing, June 28­29. 23 (8). Tanzer, A. 2001. "Chinese Walls." Forbes. November 12. Messerlin, Patrick. 2002. "China in the WTO: Antidumping and Tuan, F., and G. Cheng. 1999. "A Review of China's Agricultural Safeguard." Paper presented at Conference on WTO Acces- Trade Policy." Paper presented at International Agricultural sion, Policy Reform and Poverty Reduction in China, World Trade Research Consortium Symposium on China's Agricul- Bank, Beijing, June 28­29, 2002. tural Trade Policy: Issues, Analysis and Global Conse- OECD (Organisation for Economic Co-operation and Develop- quences, San Francisco, June 25­26. ment). 2002a. Agricultural Policies in OECD Countries: Mon- WTO (World Trade Organization). 1996. "Negotiating Group itoring and Outlook. Paris. on Basic Telecommunications." Telecommunications Ser- ------. 2002b. China in the World Economy: The Domestic Policy vices Reference Paper. Geneva, April 24. Challenges. Paris. Yu, Wusheng, and S. Frandsen. 2002. "China's WTO Commit- Pangestu, Mari, and D. Mrongowius. 2002. "Telecommunication ments in Agriculture: Does the Impact Depend on OECD Services in China: Facing the Challenges of WTO Accession." Agricultural Policies?" Paper presented at Fifth Annual Con- Paper presented to Conference on China's Accession to WTO, ference on Global Economic Analysis, Taipei. Available at Policy Reform and Poverty Reduction, Beijing, June 28­29. www.gtap.org. 2 Regional Impact of China's Accession to the WTO Elena Ianchovichina Sethaput Suthiwart-Narueput Min Zhao China's accession to the World Trade Organization The task of assessing the impact of China and its (WTO) and deeper integration into the world accession to the WTO on East Asia presents an economy present important opportunities and enormous challenge because of the complexity of challenges for the East Asia region. China's role in the changes stemming from accession and the diffi- the region is unrivaled. First, its economy is large in culty in specifying clearly what would have hap- absolute terms--constituting half the economy of pened had China not acceded to the WTO. Thus Asia, according to measures of purchasing power. the goal of this chapter is less to predict detailed Second, China has rapidly expanded its trade, changes than to provide a framework for under- almost tripling its share of global exports and more standing the impact of WTO accession on the than doubling its share of global imports over the region's economies and to offer a broad assessment period 1990­2002 and absorbing a fast-growing of this impact for different countries and country share of exports from East Asia over the last decade. groupings. The findings are drawn from quantita- Third, although its capital account is not fully con- tive and qualitative analysis, including improved vertible, China is important both as an investment computable general equilibrium (CGE) modeling destination and as a lender in global capital mar- and partial equilibrium studies, which are detailed kets. It is the world's largest host country for for- on the Web site associated with this volume.2 eign direct investment (FDI) and the largest capital Because the analysis is at an aggregated level, read- supplier among developing countries.1 And look- ers should supplement the findings here with infor- ing ahead, China will remain an important driver mation from subsectoral case studies before mak- of change in East Asia. With WTO accession, it will ing policy decisions.3 continue to open its markets to other countries' Finally, although this chapter focuses on the exports and improve its business climate. economic impact of China's accession to the WTO, The authors gratefully acknowledge helpful comments from Lu Ding of the National University of Singapore, Nattapong Thongpakde of the Thailand Development Research Institute, William J. Martin of the World Bank, and participants at seminars held at the Institute for Southeast Asian Studies of the National University of Singapore, the Thailand Develop- ment Research Institute in Bangkok, and the World Bank office in Jakarta, as well as helpful discussions with Arvind Pana- gariya of the University of Maryland at College Park and the excellent research assistance of Wallada Atsavasirilert. 21 22 East Asia Integrates the ultimate impact on the region will be driven not tries of East Asia is the subject of the next section, just by economics, but also by broader political and it is followed by our conclusions. economy considerations. China's accession to the WTO is a signal of its greater interdependence with Channels of Impact the world and its increasing international presence, which also will have repercussions in the region. Over the next decade, China's growth and greater China's expanding influence as a regional power integration into the world economy will have major may be viewed with growing concern by its neigh- effects on the region. Other countries in East Asia bors, and that influence raises issues about China's will feel the impact of China's WTO accession future leadership on economic matters in the through four main channels: region, a role traditionally played by Japan. The chapter is organized as follows. The next 1. Expansion of markets in China for their exports section briefly describes the major channels 2. Increased imports from China into their domes- through which the impacts of China's WTO acces- tic markets sion will be felt, followed by a section assessing the 3. Competition with China in third-country markets impact of accession on the newly industrializing 4. Expansion of foreign direct investment in China economies (NIEs) in East Asia--in particular, their and, potentially, outward foreign investment growing opportunities in China's markets and the from China. effects of the evolution of global production net- works. The next section assesses the scope for East Increased Access to China's Domestic Markets Asian middle-income developing countries to expand exports to China, as well as the challenges In the 1990s, exports to China spurred growth not they face from competition with China in third- only in the newly industrializing economies but also country markets. It also outlines how domestic in the developing countries of East Asia (Figure 2.1). markets are likely to be affected by increases in the The countries of the Association of South East Asia supply of exports from China and by patterns of Nations (ASEAN) have increased their exports to foreign direct investment. The impact of China's China by 390 percent and expanded their share in accession to the WTO on the lower-income coun- China's total imports from 6 percent to 9 percent. FIGURE 2.1 Exports to China from East Asia Compared with Exports to Other Partners East Asia NIEs: export markets Developing East Asia: export markets (as % of total exports) (as % of total exports) Percent Percent 40 40 35 35 30 30 25 25 20 20 15 15 10 10 5 5 0 0 198019821984198619881990199219941996199820002002 198019821984198619881990199219941996199820002002 China United States China United States Japan Other East Asia Japan Other East Asia Note: NIEs = newly industrializing countries. "East Asia Note: "Developing East Asia" refers to Cambodia, NIEs" refers to Hong Kong (China), Rep. of Korea, Singa- Indonesia, Lao PDR, Malaysia, Mongolia, Myanmar, pore, Taiwan (China). Philippines, Thailand, Vietnam. Source: IMF, Direction of Trade Statistics (various years); Source: IMF, Direction of Trade Statistics (various years); data reported from exporter country accounts. data reported from exporter country accounts. Regional Impact of China's Accession to the WTO 23 Continued growth in China's huge domestic export more not only to China, but also to the markets will fuel further export growth for the newly industrializing economies in East Asia, world and the economies of the region. In many whose own demand for imports has grown as a sectors, China's WTO accession only adds a little to result of the accession. Both the direct and indirect the already vigorous projected growth of these effects of increased access to China's markets will be markets (Figure 2.2). Nonetheless, accession will important for regional trade. cause several significant shifts. China's substantial commitments to liberalize trade in services repre- Increased Imports from China sent the most significant part of the accession pack- age (Mattoo 2002), providing national treatment to China's accession to the WTO will be accompanied foreign-funded firms and greater opportunities for by cuts in its export prices, increasing China's exporters of services. In manufacturing, China's appeal as an efficient supplier of intermediate commitments to abolish nontariff barriers and inputs. China's preaccession reforms have already reduce its import tariffs from 13.3 percent in 2001 improved the competitiveness of its exports and to 6.8 percent by the end of the implementation benefited its closest trading partners. Trade inten- period in 20104 will fuel further industrial restruc- sity indexes for 1985 and 2001 (Ng and Yeats 2003) turing. Some sectors such as the motor vehicle and suggest that trade between individual East Asian high-end manufacturing will be affected signifi- countries and China has intensified sharply since cantly by rationalization and industrial restructur- 1985. Consequently, most East Asian economies are ing. In agriculture, too, China's imports are pro- expected to benefit from further cuts in export jected to grow substantially, although the effect of prices as China continues to implement WTO- WTO-related reforms on agricultural output and related reforms over the next few years. The bene- imports is much smaller than projected by earlier fits to these countries will be evident in both studies because protection of many farm products increased output and welfare. is expected to remain virtually unchanged by the A growing segment of imports from China will end of the implementation period (Huang and be inputs in production processes, not just finished Rozelle 2002). consumer goods (Figure 2.3). China is increasingly Growth in the region's exports will also be fueled a central player in production networks. Although by the increased demand from those major trading Japan remains an important center of production- partners that benefit directly from China's acces- sharing operations in East Asia, originating about sion. The developing countries of East Asia will one-third of all regional exports of components for FIGURE 2.2 China's Imports, 1995 and 2005 China: Percent change in imports between 1995 and 2005 Foodgrains Feedgrains Oilseeds Meat and livestock Dairy Other food Beverages/tobacco times 10 Extractive industries Textiles Wearing apparel, net Wood and paper Metals Electronics Other manufactures 0 100 200 300 400 500 600 700 Without accession With accession Note: This figure captures the effects of WTO accession, encompassing all reforms since 1995. Ianchovichina and Walmsley (2003), whose work has been drawn on for the remainder of this chapter, have estimated the effects of WTO accession for 2001­10. Source: Ianchovichina and Martin (2001). 24 East Asia Integrates FIGURE 2.3 China's Growing Role in Production Networks Share of components in China's Parts and components imports from China, 2001 exports relative to share of Value of components in world trade (RCA) imports from Share in RCA China total 2.0 (millions of US$) (%) 1.5 Japan 5,587 36.7 Hong Kong (China) 13,556 45.5 1.0 Korea, Rep. of 1,695 21.1 Singapore 1,989 11.9 0.5 Taiwan (China) 1,308 18.9 109 4.9 0 Indonesia Office Telecom Electrical Malaysia 1,394 12.8 machinery equipment machinery Philippines 170 2.4 1996 2001 Thailand 1,030 13.3 Note: RCA = revealed comparative advantage. Source: Ng and Yeats (2003: Table 17.1). Source: Ng and Yeats (2003). assembly, China is finding niches; its exports of other high-technology products. Competition with parts and components increased by almost US$20 China has brought unit prices down, but thus far billion5 from 1996 to 2001. By 2001 China was other East Asian exporters have maintained their exporting more than $20 billion in parts and com- market shares in the United States and Japan. In ponents to others parts of emerging East Asia, rep- Japan, developing East Asian countries have even resenting up to 20 percent of those countries' parts managed to enlarge their market share slightly, and components trade (Ng and Yeats 2003). Thus while China has captured market share at the imports from China represent an opportunity for expense of the United States. the rest of emerging East Asia to benefit from In the future, competition is set to intensify for China's growing role in global production networks. two reasons. First, the United States, Canada, and To realize the full benefits of China's lower the European Union (EU) will abolish their import export prices, it will be important that countries quotas on Chinese textiles and apparel by 2005.6 resist pressures to protect their domestic producers China will then become a formidable competitor, and that they avoid imposing excessive safeguard especially in the apparel sector,7 pushing prices measures for this purpose. Pressures to do so are down in these important third-country markets. growing in several countries. However, succumbing Second, China will lower its own import tariffs on to them will only prolong the adjustments needed inputs for manufacturing. The effect of these tariff to realize regional comparative advantages, and will reductions on the real exchange rate will be lower distract policymakers from facilitating the transi- costs of both traded and nontraded inputs for tion of workers by putting in place the appropriate China's manufacturers. This development will labor market and safety net policies and programs. make China's products more competitive as imports, putting pressure on domestic producers in the countries that import them. Increased Competition in Third-Country Markets Competition with China in third-country markets Shifts in Investment Patterns will intensify as a result of China's accession. This competition will present a challenge for many WTO accession is likely to increase foreign direct countries, especially those with a similar compara- investment in China, as trade liberalization lowers tive advantage in labor-intensive goods. production costs and the price of capital goods and Southeast Asia competes with China in world increases the rental rates, resulting in rising returns markets for manufactures, especially labor-inten- to capital in China.8 Meanwhile, the liberalization sive products, and increasingly in higher value of rules on investment should ease flows of foreign added manufactures such as semiconductors and direct investment into previously restricted sectors Regional Impact of China's Accession to the WTO 25 such as services and automobile production. Given only on the competitiveness of the country that the substantial productivity gap that exists between exports the final product, but also on those neigh- local and foreign firms, the new FDI flows are likely boring countries that contribute various compo- to raise China's productivity. In apparel and nents at different stages of the production process. footwear, for example, the adoption of foreign This situation will create an incentive for locating technology raises productivity by 30­62 percent in investments in the countries that are part of the collective enterprises and 20­59 percent in state regional production network in which China is enterprises (Claro 2001). playing an increasingly central role. China's accession is also likely to set off changes Fifth, whereas in the past China drew heavily on in regional trade and production patterns, the the overseas Chinese community as a source of effects of which will be felt over the medium to long FDI, with WTO accession China is now able to term. The question for other East Asian economies draw on global capital markets for both FDI and is whether their own FDI inflows will increase or portfolio investment. Therefore, competition decrease as a result. It is difficult to answer this between China and other countries in the region question because much will depend on the policy for FDI may actually weaken. responses of individual countries. In addition, Last, but not least, the determinants of FDI are some factors will work simultaneously to deter- evolving over time. Agglomeration effects are mine the net regional impact of China's WTO becoming more important relative to the tradi- accession on FDI flows. tional determinants of FDI such as market size and Although countries that lose competitiveness labor costs (UNCTAD 2001). China resists easy may see declining returns to capital and FDI definitions of its comparative advantage in terms of diverted to China, several effects are likely to coun- high-tech or low-wage manufacturing, and its teract this negative impact on FDI. First, increased comparative advantage also may change apprecia- productivity and trade liberalization in China both bly in response to WTO accession. Its current com- increase the country's demand for imports and parative advantage in labor-intensive products sug- raise investment and welfare in China's trading gests that there is more scope for export partners. Martin (1993) shows that a productivity specialization vis-à-vis the newly industrializing shock in manufactures and services, for which there economies than the developing East Asian coun- is a lot of two-way trade, is more likely to raise wel- tries. However, this situation is likely to change as fare in a country's trading partners than is trade East Asia's export structure evolves.9 Over time, liberalization. The technological advance accompa- China is also likely to shift and extend its compara- nying China's liberalization will improve the coun- tive advantage from land- and labor-intensive try's competitiveness, but also increase the coun- products and low-end manufacturing to higher- try's demand for imports. end products as a result of trade-induced produc- Second, investment liberalization in China will tivity gains and savings in transaction costs from make it possible for multinational firms to further the reforms spurred by WTO accession. Increased rationalize their production processes within East productivity from WTO-related liberalization and Asia. The relief of local content requirements under higher wages--initially in China's coastal regions-- trade-related investment measures (TRIMs) will has been shown to reduce the competitive threat to encourage these firms to relocate some segments of the lower- and middle-income economies of their production from China to other countries in Southeast Asia,10 while increasing the threat to the the region. newly industrializing and more advanced middle- Third, in some sectors China's neighbors may income economies of East Asia. receive FDI flows that complement those going to According to a recent analysis, the impact of China. The scope for export specialization varies accession is significantly larger than the estimated with the degree of complementarity between China static gain, if productivity increases in services and and other countries of East Asia. high-end manufacturing are taken into account. Fourth, as FDI creates more backward and for- Every percentage point increase in productivity from ward linkages among countries in the region, the expanded competition and foreign entry in China's competitiveness of Asian products will depend not services sector implies a welfare gain for China of 26 East Asia Integrates $10 billion (1997 US$) and an increase in its gross textile exports to India, the Philippines, Vietnam, domestic product (GDP) of 2.2 percent--equivalent and other South and Southeast Asian economies is to the total estimated static gain from China's WTO also expected to slow as these countries' garment accession (Ianchovichina and Walmsley 2003). industries contract in the face of competition with China in third-country markets. China's demand for intermediate inputs and Impact of WTO Accession on Newly final products is expected to drive the export growth Industrializing East Asia and Japan of these products from its neighbors. Examples On balance, the industrialized and newly industrial- include metals and petrochemicals from Korea; izing economies in East Asia will benefit from electronics and other manufactures from Singapore; China's accession to the WTO.11 As important sup- light manufactures, petrochemicals, machinery, pliers of materials to China, these countries will see equipment, and electronics from Taiwan (China); an improvement in their terms of trade.12 In both and metals, petrochemicals, oil, and other extractive Japan and the NIEs, most of the projected increase industry products from Japan. In electronics, China in production will be driven by expanded exports to is expected to seek its additional inputs from the China. Although these countries are well positioned countries that get the largest tariff reductions-- to gain from China's accession to WTO, many of the India, the United States, and other South Asian trends that have developed from China's growing countries, and to a lesser extent Hong Kong and Sin- role in world trade are already evident. gapore--rather than from Korea, where tariffs on Japan, Taiwan (China), the Republic of Korea, electronic products are already low.13 The potential and Hong Kong (China) are expected to raise their for specialization and complementary intraindustry output of textiles in response to increased demand trade could be significant. As shown in Figure 2.4, from China's expanding garment industry. Their China already represents an important source of own garment industries will be squeezed, however, parts and components for the NIEs. particularly in the markets--North America and In automobile production, China's current plans the EU--where the quotas on Chinese textile and for restructuring its industry will make it a more apparel exports are removed. The growth of their efficient assembler of motor vehicles and eventually FIGURE 2.4 Impact of China's WTO Accession on Japan and East Asia's NIEs, 2001­10 (cumulative changes in output of selected sectors relative to baseline, millions of 1997 US$) Volume changes 8,000 6,000 4,000 2,000 0 ­2,000 ­4,000 ­6,000 ­8,000 Beverage and Textiles Apparel Processing Autos Electronics Other tobacco industries manufactures Japan Taiwan (China) Hong Kong (China) Singapore Korea, Rep. of Source: Ianchovichina and Walmsley (2003). Regional Impact of China's Accession to the WTO 27 an exporter (François and Spinanger 2002). This administrative environment for foreign investment prospect could provoke a major reorganization of are likely to be addressed in line with WTO acces- the industry across the region. Our analysis projects sion, although competition in markets for goods a contraction of automobile production in Japan and services is expected to intensify. and the NIEs. While the benefits to Japan and the NIEs from The NIEs will benefit from China's increased China's WTO accession are clearly positive, over demand for services. Indeed, accession is likely to time China's likely shift to extend its comparative increase demand for all types of services, including advantage into higher-end products implies that transport and communications, which these these countries will face increasing competitive economies are well positioned to provide.And it will challenges, as is already happening in selected sub- enhance the role of Hong Kong as a financial center sectors and regions. serving the mainland's investment needs and pro- viding investment services (Deutsche Bank 2001). Impact of WTO Accession on the Investment flows into the NIEs are unlikely to Middle-Income Developing fall as the result of China's WTO accession. The Countries of East Asia returns to capital in these countries will rise relative to the baseline (though not by as much as in Overall, China's trade liberalization and growth China).The reason is that the NIEs are suppliers of will have a mixed impact on these countries-- raw materials to China, rather than competitors of Indonesia, Malaysia, the Philippines, and Thailand China, and thus the prices received for their exports (Figure 2.5). Some factors are common to these will tend to rise. countries. China's market presents sizable opportu- For Japan, the major impact of China's WTO nities, including links with China as it plays an accession is that China will become a more attrac- increasingly larger role in global production net- tive destination for Japanese investments. After five works. At the same time, similarity in export struc- years of strong growth, Japanese FDI to China ture suggests the threat of increased competition in dropped substantially in 1996­99 (Ministry of third-country markets from increased exports from Finance, Japan), less because of Japanese firms' China, with the impact of accession itself concen- financial difficulties than because of a difficult trated in a few sectors--notably, apparel and tex- market environment in China (Marukawa 2001). tiles, where adjustments are likely. Generally, and Some of the concerns about China's weak legal and increasingly over time, trade-induced changes in FIGURE 2.5 Impact of China's WTO Accession on Developing Economies of East Asia, 2001­10 (cumulative changes in output of selected sectors relative to baseline, millions of 1997 US$) Volume changes 1,500 1,000 500 0 ­500 ­1,000 ­1,500 Agriculture Textiles Apparel Processing Autos Electronics Other industries manufactures Indonesia Vietnam Malaysia Philippines Thailand Source: Ianchovichina and Walmsley (2003). 28 East Asia Integrates China's productivity will make China a stronger Some of the top exports from Indonesia to partner and will ensure that the benefits from mar- China will also benefit significantly from reduc- ket opportunities outweigh third-country market tions in nontariff barriers (Table 2.2). Nearly 20 competitive challenges. The details, however, will percent of Indonesian exports to China are in depend on the situation and response of the indi- products--especially palm oil, rubber, and vidual countries. processed oil--on which quantitative restrictions will be lifted (Table 2.2). Palm oil exports, in partic- ular, should gain significantly from the relaxation Indonesia of quantitative restrictions (QRs).16 Palm oil, which In the decade ahead, China's liberalization and eco- was previously subject to an import license, is now nomic expansion open several opportunities for subject to tariff-rate quotas (TRQs) with accession. Indonesia but also pose risks, both as an exporter Initial quota levels are set significantly above and as a destination for investment. The challenge import levels prior to accession and are slated to will be to manage the transition to realize those increase significantly by 2005.17 TRQs on rubber opportunities. will take effect in 2004, when quotas are elimi- Indonesia's exports to China stand to benefit nated.18 Urea also will be subject to a TRQ, with a significantly from China's growth and liberaliza- final quota level in 2006 of more than 2.5 times the tion.14 Much of the tariff reduction under China's initial quota level and an in-quota tariff rate of 4 WTO accession will occur by 2004 (Table 2.1), but percent.19 Most categories of timber will also be lib- China is free to offer tariff rates below these bind- eralized within three years of accession. ings. For example, under the "Early Harvest" aspect Indonesia's overall export volume will continue of the Framework Agreement on China-ASEAN to grow. Demand from China will be compounded Comprehensive Economic Cooperation, which by increased import demand from China's closest would establish an ASEAN-China Free Trade Area, trading partners--Japan, the NIEs, the EU, and the a range of Indonesian exports will face lower tariffs United States--which themselves have benefited than other countries' exports.15 from China's growth and WTO accession. However, TABLE 2.1 Weighted Average Tariffs Facing Exports to China: Indonesia and Thailand, 2001­08 (percent) 2001 2002 2003 2004 2005 2006 2007 2008 Indonesia 8.20 8.66 5.82 4.91 4.73 4.65 4.58 4.51 Thailand 10.19 8.91 8.41 4.86 4.63 4.43 4.25 4.05 Note: The weighted average tariff was calculated on the top 100 exported products to China, representing more than 85 percent of each country's exports to that country. The in-quota tariff rate is used for tariff-rate quota products. Sources: UN Comtrade data; China's WTO accession agreement; authors' calculations. TABLE 2.2 Shares of Exports to China Affected by Lifting of Quantitative Restrictions (QRs): Indonesia and Thailand (percent of total exports, 1995­99) Percent of exports Country facing QRs Key products (% of total exports) Indonesia 16.5 Palm oil (7.3), rubber (2.9), processed oil (0.6) Thailand 30.6 Rice (9.24), rubber (11.8), cane sugar (5.8) Sources: UN Comtrade data; China's WTO accession agreement; authors' calculations. Regional Impact of China's Accession to the WTO 29 competitive pressures are likely to shift Indonesia's those of competing imports from China were manufacturing structure. Exports from electronics deemed to be most at risk. For Thailand and and other manufacturing industries are projected Indonesia, the results indicate that 15­25 percent of to increase, as are exports of land-intensive prod- exports to the United States and Japan are at risk ucts--food and feed grains and wood products-- from growing competition from China (Tables 2.3 and other raw materials, including energy prod- and 2.4). ucts.20 Indonesia will also increase its exports of Indonesian exports are more at risk in the oilseeds, sugar, and cotton to China. United States than in Japan. Among Indonesia's At the same time, the increasing similarity in exports to the United States in 1995­99, the top five export structure between China and Indonesia sug- products23 accounted for more than 30 percent, gests a potential threat from competition in third- and the top 100 products for 84 percent of total country markets.21 Indonesia's apparel sector will exports. Of Indonesian exports to the United have to adjust. Indonesian apparel is sold mainly in States, almost half (47 percent) appear to have unit North America and Western Europe, where it will values close to those of competitor products from be particularly vulnerable to the abolishment of China, and 16 percent are in the at-risk cate- quotas on Chinese apparel exports. With a possible gories--those in which unit values are close to decline in apparel output, Indonesia's textile sector those of China and in which China supplies more will also come under pressure, although there will than 10 percent of imports (Table 2.3). Using the 5 be niches for its expansion. percent threshold, nearly a quarter of Indonesian The risk to exports to third-country markets exports to the United States are at risk, including was confirmed by a market-by-market and prod- Indonesia's second and third most important uct-by-product analysis carried out for Indonesia. export products to the United States.24 For this analysis, we identified "exports at risk" to In the Japanese market, Indonesia's export the U.S. and Japanese markets based on their structure has been even more concentrated than in importance to the exporting country and the extent the United States, with the top five products25 to which they compete with similar products from accounting for more than 44 percent and the top China.22 Exports in product categories that are 100 for more than 85 percent of total exports in characterized by both a high share of Chinese 1995­99. About 11 percent of Indonesia's exports imports (at least 5 percent) and unit values close to are in at-risk categories, with unit values close to TABLE 2.3 Market-by-Market and Product-by-Product Analysis of Indonesian Exports: China's Market Share and Closeness of Unit Value (UV) as Proxies for Potential Risk China's import share China's import share < 10% > 10% < 5% > 5% Share in U.S. market (%) (n = 165 products) UV close 30.5 16.2 22.8% 24.0 UV not close 16.1 21.1 14.1% 23.1 Share of market in analysis 83.9 83.9 Share in Japanese market(%) (n = 188 products) UV close 23.8 11.0 18.5 23.0 UV not close 24.6 11.7 23.3 13.1 Share of market in analysis 71.2 77.9 Note: The table shows calculations using two (arbitrary) threshold levels for Chinese imports: 5 percent and 10 percent of the total import market. Sources: UN Comtrade data; China's WTO accession agreement; authors' calculations. 30 East Asia Integrates TABLE 2.4 Market-by-Market and Product-by-Product Analysis of Thai Exports: China's Market Share and Closeness of Unit Value (UV) as Proxies for Potential Risk China's import share China's import share < 10% > 10% < 5% > 5% Share in U.S. market (%) (n = 150 products) UV close 23.9 8.7 18.1 14.5 UV not close 22.1 16.4 19.2 19.4 Share of market in analysis 71.1 71.1 Share in Japanese market (%) (n = 189 products) UV close 28.0 20.8 23.9 24.9 UV not close 17.2 16.8 11.9 22.1 Share of market in analysis 82.8 82.8 Note: The table shows calculations using two (arbitrary) threshold levels for Chinese imports: 5 percent and 10 percent of the total import market. Sources: UN Comtrade data; China's WTO accession agreement; authors' calculations. those of competing Chinese products and with expand in China and Indonesia simultaneously. China supplying more than 10 percent of imports. And, like other ASEAN middle-income countries, Using the 5 percent threshold, 23 percent of Indonesia has the potential to develop its role as a Indonesia's exports are potentially at risk; they supplier of specific parts to an automobile produc- include some of the most important exports to tion network, given the restructuring of the indus- Japan.26 try now taking place in the region.27 Indonesia is likely to adjust to the increased Indonesia will need to tailor its strategy to grasp competition in apparel by increasing its specializa- the opportunities for increases in trade and invest- tion in wood and paper products and in light, high- ment flows if it is to offset the declines that are pro- end, and other manufactures. Exports of these jected in its exports to the EU, Japan, and the products could increase, primarily in response to United States. Key elements will be measures to rising demand in China. restore investor confidence and increase competi- China's increased attractiveness for investors tiveness. Meanwhile, Indonesia will have to avoid again provides both opportunities to Indonesia to protecting its domestic producers with excessive expand its intraindustry trade and potential safeguard measures in order to facilitate an adjust- threats. One approach to assessing Indonesia's abil- ment in the manufacturing sector that responds to ity to prevent a decline in foreign direct investment the opportunities in China's markets. Measures is to look at the sectoral changes likely to occur. If such as the recently introduced temporary safe- the sectors that are expanding (contracting) are the guards against garment imports will only prolong ones that already account for a large share of FDI, the adjustments that Indonesia needs to make to then FDI is likely to increase (decrease) as well. Sec- realize its regional comparative advantages. tors that are growing less rapidly in Indonesia-- apparel and textiles as well as assembly opera- Malaysia tions--have received foreign direct investment in the past, whereas any expansion in agriculture Among the developing middle-income countries in likely involves little FDI. Opportunities exist for East Asia, Malaysia is likely to be the one most pos- Indonesia to participate in global production net- itively affected by China's accession to the WTO.28 works--cosmetics, machinery, and audiovisual Overall, the effect on trade will be positive, as equipment, for example--in which FDI may Malaysia becomes a more important trading part- Regional Impact of China's Accession to the WTO 31 ner to China. Flows of foreign investment will not sion to the WTO, China will demand more raw be much affected, because the returns to invest- materials and land-intensive products, and the ment in Malaysia are expected to rise only slightly Philippines is well positioned to increase exports of over the accession period. food and feed grains, cottons, sugar, vegetables, and Malaysia's exports of wood products and other fruits in response.32 Food grain exports, in particu- manufactures will likely increase as a result of WTO lar, are expected to rise substantially as a result of accession, relative to the baseline, in response to China's accession. Output and exports of light increased demand for these products in China and manufactures are expected to grow faster with than Taiwan (China).29 Malaysia's overall volume of agri- without China's accession. Demand from China cultural exports is not projected to change much, but will be compounded by increased import demand exports of oilseeds, sugar, livestock, and cotton to from China's closest trading partners, which them- China and Taiwan (China) are likely to increase as a selves will have benefited. result of China's WTO accession. Demand from At the same time, because the Philippines has China will be compounded by a higher import comparative advantages in many areas in which demand from China's closest trading partners-- China is strong, China's WTO accession will inten- Japan, the NIEs, the EU, and the United States-- sify competition between the two countries. Com- which themselves have benefited from China's petition with China will be particularly acute in growth and WTO accession. apparel produced for North America and Western Just like the other developing economies of East Europe because of the removal of quotas on Asia, Malaysia is expected to lose some of its share in China's textiles and apparel in these markets. Over- the world apparel market as a result of China's all, the impact of accession on the Philippines' out- accession.30 In textiles, however, output will be little put of apparel and textiles will be negative. Even so, affected, and while the textile exports of all the other the Philippines occupies certain niches within the developing countries in the region will contract as a export market that may be less exposed to Chinese result of China's accession, Malaysia's exports will competition. hold their ground, buttressed by increased demand The impact of China's accession on the Philip- from China's apparel industry. In this way, Malaysia pines will also depend on investment flows. As in is more similar to the NIEs in the region. Indonesia and Thailand, the sectors that are grow- Malaysia's involvement in intraregional produc- ing less rapidly--apparel, textiles, and assembly tion networks and the improving quality of its labor operations--have been targets of FDI in the past. and infrastructure position it well as a destination However, looking ahead, the Philippines is poised for foreign direct investment that complements to take part in global production networks in elec- investment in China. By 2001 Malaysia was already tronics, machinery and equipment, processing, and exporting $1.3 billion to China and importing $1.4 light manufacturing, where FDI may well expand billion from China in parts and components (see in China and the Philippines simultaneously. In Ng and Yeats 2003). During China's recent expan- 2001 imports from China and exports to China in sion, Malaysia received significant FDI in the infor- parts and components totaled $170 million and mation technology-related and electronics indus- $342 million, respectively. This represents only a 2 tries; FDI in electronics grew 40 percent from 1997 percent share of the total parts and components to 2001. In the automobile sector, Malaysia may trade for the Philippines, suggesting considerable position itself to benefit from participating in inter- opportunity for expansion (Ng and Yeats 2003). national production networks--for example, at The Philippines'ability to move up the value chain present it is producing steering gears. Alternatively, and capture more benefits from China's accession it may maintain its more protectionist stand, which will depend on what strategy the country adopts. may harm its prospects to attract additional FDI. Many of the sectors that will suffer from China's accession to the WTO use unskilled labor intensively, while those sectors that will expand use land or The Philippines skilled labor more intensively. Wages of unskilled Philippine exports to China stand to benefit from workers may well come under pressure. The potential China's growth and liberalization.31 With its acces- impact on urban inequality and vulnerability in the 32 East Asia Integrates short term will have to be carefully monitored and industries are projected to increase, as are exports addressed, with increasing pressure to facilitate the to China of land-intensive products--oilseeds, adjustment process in the labor market. sugar, and wood products. Cotton production may expand as demand for cotton increases in response to the expansion of Taiwan's textile sector. China's Thailand accession will have a negative impact on growth in Thailand is well positioned to expand its manufac- the apparel and textile sectors.36 turing base, despite the fact that China's WTO The increasing similarity in export structure accession may present a challenge to the Thai econ- between China and Thailand suggests, however, a omy as competition in the textile and apparel sector potential threat from competition in third-country strengthens. markets. A market-by-market and a product-by- The intensity of trade between Thailand and product analysis carried out for Thailand suggest China (Ng and Yeats 2003) in 2001 suggests that the that Thai exports are less at risk in the United States planned reductions in protection of China's mar- than in Japan.37 In the U.S. market, the specific kets offer some good opportunities for exporters. products that appear most at risk from Chinese For Thailand, much of the tariff reduction will competition do not include Thailand's most occur by 2004 (Table 2.1); it is receiving an average important export products to the United States.38 tariff reduction of more than six percentage points. Thailand's exports to the United States are heavily These rates represent the statutory tariff bindings concentrated. Based on data for 1995­99, the top under China's WTO accession, but China is free to five products account for more than 25 percent of offer tariff rates below these bindings as agreed, for total Thai exports, and the top 100 products example, in 2003 under the "Early Harvest" aspect account for 81 percent. Less than 9 percent of Thai- of the ASEAN-China Framework Agreement. land's U.S. exports fall into the at-risk categories-- Thailand's agricultural exports to China stand that is, where unit values are close to those of Chi- to benefit significantly from the liberalization of nese products and where China's import market nontariff barriers.33 More than 30 percent of Thai share exceeds 10 percent (Table 2.4).39 For products exports to China are in products--especially rice, in which China supplies 5­10 percent of U.S. rubber, and cane sugar--on which quantitative imports, roughly 15 percent of Thai exports are at restrictions will be lifted (Table 2.2).34 And, under risk. the WTO accession agreement, China confirms that Thailand's exports to Japan are more vulnerable tariff-rate quotas will be allocated with historical than those to the United States. They are less con- trade flows to end users in mind, and imports will centrated than those to the United States, with the be allocated to the full limit of the quota established top five export products40 accounting for only 21 for each calendar year based on demand in the Chi- percent of total Thai exports and the top 100 for 76 nese market. Because Thailand's import share in percent. But nearly 21 percent of Thailand's exports many of the products subject to tariff-rate quotas are in the at-risk category--where the unit values has been high historically, it can expect to be allo- of Thai products are close to those of their Chinese cated high quotas.35 Thailand's food processing competitors and where China's share of imports is industries also are well placed to expand in China's greater than 10 percent.41 Four of Thailand's top 10 markets. exports to Japan appear to be at risk on this basis.42 Thailand's overall export volume will continue Based on the 5 percent threshold, nearly 25 percent to grow. Demand from China will be compounded of Thailand's exports to Japan are at risk. by increased import demand from China's closest As for investment flows, the opportunities may trading partners--Japan, the NIEs, the EU, and the outweigh the risks. Sectors that are now contract- United States--which themselves have benefited ing in Thailand--apparel and textiles as well as from China's growth and WTO accession. However, assembly operations--have historically accounted competitive pressures are likely to shift Thai manu- for important shares of inward FDI. One-fourth of facturing increasingly into electronics and other FDI has gone to the hotel and restaurant sectors, manufactures, especially metals and petrochemical which have room for expansion. At the same, Thai- products. Exports from processing and electronics land is poised to expand its role in global networks Regional Impact of China's Accession to the WTO 33 producing electronics, metals, petrochemicals, and Impact of WTO Accession on the other manufactures. In the electronics industry, Low-Income Countries of East Asia when FDI flows to China gained momentum in the late 1990s, FDI flows to Thailand in these sub- The impacts of China's accession to the WTO will sectors continued growing, reaching 27 percent of vary widely among Cambodia, the Lao People's the FDI in Thai manufacturing in 1997­2001. In Democratic Republic (Lao PDR), and Vietnam. 2001 imports from China of parts and compo- Cambodia is particularly vulnerable because of its nents already represented 13 percent (or more heavy emphasis on apparel exports. Lao PDR, by than $1 billion) of Thailand's total parts and com- contrast, is likely to be little affected in either China ponents imports, and exports to China of parts or third-country markets. Vietnam falls somewhere and components totaled nearly $1 billion (Ng and in between, with competitive pressures on its Yeats 2003). exports matched by growing market opportunities. In Thailand as in the Philippines, many of the Regional arrangements are playing an impor- sectors with exports at risk use unskilled labor tant role in ensuring that the low-income countries intensively, while the expanding sectors use land or of East Asia benefit from regional trade. For exam- skilled labor more intensively. This finding suggests ple, under the "Early Harvest" aspect of the that growth and change in China will have the effect ASEAN-China Framework Agreement announced of lowering the wages of unskilled Thai workers rel- in November 2002, China granted these countries ative to those of skilled workers. The potential most-favored-nation (MFN) status, even though impact on urban inequality and vulnerability in the they have not yet acceded to the WTO. short term will have to be carefully monitored and addressed. As unemployment in the labor-intensive Cambodia sectors rises, the pressure to facilitate the adjustment process in the labor market will increase. For Cambodia, it will be important to capitalize on The extent to which Thailand will exit from any opportunities from China's growth. Cambodia assembly-type production or upgrade its capabili- will see some benefits from China's import tariff ties will depend largely on the policies it pursues-- reductions,43 but, most important, some of Cam- either embracing trade-induced competitiveness bodia's top exports--especially rubber and wood-- and productivity gains or submitting to short-term will benefit from the reduction in quantitative protectionist pressures. Manufacturers are already restrictions (Table 2.5). The "Early Harvest" provi- complaining about low-cost imports of electrical sion of the ASEAN-China Framework Agreement appliances and motorbikes from China. Also also gives Cambodia MFN treatment. Growth in important are supply factors, including the local China overall is expected to be a powerful source of availability of engineering and sourcing capabili- external demand for Cambodia, including for ties, and government incentives for upgrading tech- cross-border activity that may not be well captured nology. Thailand's private business community is in official statistics. already exploring niche opportunities within labor- China's accession to the WTO could pose a dou- intensive sectors. ble threat to Cambodia: to its balance of payments, TABLE 2.5 Share of Exports to China Affected by Lifting of Quantitative Restrictions (QRs): Cambodia and Lao PDR (percent of total exports,1995­99) Percent of exports Country facing QRs Key products (% of total exports) Cambodia 11 Rubber (11); timber (2.9); plywood (4.64) Lao PDR 0.6 Rice (0.6); timber (62.4) Sources: UN Comtrade data; China's WTO accession agreement; authors' calculations. 34 East Asia Integrates from lost exports, and to many of its households, Singapore is another important market. Cambo- from lost jobs and lower wage incomes. Cambodia's dia's exports to Singapore are extremely concen- exports are already highly correlated with China's, trated, with natural rubber accounting for nearly and many of Cambodia's leading exports are in the 60 percent, and their exposure to Chinese exports is apparel and textile categories where China is cur- quite limited. Even at the 10 percent import share rently quota-constrained. threshold, less than 3 percent of Cambodian The United States makes up 40 percent of Cam- exports appear to be at risk. bodia's export market. Cambodia's exports to the All this said, Cambodia faces major challenges in United States have been very concentrated, with 94 adjusting its economy. Reforms in governance and percent of them in textiles and apparel. Five prod- other areas need to be speeded up to ensure that ucts--all apparel--accounted for nearly half of total alternative production networks develop. exports in 1995­99. It is true that a high proportion (68 percent) of Cambodian exports to the United Lao People's Democratic Republic States have unit values that are not close to those of their Chinese competitor products, and thus the Lao PDR is less exposed than Cambodia to the share of exports that are at risk at the 10 percent impact of China's accession to the WTO. Although share threshold appears low (9 percent) (Table 2.6). the effects of tariff reduction will be quite small,44 But this estimate significantly understates Cambo- the phase-out of quantitative restrictions associated dia's likely true exposure. Once the quota restric- with WTO accession will create market opportuni- tions on China are lifted, its exports to the EU and ties for Lao PDR. The main benefit of the QR liber- North America are likely to capture more than 10 alization will be felt in timber exports,45 which percent of these markets. If Cambodia's exports in make up 62 percent of total exports to China, but these categories are viewed as being in a category will also affect other key exports such as dried with a high Chinese import share, the picture fruits, "other pharmaceutical plants," and coffee. changes dramatically: Roughly 30 percent of Cam- No appreciable losses are expected in Lao PDR's bodia's exports show themselves to be at risk. most important export market, Thailand, which Cambodian manufacturers continue to hope absorbs one-fourth of its total exports.46 Ten per- they can compete in various niche markets. But cent of Laotian exports go to Japan, and 45 percent they face some significant administrative hurdles of these are in a single product category--builders and government interventions that tend to negate and joinery.47 Only 4 percent of Lao PDR's exports Cambodia's comparative advantage against its to Japan are in at-risk categories, where unit values regional competitors, particularly in the context of are close to those of competitor products from low labor input costs. China and where China's import share is more than TABLE 2.6 Market-by-Market and Product-by-Product Analysis of Cambodian Exports: China Market Share and Closeness of Unit Value (UV) as Proxies for Potential Risk China's import share China's import share > 10% + all < 10% > 10% < 10% apparel, textiles Share in U.S. market (%) (n = 64 products) UV close 22.0 9.5 2.4 29.0 UV not close 39.3 28.7 1.0 67.0 Share of market in analysis 99.4 99.4 Note: The figure shows calculations using two (arbitrary) threshold levels for Chinese imports: 5 percent and 10 percent of the total import market. Sources: UN Comtrade data; China's WTO accession agreement; authors' calculations. Regional Impact of China's Accession to the WTO 35 10 percent (see Table 2.7). (Using the 5 percent tors likely to expand include food processing, light import share threshold, the exposure increases to manufacturing, metals and petrochemicals, elec- 19 percent, but sensitivity analysis shows only an tronics, and other manufactures (Figure 2.5). insignificant impact on the terms of trade.) To date, Vietnam has been particularly successful Although trade with China clearly represents at expanding its apparel and other labor-intensive opportunities for Lao PDR, the likely impact of exports--apparel is an important growth sector for China's WTO accession on the country may be lim- Vietnam. By global standards, Vietnam's exports of ited. Its policymakers therefore have some breath- garments are small,49 but at the national level they ing space to implement their broader development are important. They account for more than 14 per- agenda. cent of Vietnamese exports and provide a livelihood for thousands of unskilled workers, many of whom are female. The country has exported apparel Vietnam mainly to the EU and Japan. More recently, the Viet- China's WTO accession will present Vietnam with a nam-U.S. Bilateral Trade Agreement has opened up combination of expanding market opportunities in the U.S. market to apparel from Vietnam,50 but China's growing market and serious competitive China's WTO accession calls into question how threats. The ability of the economy and particularly much of a share Vietnamese exporters will be able to the private sector to react flexibly will be critical. capture in this market. Vietnam has two years to Vietnam itself is preparing for integration with the establish itself before the 2005 removal of the quota rest of the world. The country has made a bold step on Chinese imports. Estimates suggest that China's toward trade reform by signing the Vietnam-U.S. entry to the WTO will depress Vietnam's apparel Bilateral Trade Agreement. It is also in the process output and exports relative to the baseline by 2010 of implementing the ASEAN Free Trade Area and negatively affect its terms of trade (Figure (AFTA) and getting ready for accession to the 2.5).51 Therefore, it is important for Vietnam to use WTO. Despite these reforms, however, its economy the period 2002­05 to advance the reform process is still relatively closed (Ianchovichina 2001), and it and smooth the structural adjustment that will need is still in the process of establishing its export base. to take place. On the positive side, increased access to China's For Vietnam, trade liberalization by China domestic market will stimulate Vietnamese exports implies lower costs for Chinese imports, which rep- across most product categories and further inten- resents an important opportunity to reduce the sify trade between the two economies.48 China's costs of Vietnamese manufacturing overall. At the accession will create opportunities for expanding same time, liberalization may represent tougher rice production and increasing exports to China of competition for its own manufacturers. Vietnam oilseeds, sugar, and cotton, for example. Other sec- has already imposed temporary safeguard measures TABLE 2.7 Market-by-Market and Product-by-Product Analysis of Lao PDR's Exports: China Market Share and Closeness of Unit Value (UV) as Proxies for Potential Risk China's import share China's import share < 10% > 10% < 5% > 5% Share in Japanese market (%) (n = 11 products) UV close 67.3 4.0 52.2 19.1 UVnot close 24.3 4.3 24.3 4.3 Share of market in analysis 100.0 100.0 Note: The figure shows calculations using two (arbitrary) threshold levels for Chinese imports: 5 percent and 10 percent of the total import market. Sources: UN Comtrade data; China's WTO accession agreement; authors' calculations. 36 East Asia Integrates on imports of motorbikes. Such measures can only Dynamic gains will swamp the static impact prolong the adjustments that Vietnam needs to from China's WTO accession. The liberalization of make to realize its regional comparative advantages. China's services sector will lead to expanded Vietnam will have to greatly improve its busi- exports of services such as transport logistics, com- ness environment if it is to compete successfully for munication, business, and tourism services, and the foreign investment with China and reap net gains productivity from expanded competition and for- from China's growing role in the region. Without eign entry in China's services sector will fuel such action, the net impact of China's WTO acces- China's economy and demand. Because trade sion could be negative for Vietnam. It will be intensity with China is high for all emerging East important to resist protectionist pressures from Asian economies, they stand to benefit from this domestic producers and to act aggressively to estab- dynamic growth. Growth in the region's exports lish its garment exports in the United States. will also be fueled by the increased demand from China's major trading partners that benefit directly from China's accession to the WTO. Conclusions China's accession and deeper integration into The biggest beneficiary of China's accession to the the world economy present the East Asia region World Trade Organization is China itself, and most with a range of opportunities and challenges. Given of the benefits are associated with China's own the complexity of the changes, this assessment of trade liberalization. China's accession and growing the impact on the region's economies is merely sug- role in the world economy also have important gestive. However, it provides a framework for fur- implications for the rest of East Asia. ther understanding the likely changes, so that East Emerging East Asia is in a good position to ben- Asian economies can position themselves to grasp efit from China's agricultural liberalization. China the opportunities and manage the challenges. already is a major agricultural market for these countries, and the markets of most interest to Endnotes developing East Asia--oilseeds, sugar, processed foods--are poised to open further. Although the 1. It is difficult to judge the accuracy of the FDI data because they reflect "round-tripping" investments undertaken from initial protection in China's food grain sector might China to take advantage of concessions enjoyed only by be less than other studies have suggested (e.g., for foreign investors. The outflows do not include flows rice), import demand is still expected to increase through Hong Kong (China). considerably. 2. Descriptions of the CGE methodology and detailed results are given in the background paper by Ianchovichina and The potential for specialization and comple- Walmsley (2003) at www.worldbank.org/eaptrade. mentary intraindustry trade in the manufacturing 3. For information on individual sectors, see Yusuf, Altar, and sector could be significant. China is increasingly a Nabeshima (2003). 4. These are weighted average tariffs computed using trade central player in global production networks, weights for 2001; see Ianchovichina and Martin (2002). See including electronics and machinery. And its trade www.worldbank.org/eaptrade for the estimated evolution with the rest of East Asia in parts and compo- of tariff rates by product group, 1997­2010. 5. Dollar amounts are current U.S. dollars. nents--both imports and exports--represents an 6. China will be subject to additional textile safeguard quotas opportunity for the rest of emerging East Asia to until 2007, but these will be applicable for only one year at benefit from China's growing role. In the automo- a time, unlike the existing quotas that were put in place for bile sector, however, China's current plans for an indefinite period. 7. This is a consensus finding supported by Deutsche Bank restructuring its industry will make it a more effi- (2001), Ianchovichina and Martin (2003), Ianchovichina cient assembler, leading to a contraction of produc- and Walmsley (2003), and Wang (2002). The textile indus- tion in Japan and the NIEs. Likewise, the abolish- try in general will also be hurt, though not nearly as much as the apparel industry, because some of these countries ment of import quotas on Chinese textiles and will start exporting textiles to China and other NIEs. apparel in key markets by 2005 will make China a 8. McKibbin and Tang (2000) and Ianchovichina and Walms- formidable competitor, especially in the apparel ley (2003) discuss in detail the effect of trade liberalization on rates of return to capital and foreign investment. sector, and will lead to restructuring, with a partic- 9. All East Asian economies saw increases in their share of ularly significant impact on quota-dependent low- manufactured exports during the first half of the 1990s, and and middle-income countries. all saw changes in the structure of manufactured exports. In Regional Impact of China's Accession to the WTO 37 the 1990s the NIEs increased their share of electronics and able indicates the importance of exports in that product cate- information technology products and China increased its gory to the sending country. The second variable was imports share of electronics and telecommunications exports. from China as a share of total imports in that product category 10. Even though wage levels and productivity lag in China's (e.g., garment imports from China as a share of total garment hinterland, in the short to medium term the lagging imports in the United States). In the country-by country regions may not be able to compete successfully with other analyses, we looked at two (arbitrary) threshold levels for the less developed East Asian economies in attracting footloose share of Chinese imports--10 percent and 5 percent of total export-oriented industries because of high transport and imports. The third variable was unit values of imports in the other infrastructure costs. product categories in question. Product categories mask a 11. These results from CGE modeling are consistent with find- variety of differences; unit values provide an additional indica- ings of other CGE modeling work. See Li and others tion of how similar products are and therefore how likely there (2000), Deutsche Bank (2001), Ianchovichina and Martin is to be a terms-of-trade impact from competition. If the unit (2003), and Wang (2002). values of imports from both sending countries were on the 12. See Appendix Table 2.3 in the technical appendix to this same side of the average unit values for the product category in chapter at www.worldbank.org/eaptrade. question, we deemed them to be"close." 13. Differences in tariff cuts reflect differences in export com- 23. These products are technically specified natural rubber (9 position by exporting country. percent of exports); footwear with outer sole of leather (8.2 14. Ng and Yeats (2003) estimate a distance-adjusted trade percent); video recording or reproducing apparatus (5.4 intensity index of 1.36 for China and Indonesia. percent); plywood with outer ply of tropical wood (5 per- 15. This group of products includes roasted decaffeinated cof- cent); and other rubber footwear (3.1 percent). fee, palm kernel or babassu oil, cocoa powder, soap, cath- 24. Footwear with leather soles and video recording or reproduc- ode ray tubes, and cane and bamboo furniture. ing apparatus--which accounted for 8 percent and 5 percent 16. Between 1995 and 1999 Indonesia was the second largest of Indonesian exports to the United States, respectively. exporter of palm oil to China after Malaysia. In 1999 it sup- 25. Plywood with an outer ply of tropical wood (19.4 percent); plied China with 355,172 metric tons. frozen shrimp and prawns (11.5 percent); bituminous coal 17. The initial quota for palm oil is 2.1 million metric tons, (7.1 percent); nickel mattes (3.5 percent); and nonalloyed which will rise to 3.168 million metric tons in 2005. In 1999 aluminum (2.9 percent). Chinese imports of palm oil totaled 1.193 million metric 26. They are bituminous coal and cotton yarn other than tons. The in-quota tariff rate is set at 9 percent throughout sewing thread. the period, and while the initial out-of-tariff rate is as high 27. In 2001 Indonesia exported $313 million to China and as 63.3 percent, it is slated to decline constantly to 9 percent imported $110 million from China in parts and compo- in 2006. Although TRQs will be allocated to both state and nents (Ng and Yeats 2003). nonstate trading enterprises, the reduction in the propor- 28. Although Malaysia faces some threats because of its high tion of state trading enterprises from 42 percent to 10 per- wages and heavy reliance on mid-range electronics, the cent during the implementation period and the elimina- economy-wide, medium-term analysis highlights equally tion of the tariff-rate quota on January 1, 2006, should important opportunities for gains. facilitate development of the market. 29. See Appendix Table 2.2 in the technical appendix to this 18. For natural rubber imports, the total initial quota level is set chapter at www.worldbank.org/eaptrade. at 429,000 metric tons, with an annual quota growth rate of 30. See Appendix Table 2.2 in the technical appendix to this 15 percent after 2004, when the quota will be eliminated. chapter at www.worldbank.org/eaptrade. Phasing-out dates are as of January 1 of the calendar year 31. Ng and Yeats (2003) estimate a distance-adjusted trade specified. Natural rubber is a product subject to designated intensity index of 2.29 for China and the Philippines. trading (Annex 2B of China's WTO accession agreement), 32. See Appendix Table 2.2 in the technical appendix to this which is slated to be liberalized within three years of acces- chapter at www.worldbank.org/eaptrade. sion. Under designated trading, the Chinese government 33. The annual average value of Thailand's exports to China is authorizes only certain firms to engage in international $2.14 billion. Thailand's top exports to China include parts trade. Initial and final tariff rates are bound at 20 percent. for automated data processing machines, rice, rubber, cane 19. Urea accounted for 1.3 percent of Chinese imports from sugar, and plastics. Indonesia. Previously, urea was subject to both an import 34. According to China's WTO Protocol of Accession, tariff- license and quota. These restrictions will be replaced by tar- rate quotas will apply to grains, sugar, and cotton for which iff-rate quotas upon accession. The initial quota for urea is out-of-quota tariffs are quite high. Otherwise, after the about 1.3 million metric tons, increasing to 3.3 million met- phase-in period the tariffs will range between just 1 and 15 ric tons in 2006. The out-of-quota tariff rate is 50 percent. percent--representing substantial liberalization over 2001 20. Indonesia could increase its oil sales. However, because levels (Anderson, Huang, and Ianchovichina 2002). Indonesia is a member of the Organization of Petroleum 35. In 1995­99 Thailand supplied more than 84 percent of Exporting Countries (OPEC), its production is constrained China's imports of rice and nearly 60 percent of its imports by OPEC quotas. of natural rubber. Thailand supplied almost 33 percent of 21. During 1995­99, based on 2,700 products at the five-digit China's imports of raw cane sugar, but it is only one of 12 Standard International Trade Classification (SITC) level, countries that have initial negotiating rights. the correlation between Indonesia's export structure and 36 See Appendix Table 2 in the technical appendix to this China's rose from 0.174 to 0.302. chapter at www.worldbank.org/eaptrade. 22. In each major import market we looked at three variables. The 37. See the description of the methodology in the earlier section first was imports in a particular product category as a share of on Indonesia for which a similar analysis was carried out. total imports from that country (e.g., garment imports from 38. The top export products include digital monolithic units Thailand as a share of total imports from Thailand). This vari- (7.7 percent); frozen shrimp and prawns (7.5 percent); pre- 38 East Asia Integrates cious metal jewelry (4.1 percent); prepared and preserved Conference on Global Economic Analysis, Monash Univer- crustacea (3.6 percent); and leather sole footwear (2.4 per- sity, Melbourne, Australia, June 28­30. cent). Francois, Joseph F., and Den Spinanger. 2002. "Market Access in 39. These shares denote minimum levels of exposure, because Textiles and Clothing." Paper prepared for Informing the the analysis covers only the top 150 and 165 products. Doha Process: New Trade Research for Developing Coun- 40. They are smoked sheets of natural rubber (6.4 percent); tries, Cairo, May. frozen shrimp and prawns (5.7 percent); frozen poultry Hertel, T. W., T. L. Walmsley, and K. Itakura. 2001. "Dynamic parts (3.7 percent); preserved and prepared crustacea (2.8 Effects of the `New Age' Free Trade Agreement between percent); and parts for automated data processing Japan and Singapore." Journal of Economic Integration 16 (4): machines (2.5 percent). 446­48. 41. Among the top 200 export products, unit value data are Huang, J., and S. Rozelle. 2002. "The Nature of Distortions to missing for 11. We focus on the remaining 189 products, Agricultural Incentives in China and Implications of WTO which collectively accounted for 82.8 percent of Thailand's Accession." Paper presented at Seminar on WTO Accession, exports to Japan. Policy Reform and Poverty Reduction in China, World Bank, 42. The four exports are frozen poultry parts; prepared and Beijing, June 28­29. preserved crustacea; other frozen, dry, salted mollusks; and Ianchovichina, Elena. 2001. "Trade Barriers in Vietnam and seats with wooden frames. China: A Comparison." World Bank, Washington, D.C. 43. The weighted average tariff for the top 100 exports to Processed. China will decline from a relatively low level of 4.3 percent ------. 2003. "GTAP-OD: A Model for Analyzing Trade in 2001 to 2.6 percent in 2005. Reforms in the Presence of Duty Drawbacks." GTAP Techni- 44. The weighted average tariff for the top 100 exports to China cal Paper 21, Purdue University. will decline from 7.3 percent in 2001 to 6.0 percent in 2005. Ianchovichina, Elena, and William J. Martin. 2001."Trade Liber- 45. Timber trading will be liberalized within three years of alization in China's Accession to the World Trade Organiza- accession. tion." Journal of Economic Integrations 16 (4): 421­45. 46. Nearly 62 percent of exports to Thailand are of wood and ------. 2003. "Economic Impacts of China's Accession to the related products--categories in which a large Chinese WTO." Policy Research Working Paper 3053, World Bank, export supply response is not expected. Washington, D.C. 47. Lao PDR exports 11 products to Japan--largely wood and Ianchovichina, Elena, and Terrie Walmsley. 2003."The Impact of wood products (74 percent), coffee (24 percent), and some China's WTO Accession on East Asia."Policy Research Work- cotton shirts and footwear (2 percent). ing Paper 3109, World Bank, Washington, D.C. Processed. 48. Trade intensity indexes in Ng and Yeats (2003) show that IMF (International Monetary Fund). Various years. Direction of the intensity of trade between China and Vietnam grew Trade Statistics. Washington, D.C.: IMF. markedly between 1985 and 2001. Li, S., Z. Wang, Z. Fan, and L. Xu. 2000. WTO: China and the 49. In 1997 Vietnam had less than a 1 percent share of the World. Beijing: China Development Press. global apparel market. Martin, William J. 1993. "The Fallacy of Composition and 50. Vietnamese apparel exports to the United States have grown Developing Country Exports of Manufactures." World Econ- rapidly over the past couple of years. In 2001 articles of omy 16 (2): 159­72. apparel and clothing accessories exported to the United States Marukawa, T. 2001. "Japanese Foreign Direct Investment and amounted to $57 million. In 2002 exports of these items China's Industrial Development: Focusing on Automobile, exploded to nearly $900 million, and to more than $200 mil- Electronics, and Textile Industries." Paper presented at Con- lion in January 2003 alone (http://dataweb.usitc.gov/). In ference on Japan and China: Economic Relations in Transi- response, the U.S. government was prompted to imposed tion, Tokyo, January 18­19. quotas on Vietnamese garment exports. Mattoo, Aaditya. 2002. "China's Accession to the WTO: The Ser- 51. See Fan and Li (2000). Also see Appendix Table 2.3 in the vices Dimension." Paper presented at Seminar on WTO technical appendix to this chapter at www.world Accession, Policy Reform and Poverty Reduction in China, bank.org/eaptrade. World Bank, Beijing, June 28­29. McKibbin, W., and K. Tang. 2000. "Trade and Financial Reform in China: Impacts on the World Economy."World Economy 23 (8). References Ng, F., and A. Yeats. 2003. "Major Trade Trends in East Asia: What Are Their Implications for Regional Cooperation and The word processed describes informally reproduced works that Growth?" Policy Research Working Paper 3084. World Bank, may not be commonly available through libraries. Washington, D.C. Spihauger, Den. 1999. "Textiles beyond the MFA Phase-Out." Anderson, K., J. Huang, and E. Ianchovichina. 2003. "Impact of World Economy 22 (4): 455­76. China's WTO Accession on Farm­Nonfarm Income UNCTAD (United Nations Conference on Trade and Develop- Inequality and Rural Poverty." Policy Research Working ment). 2001. World Investment Report 2001: Promoting Link- Paper 3052, World Bank, Washington, D.C. ages. Geneva. Claro, S. 2001. "Tariff and FDI Liberalization: What to Expect Wang, Z. 2002. "WTO Accession, `Greater China' Free Trade from China's Entry into WTO?" Paper presented at Eighth Area, and Economic Relations across the Taiwan Strait." Annual Conference on Empirical Investigations in Interna- Paper presented at Fifth Conference on Global Economic tional Trade, Purdue University, November 9­11. Analysis, Taipei, June 5­7. Deutsche Bank. 2001. Quantifying the Impact of China's WTO Yusuf, Shahid, M. Anjum Altar, and Kaoru Nabeshima, eds. Entry. DB Global Market Research, December 14. 2003. Global Production Networking and Technological Fan, Z., and S. Li. 2000. "The Implications of Accession to WTO Change in East Asia. Washington, D.C.: World Bank and for China's Economy." Paper presented at Third Annual Oxford University Press. 3 New Regionalism: Options for East Asia Mari Pangestu Sudarshan Gooptu Half of world trade is now conducted under prefer- and arrangements aimed at broader cooperation-- ential trade arrangements (WTO 2000a, 2000b), up for example, ASEAN + 3 (China, Japan, and Korea) from 40 percent in 1988­92 (Lloyd and Crosby, and ASEAN + CER (Australia, New Zealand). New 2002). In East Asia, the past several years have seen early warning systems for macroeconomic imbal- a plethora of proposals for new preferential ances, as well as numerous technical assistance and arrangements, both bilateral and regional. information exchange arrangements, are also being Although very few of these proposals have reached considered. the negotiation stage or have been formalized, sev- Preferential trade arrangements are sharply at eral economies in the region are seriously engaged odds with previous East Asian preferences for a in developing new preferential trade relationships. nondiscriminatory approach to reform.3 In the These include China and two countries, Japan and mid-1990s economies in the region pursued a uni- the Republic of Korea, that were formerly staunchly lateral and nonpreferential route to trade liberal- against the preferential route to trade liberalization. ization--an approach that is embodied in the The most significant of the new initiatives in principles of the Asia-Pacific Economic Coopera- terms of members is the ASEAN (Association of tion (APEC) forum and supported by evidence Southeast Asian Nations)-China Free Trade Area that such a commitment contributes to growth, now under negotiation.1 If agreement is reached, it development, and greater integration in economic will be the world's biggest free trade area, encom- terms and more broadly.4 The East Asian passing 1.7 billion people, a collective gross domes- economies, especially, undertook unilateral liberal- tic product (GDP) of almost US$2 trillion,2 and ization mainly on a most-favored-nation basis, intraregional trade of $1.2 trillion. Other recent encouraged by regional peer pressure and World proposals include bilateral cooperation between Trade Organization (WTO) commitments, among Japan and Korea and between Japan and Singapore other factors. This chapter draws on background papers prepared for this study by Peter Lloyd and Mark Crosby (2002) and He Fan (2002), as well as other papers listed in the references. We have also benefited from the excellent research assistance that was provided by Qing Lani Wu from the Kennedy School of Government, Harvard University, during her internship at the World Bank. 39 40 East Asia Integrates Regional integration in trade should be viewed of 1997­98. The crisis showed that rapid deprecia- as only one element (but a critical one) of a broader tion of one country's currency could adversely set of policy and institutional changes currently affect the export competitiveness of other coun- taking place in East Asia and aimed at regional eco- tries, especially neighbors producing the same nomic cooperation. The wider the country and sec- products for the same export markets (Eichengreen toral coverage of regional trade agreements (RTAs), 2001). The crisis initially propelled countries to the greater will be the gains from removing intrare- explore options for monetary cooperation and gional barriers through them. Thus liberalization macroeconomic policy coordination,5 but, by high- goals can be well served by moving on several lighting the economic interdependence of the fronts. The outcome will depend crucially on main- region, it has also given rise to proposals for taining a common set of principles for each regional cooperation in trade and investment (see regional integration agreement that focus on trade Box 3.1.). facilitation measures, allowing the most liberal A second key factor in the new trend is the per- rules of origin, and yet at the same time act as a cat- ceived need by other economies of the region for alyst for multilateral trade liberalization. stronger cooperation with China, both as a growing import market and as a rising competitor in export markets. Over the decade 1990­2000, China experi- Reasons for New Regionalism enced an average real rate of growth of 10 percent a The new regionalism in East Asia seems to be moti- year, and its exports quadrupled from some $62 bil- vated by several factors. The first is the need to lion to $250 billion. Meanwhile, investment flows reduce the risks of financial contagion and unusual to China increased dramatically, from some $3 bil- exchange rate instability, the damaging effects of lion to more than $40 billion (ASEAN-China which were made clear by the Asian financial crisis Expert Group on Economic Cooperation 2001). BOX 3.1 Forms of Regional Preferential Trading Arrangements Statistics on regional trading arrangements noti- · Common market: free trade area among fied to the WTO show that more than twice as members in goods, services, capital, and many preferential trading arrangements (197) labor. Entails removal in those areas of all bar- are in force today as 10 years ago (WTO 2002b). riers and trade regulations that restrict trade Nearly all WTO Members (97 percent) now par- among members. ticipate in RTAs, and many belong to more than · Single market: a common market plus harmo- one. Some, like member countries of the Euro- nization of all laws, regulations, and taxes pean Union (EU) and Mexico, belong to more that affect market prices. than 10 RTAs (Crawford and Laird 2000). EU- · Economic union: common market and unifica- centered or -related RTAs make up about half of tion of economic institutions and coordina- the RTAs in force. RTAs come in many forms: tion of economic policy among member · Sectoral preferential trading arrangements: countries. Supranational institutions are preferential tariffs in selected sectors and established with decisions binding all mem- goods. Examples are the European Coal and bers. Involves considerable surrender of Steel Community and the ASEAN Preferential national sovereignty. Trade Agreement scheme, which preceded · Monetary union: a single currency and a sin- the AFTA scheme. gle central bank. · Free trade area: removal of tariffs and other · Fiscal union: equal tax treatment in one coun- trade regulations that restrict trade among the try of enterprises and persons from anywhere members (may apply to goods and services). within the RTA region. · Customs union: a free trade area plus harmo- nization of tariff rates and trade regulations Sources: Lloyd and Crosby (2002) and Chia that face third countries (may apply to goods (2002). and services). New Regionalism: Options for East Asia 41 A third factor is the interest of business commu- breakthroughs in the multilateral trade liberaliza- nities in getting preferential access to foreign mar- tion presently being considered at the global level. kets, especially when these are imperfectly compet- There is concern that pursuit of a regional itive markets in which some form of establishment approach might deflect attention from the multilat- is required. The benefits from being the first eral agenda, that it could hold East Asia back from movers in such an environment are significant. The reaching its full economic potential, and that it greater tradability of many services and the growth could increase economic and political tensions in of foreign direct investment (FDI) have con- the region. Experience shows that a haphazard and tributed to this focus in policymaking. uncoordinated proliferation of preferential agree- Other factors include the move by many ments can increase the costs of doing business. economies, especially the more developed in the With an eye toward helping the economies of the region, to lower their average tariffs; the growing region make informed decisions on trade policy and recognition of the value of harmonizing standards economic integration, this chapter reviews current and regulations, so that they do not impede trade; and proposed regional trading arrangements and and the higher concentration of trade among discusses issues in their design. The first major sec- regional partners, especially in East Asia. These tion that follows outlines the agreements now in changes have affected countries' assessment of the force and under discussion in East Asia. The next costs and benefits of entering into preferential section examines what the literature reveals about agreements. Some countries are also seeking to the benefits and risks of such arrangements and forge new agreements as a defensive response to draws out some policy implications for East Asian arrangements being created elsewhere. economies. It includes a look at whether a regional Agreements on economic cooperation offer approach can complement a multilateral approach. opportunities to build a sense of community or to The evolving relationship between China and repair past tensions between neighboring ASEAN and the implications for their cooperation economies. Membership in regional trading agreement now under negotiation are examined in arrangements and informal economic cooperation the next section. It is followed by a discussion of forums provides occasions for numerous meetings regional financial cooperation and our conclusions. between senior officials, ministers, and leaders, and apparently it has served as a crucial component of Regional Trade Agreements in the community building that has taken place in the East Asia region, especially in ASEAN. APEC meetings have also provided an important forum in which to dis- Bilateral and multilateral trading arrangements in cuss economic and, increasingly, noneconomic East Asia are proliferating and becoming more issues of concern to members without having to complex. Chia (2002) distinguishes between eco- negotiate. The discussions and negotiations that do nomic regionalization and regionalism. Regionaliza- take place, including issues not yet covered in the tion refers to the greater economic interdependence WTO, are also seen as important capacity-building that results from increased intraregional trade, exercises. investment, technology, and migration flows with- Finally, countries cite their perceptions of a slow out any formal framework of cooperation. It is pace of progress in trade liberalization, or their often termed market-driven integration because it expectations of poor prospects under WTO and occurs as each economy undertakes its own unilat- existing regional cooperation mechanisms such as eral process of opening up, as well as fulfilling its ASEAN and APEC and the example set by multilateral commitments. Regionalism refers to increased economic integration in Europe and formal economic cooperation and economic inte- North America. gration arrangements, and agreements between two or more countries that are designed to achieve economic growth through trade and investment Organization of This Chapter liberalization and facilitation. The chances for a new WTO trade round are now Market-driven integration or regionalization has brighter, and East Asia stands to gain from any been occurring in East Asia since the mid-1980s 42 East Asia Integrates through increased trade and investment linkages. Some may never be implemented because many The process has been driven by unilateral reforms in unresolved issues remain, and others may only individual economies and by the logic of the "flying become implementation agreements associated geese" pattern of relocating production processes to with APEC processes. cheaper areas abroad as domestic costs rise. Firms The arrangements mooted or under discussion moved their production processes from Japan to the are all free trade areas, not customs unions. Like East Asian newly industrializing economies (NIEs) regional trading arrangements elsewhere in the in the 1970s and early 1980s, to Southeast Asia in the world, they extend beyond the traditional areas of mid-1980s to early 1990s, and to China in the mid- trade policy--such as tariffs and nontariff mea- 1980s to mid-1990s (Xu and Song 2000).6 To sup- sures--and into investment, services, and stan- port the vertical specialization and division of pro- dards. These are areas covered by the WTO, but duction along the value chain, intraregional trade where the progress of WTO negotiations might be has grown rapidly in components, parts, and inputs. seen as too slow (e.g., in services), or where WTO Table 3.1 lists actual and potential trading coverage is very limited (e.g., the investment issues arrangements involving East Asian economies.7 in goods production). All the discussions refer to Most of these arrangements are still in the form of WTO consistency as well as to other rules such as proposals and studies, or at the negotiation stage. open access and comprehensive coverage. TABLE 3.1 Proposed and Actual Regional Trading Arrangements Involving East Asian Countries Type of agreement Status Year Bilateral Asia Pacific China­Hong Kong (China) Closer economic partnership Signed 2003 Singapore­Australia Free trade area Signed 2003 Singapore­Canada Free trade area Under negotiation 2001 Singapore­Chile Free trade area Under negotiation 2000 Singapore­Japan Free trade area Signed 2002 Singapore­ Korea, Rep. of Free trade area Proposal Singapore­Mexico Free trade area Under negotiation 1999 Singapore­New Zealand Closer economic partnership Signed 2001 Singapore­Taiwan (China) Free trade area Proposal/study 2002 Singapore­USA Free trade area Signed 2003 Korea, Rep. of­Australia Free trade area Official discussions 2000 Korea, Rep. of­Chile Free trade area Signed 2002 Korea, Rep. of­China Free trade area Proposal/study Korea, Rep. of­Japan Free trade area Official discussions/ study 1998 Korea, Rep. of­Mexico Free trade area Official discussions/ study 2000 Korea, Rep. of­New Zealand Free trade area Official discussions/ study 2000 Korea, Rep. of­Thailand Free trade area Proposal/study 2001 Korea, Rep. of­USA Free trade area Under negotiation 2001 Japan­Canada Free trade area Proposal/study 2002 Japan­Chile Free trade area Official discussions/ study 2000 Japan­China­Rep. of Korea Free trade area Proposal 2002 Japan­Mexico Free trade area Official discussions/ study 1998 New Regionalism: Options for East Asia 43 TABLE 3.1 (continued) Type of agreement Status Year Japan­Philippines Free trade area Proposal 2002 Japan­Taiwan (China) Free trade area Proposal Japan­Thailand Closer economic partnership Proposal/study 2002 Taiwan (China)­New Zealand Free trade area Proposal Taiwan (China)­Panama Free trade area Proposal Hong Kong (China)­New Zealand Closer economic partnership Official discussions 2001 Thailand­Australia Free trade area Under negotiation 2002 Thailand­Croatia Free trade area Proposal 2001 Thailand­Czech Republic Free trade area Proposal 2001 Thailand­India Free trade area Proposal 2002 USA­Philippines Free trade area Proposal 2002 USA­Taiwan (China) Free trade area Proposal 2002 Regional plus AFTA Free trade area Being implemented 1992 AFTA + CER Closer economic relations Official discussions/ study 2000 ASEAN + China Free trade area Official study/ negotiation 2001 ASEAN + India Regional trade and investment agreement Proposal 2002 ASEAN + Japan Closer economic partnership Official discussions 2002 ASEAN + Korea, Rep. of Free trade area Official discussions 2002 Singapore + EFTA Free trade area Signed 2002 ASEAN + 3 Free trade area Official discussions/ study 2000 EU + ASEAN Trans Regional EU­ASEAN Trade Initiative (TREATI) Proposal 2003 New regional Japan­Korea, Rep. of­China Free trade area Official discussions/ study 2000 Pacific 5 Free trade area Proposal 1997 Notes: EFTA (European Free Trade Area): Switzerland, Iceland, Liechtenstein, and Norway. Pacific 5: Singapore, Australia, New Zealand, United States, and Chile. ASEAN: Brunei, Cambodia, Indonesia, Lao PDR, Malaysia, Myanmar, the Philippines, Singapore, Thailand, Vietnam. ASEAN + 3: ASEAN plus Japan, Korea, and China. AFTA = ASEAN Free Trade Area. CER = Australia­New Zealand, already associated under the Australia-New Zealand Closer Economic Relations Trade Agreement (ANZCERTA). Sources: Authors' compilations from various media sources. Geographic proximity is not a key considera- ASEAN Free Trade Area tion. Some agreements are between close neigh- bors, but others are between trans-Pacific part- AFTA was until recently the only preferential ners, such as Mexico and Singapore or Korea and regional trade agreement in East Asia. It now has 10 Chile. Several take ASEAN/AFTA (ASEAN Free member economies with a combined population of Trade Area) as their focal point and build upon nearly half a billion and a combined GDP almost as AFTA. large as China's. AFTA became effective on January 44 East Asia Integrates 1, 2002, for the original six ASEAN members. Longer investment, and finance between China and north- timetables apply to the newer members: 2004 for east Asia, ASEAN, the economies of Hong Kong Vietnam, 2006 for Lao People's Democratic Republic (China) and Macao (China), and East Asia more (Lao PDR) and Myanmar, and 2008 for Cambodia. widely. At a recent meeting in Hong Kong, China Formal economic integration under AFTA is announced the possibility of a Hong Kong, Macao, confined mainly to tariff reductions, and intra- and China free trade area.9 And there have been ASEAN trade already enjoys low tariffs. AFTA's tar- discussions of separate bilateral agreements get is for 0­5 percent tariffs for intra-ASEAN trade between China and Asian countries such as Korea among the original six signatories. In fact, 90 per- and Thailand. The recently completed ASEAN + 3 cent of intra-ASEAN trade can already be con- (China, Japan, and Korea) Bilateral Swap Arrange- ducted at tariffs in the 0­5 percent range. The aver- ment of the Chiang Mai initiative (CMI) is perhaps age ASEAN common effective preferential tariff a precursor to expanded trade arrangements and (CEPT) fell from 12.76 percent to 2.91 percent over greater financial and macroeconomic cooperation the period 1993­2002. Because individual ASEAN and coordination.10 economies were already liberalizing their trade Proposals for northeast Asian economic cooper- before AFTA came into force, AFTA's common ation involving China, Korea, and Japan face many effective preferential tariff rates differ little from the obstacles such as agricultural protectionism and most-favored-nation (MFN) rates, and indeed for the complex and shifting political relations in two-thirds of the items in AFTA's inclusion list the northeast Asia.11 Proposals such as a Northeast two tariffs are the same. As a result, not much of Asian Council and a Northeast Asian Development intra-ASEAN trade has come in under the CEPT Bank have not gone far (He Fan 2002). tariff (Soesastro 2001). The most concrete proposal is for economic The integration of trade and investment--as cooperation between China and ASEAN. In well as, to some extent, services and factor move- November 2000 the fourth meeting of the ASEAN ments--among AFTA members stems more from + 3 leaders created a task force to study the possibil- the unilateral liberalization that has been under- ity of establishing an ASEAN free trade zone, and in taken by individual members and from the market November 2001, "to everyone's surprise," ASEAN forces integrating their economies, especially those leaders and Chinese premier Zhu Rongji endorsed that are close neighbors, than to the agreement the establishment of a free trade area between itself. Different AFTA members have different trade China and ASEAN economies within 10 years in policy regimes: Some have chosen to multilateralize the context of the Framework Agreement on their AFTA commitments; others have chosen the China-ASEAN Comprehensive Economic Cooper- same schedule of tariff reductions as AFTA but ation (He Fan 2002).12 Negotiations are ongoing, with a slightly higher end point on an MFN basis including on issues of product coverage. compared with the AFTA preferential rate. The envisaged scope of the agreement is quite AFTA has evolved more gradually than regional comprehensive. On trade in goods it goes beyond cooperation schemes in North America and Europe. the removal of tariffs and nontariff barriers to Strong expressions of political commitment, and include trade facilitation measures such as con- proposals to widen ASEAN cooperation, have not formity of standards and procedures, and it also always been followed up with concrete and broad- covers trade in services. The aim is to introduce tar- ranging implementation.8 A comprehensive and iff reductions for a certain group of products over integrated framework to cover investment, services, three years beginning prior to January 1, 2004, and trade and investment facilitation measures, compe- for other products over the agreed time frame of tition policy, and antidumping has been proposed, the free trade area--either to 2010 or to 2013. but it has not progressed very far. Other ASEAN Proposals Proposed Regional Arrangements Involving China In response to the proposed creation of the ASEAN- Various initiatives have been discussed in the last China Free Trade Area, other countries have now few years to strengthen cooperation in trade, shown interest in free trade agreements with New Regionalism: Options for East Asia 45 ASEAN. Japanese prime minister Junichiro Koizumi monitoring short-term capital flows, developing recently suggested an ASEAN-Japan Comprehen- early warning systems, assessing regional financial sive Economic Partnership, an arrangement that vulnerabilities, and preventing a future financial would include traditional as well as new elements crisis. The Asian Development Bank has dedicated related to facilitation, standards, and other forms of a monitoring unit to undertaking this surveillance cooperation. Korea has made a similar proposal, task as well as to providing training for central bank and at the APEC 2002 meeting in Mexico, even the officials and promoting other regulatory and mon- United States made a similar overture. President itoring initiatives. George W. Bush put forward a new trade initiative The most concrete initiative is the Chiang Mai called the U.S. Enterprise for ASEAN Initiative, initiative of May 2000. CMI expanded the existing under which ASEAN could liberalize its trade with ASEAN swap arrangements to include all ASEAN the United States. No timetable has been set. countries and set up a network of bilateral cur- In the last year or so, Thailand has begun a more rency swap and repurchase arrangements among deliberate pursuit of bilateral and regional arrange- ASEAN + 3 countries. The aim of the initiative is ments. Observers are concerned, however, about to provide additional short-term hard currency for how such arrangements will affect ASEAN eco- countries facing possible liquidity shortfalls. In nomic regionalism and cooperation. addition, the CMI will seek to achieve better mon- ASEAN is in a unique position at the hub of a itoring of capital flows, regional surveillance, and series of proposed preferential arrangements, but training of personnel (Henning 2002). The ceiling probably because of its own internal weaknesses for withdrawal is not high and would not have and lack of leadership it does not seem to be mak- been sufficient to prevent the liquidity shortfall ing the most of this position. Therefore, to achieve during the 1997­98 financial crisis. Some bilateral success ASEAN would be well advised to ensure agreements have reached their ceiling withdrawal that it has a clear vision of what it wants to achieve of $3 billion. The maximum available for with- from regionalism and that it has a coherent frame- drawal under each bilateral swap agreement is to work for deciding which agreements to enter. be negotiated between countries, because the terms of collateral and conditionalities need to be determined. By the end of March 2003, 12 bilateral Regional Financial Cooperation swap agreements involving $39 billion had been Since the Asian financial crisis, East Asian concluded under the CMI. In addition, two more economies have accelerated cooperation in the bilateral swap agreements proposed by Indonesia financial sector and macroeconomic manage- with China and Korea, respectively, were under ment.13 The initiatives range from institutional-- way. The implementation of the CMI bilateral such as the creation of an Asian Monetary Fund, swap agreements is to be reviewed in 2004, at sparked by the initial dissatisfaction with the Inter- which time the ASEAN + 3 group of countries may national Monetary Fund (IMF) bailout programs decide to amend the CMI framework arrange- at the beginning of the crisis--to the currency swap ment, to make it permanent, or to begin a process lines of the Chiang Mai initiative. In practice, the to transform it into a more formal institution for efforts have focused more on stabilizing currencies foreign exchange pooling.14 The June 2003 APEC than on coordinating macroeconomic policy or Summit launched the $1 billion Asian Bond Fund, achieving overall financial stability. a scheme that aims to reinvest a small portion of Financial cooperation to date has focused on the reserves of Asia's central banks within the strengthening surveillance and supervisory mea- region.15 It also endorsed initiatives to contain the sures within the region and on swap agreements. economic damage from the severe acute respira- ASEAN + 3 established an ASEAN + 3 Surveillance tory syndrome (SARS). Process in November 1999, based in the ASEAN According to He Fan (2002), China was not ini- Secretariat in Jakarta, to encourage the coordina- tially enthusiastic about financial cooperation in tion of macroeconomic and financial policies. The East Asia, but it has changed its attitude out of con- following issues will be covered by finance minis- cern that if it does not play a positive role in the ters at their meetings: enhancing cooperation in process, it will be kept out by other countries. The 46 East Asia Integrates Chinese government reportedly feels that trade and achieve. For smaller states, RTAs also offer the technological cooperation is more fundamental opportunity to build capacity in trade negotiations. and should be preconditions for financial coopera- But preferential agreements also carry risks, one tion. According to McKinnon (2001), a U.S. dol- of which is that trade may be diverted away from lar­linked currency standard can be used to achieve more efficient nonmembers to less efficient mem- regional exchange rate stability. Mundell (2003) bers. If a preferential agreement diverts more trade supports the idea of having an Asian "currency than it creates, it will yield smaller gains in effi- anchor," but believes that neither the Chinese yuan ciency than would multilateral liberalization. The nor the Japanese yen is a suitable candidate at the larger the difference between the preferential RTA present time. Meanwhile, Kuroda and Kawai (2003) tariff and the external tariff imposed on imports note that a U.S. dollar­ based exchange rate regime from nonmembers, the greater is the trade diver- makes an economy susceptible to fluctuations in sion effect. Trade also may be diverted if members effective exchange rates (e.g., when the dollar-yen of a free trade area impose rules of origin, because rate became volatile in 1995 and 1998). Such risks such rules may cause imports to be redirected need to be managed as well. The ASEAN Task Force through the member country that has the lowest on ASEAN Currency and Exchange Rate Mecha- external tariff. nism, which was established in March 2001, is dis- The literature on the benefits and costs of cussing ways of coordinating exchange rate policies regional trade agreements does not offer conclusive in the region.16 guidance on whether trade creation will outweigh trade diversion or on what such agreements will contribute to welfare.17 The view of Srinivasan, Costs and Benefits of Economic Whalley, and Wooton (1993) is that quantitative Regionalism assessments do not offer sufficient grounds for What can countries hope to achieve from preferen- either vigorous support of or vigorous opposition tial trade agreements (PTAs)? What does the litera- to preferential agreements. They reviewed work in ture say about the economic costs and benefits of this field based on computable general equilibrium these agreements? (CGE) models and found that the welfare effects of The economic benefit of a preferential trade preferential trade agreements had probably been agreement is the gains in economic efficiency positive but not very large.18 achieved from trade liberalization. A member Krueger (1995) found that trade diversion country will allocate resources to sectors in which it tended to occur less in those RTAs whose members has a comparative advantage vis-à-vis other mem- had high levels of bilateral trade before entering the bers, and trade is created. The greater size of the agreement. And Yeats (1998), looking at experience combined market can also contribute to economies in MERCOSUR (Mercado Común del Sur/South- of scale and attract foreign direct investment, ern Common Market Agreement), found that trade although because the benefits depend on the size of diversion can be minimized if external tariffs are markets as well as on the producers in the RTA, low and if members become more open by further they vary widely across countries. reducing their external tariffs as they enter the RTA. Other than the traditional economic benefit of Soloaga and Winters (1999), in a comprehensive greater efficiency, countries enter into RTAs for study of nine arrangements over 17 years, found no many development-related reasons. Such arrange- indication that the rise in preferential agreements ments offer the chance to create larger regional in the 1990s significantly boosted intrabloc trade. markets that are more attractive to foreign At the same time, for trade diversion they found investors; to sequence liberalization by opening up significant effects only in the European Union (EU) in stages; to lock in place unilateral reforms; and to and European Free Trade Area (EFTA). pursue structural change jointly with other mem- Davis, McKibbin, and Stoeckel (2000) found bers. Even though liberalization, trade facilitation, that a joint free trade arrangement between AFTA and structural reform are often thwarted by vested and CER would yield worthwhile benefits19: for interests, regional cooperation for the joint promo- AFTA an extra $25.6 billion and for CER an addi- tion of these processes can make them easier to tional $22.5 billion in GDP in net present value New Regionalism: Options for East Asia 47 (that is, in discounted cash flow) terms. The real "theory of new regionalism" complements and consumption gain, allowing for the ability to shift supplements the World Trade Organization. The spending through time, would be 1 percent by 2005 current initiatives in Asia and Asia-Pacific regional- for AFTA and 0.6 percent for CER.20 ism are responses to regionalism happening else- However, the same study also found that if where in the context of globalization, information APEC proceeds on schedule, the additional gains communication technology, and knowledge-based from the AFTA-CER arrangement would be rela- economies. tively small: those to AFTA would be just over $10 billion, and those to CER members would be just Free Trade or a Bloc-ed Up World? under $2 billion. Two reasons are given for this finding: (1) trade between AFTA and CER is small The question of whether regional trade arrange- compared with members' trade with APEC as a ments are stumbling blocks or building blocks in whole; and (2) APEC is not preferential, so that attempts to achieve a more open multilateral sys- some of the APEC gains could in fact be attributa- tem arises often. RTAs have not always been effec- ble to the AFTA-CER connection, to the extent that tive in locking in unilateral reforms,22 and it is not this arrangement encourages further liberalization clear whether, when compared with a multilateral by APEC members. approach, they are able to achieve deeper and faster Bayoumi and Mauro (2001) also examined the liberalization. Negotiations of RTAs have arguably costs, benefits, preconditions, and implications of been no more successful than multilateral negotia- an ASEAN regional currency arrangement that is tions in dealing with sensitive sectors and issues. assumed to culminate in a regional currency. They Indeed, RTA negotiations, like multilateral negotia- are of the view that, on economic criteria, ASEAN tions, have tended not to cover sensitive sectors appears less suited for a regional currency arrange- such as agriculture. Furthermore, regional arrange- ment than Europe before the Maastricht Treaty, ments may divert attention and resources away although the difference is not large.21 The experi- from multilateral and unilateral efforts to liberalize ences of the European Monetary Union (EMU) and facilitate trade. Especially in small economies, indicate that the path toward a common currency is concentration on negotiating regional agreements a difficult one. It requires a firm political commit- could have dire implications if it slows down ment to ensure that an attempt to form a regional progress on these fronts. currency arrangement is not viewed as simply These concerns aside, several recent studies have another fixed exchange rate regime open to specu- suggested that, under some circumstances, uncoor- lative crises. dinated preferentialism could lead to global free Clarete, Edmonds, and Wallack (2003) provide trade. Their results are derived from special condi- empirical analysis to show that the impact of PTAs tions--in which just one (differentiated) good or on trade flows varies widely across PTAs. They also service is traded in oligopolistic markets--but they find that the establishment of APEC and the EU provide some guidance and focus for cooperation. expanded trade significantly, both among members For one thing, consider the case in which, driven and to the rest of the world, providing evidence by producers' interest in getting access to rents in that PTAs can create rather than divert trade. foreign markets, economies form a bloc to which Although they used export data in this analysis, others seek membership. If the bloc's constitution their findings are consistent with those of Frankel says that everyone who wants to join must be let in (1997), who used total trade data (i.e., sum of on the same terms as the original members, then exports and imports). Frankel attributed the strong the excluded economies continue to seek to join positive effect on trade to the large share of total until everyone is a member. world trade accounted for by the APEC member Another possibility is the merging of blocs. countries. Meanwhile, Low (2003) concedes that Andriamananjara (1999), who considers the incen- bilateral and cross-regional trading arrangements tives to merge smaller blocs, starts with a situation are still "second best," and that broader regionalism in which each economy forms a bloc with one and multilateralism are still superior. The Doha neighbor. In the next stage, the members of one ministerial declaration has acknowledged that this agreement merge with those of one other agree- 48 East Asia Integrates ment, and so on. Could this process continue until tions, in turn, could stop the proliferation of agree- global free trade is reached, or will it stop before ments before free trade is reached. The issues that then? He finds that global free trade can be attained, are now on the trade liberalization agenda--for but only if the general level of inter-bloc tariffs is example, in services, standards, and investment-- low enough. When tariffs are low, the oligopolistic are likely to make preferential agreements increas- firms that populate this model can make greater ingly complex. The outcome, in practice, is likely to profits with unrestricted access to all markets. be a host of different agreements between different These two situations suggest that it would be groups of partners--the "spaghetti bowl" effect. worthwhile to introduce rules on the use of prefer- Latin America's overlapping preferential agree- ential agreements and, in particular, rules on acces- ments have produced this effect. Experience shows sion to such agreements. When bloc members have that in these circumstances the costs of doing busi- discretion over who can join, they will stop letting in ness increase, pushed up by inconsistencies among others before global free trade is reached. Therefore, different agreements--for example, different sched- without a rule on accession, free trade blocs will ules for phasing out tariffs; different rules of origin stop growing at a point short of universal free trade. and exclusions, conflicting product standards, and Network effects are a factor that could drive the differences in rules on antidumping and other regu- outcome of a preferential approach further toward lations and policies. The more dimensions there are free trade. When direct network effects exist, the to the agreements and the more agreements there value of a product increases with the number of are, the wider is the scope for such inconsistencies. consumers who use a compatible product. Gandal And if the costs of doing business increase, further and Shy (2001) show that when these effects are sig- trade diversion effects may arise. nificant, the incentive to form a standards union disappears. Distribution of Benefits: Hubs and Spokes Freund (2000), by examining serial bilateralism, provides another perspective on the process of con- Many of the proposed regional trading agreements solidation.23 She finds that the benefits of a bilateral in East Asia are of the so-called hub and spoke pat- agreement include higher profits from preferential tern.24 The concern here is that the balance of access to foreign markets and higher consumer sur- advantages in hub and spoke arrangements will plus at home. These gains always outweigh the loss tend to favor the hub because of its stronger bar- of profits in the home market and the loss of tariff gaining position and the greater attractiveness to revenue that results from offering foreign suppliers investors of its central location. Agreements on access at preferential rates. This is the case no mat- contentious matters will also tend to be tailored to ter what agreements are already in place and what the demands of the hub. The spokes, for their part, other economies have done. The best strategy, then, having secured their preferential access, will have is for each economy is to have a bilateral agreement an incentive to oppose the admission of new mem- with every other economy. In this model, in the bers into the circle. The rest of this section summa- presence of uncoordinated bilateralism the out- rizes the key findings of the substantial literature come is free trade. on hub and spoke arrangements. In reality, as Freund points out, costs are associ- Hub and spoke agreements do not provide equal ated with having a series of bilateral agreements, market access to all participants. Even if tariffs were including those associated with complex rules of removed along each spoke, the spoke countries origin. The issue of sensitive sectors is also ignored. would still not have free access to each other's mar- And the more dimensions there are to an agree- kets but only to that of the hub. ment, the more scope there is to apply discrimina- Snape (1996b) explains that in such systems tion in a variety of ways and the more difficult it there is an incentive to create tailor-made agree- becomes to bolt different agreements together ments to deal with products that the hub country (Snape 1996a; Findlay 2001). Cooperation with one regards as contentious. He also notes how small set of partners could make it more difficult to har- countries have incentives to join preferential agree- monize with other partners, and it could impede the ments, especially as more and more countries sign signing of new preferential agreements. Such condi- up with the hub economy. Indeed, the hub and New Regionalism: Options for East Asia 49 spoke system can "spread like a rash" (Baldwin (1996b) argues that each spoke country has paid a 1997). The spoke economies may or may not have price for its preferential access to the hub country, deals with each other--see, for example, Snape, and that it will resist further reductions of tariffs on Adams, and Morgan (1993); Anderson and Snape an MFN basis that erode the value of its special (1994); Wonnacott (1996); Snape (1996b). deals on sensitive products. Compared with all the other spokes, the hub At the least, the trade policy of the hub could economy benefits from its preferential access to each become a source of conflict among current and spoke economy. Only firms based at the hub get prospective members of the arrangement (Andria- duty-free inputs from each spoke. The hub country mananjara 1999). Spokes may seek to have new also gains if it diverts investment from each of the members come in with fewer and fewer benefits. spokes; its favored position gives producers access This objective may stem not only from domestic not only to the domestic market but also to those of interests in the spoke countries, but also from for- all the spoke economies. Furthermore, producers eign investors who have invested in these countries based in the hub are likely to be able to get more in order to gain access to the hub. (These investors inputs at low or zero tariffs than those based in the may be originally from the hub country, and they spokes, because they can source both from the hub will not be without influence in their old home.) and from any of the spokes. An inefficient pattern of In short, the hub and spoke structure is fraught investment may be perpetuated, as a result of iner- with many risks. Economies that are large enough tia, even if a hub and spoke system evolves into a have strong incentives to assume the role of hub free trade area (Wonnacott 1996). economy and to dominate a group of complemen- A spoke economy does not gain from free trade tary economies in their region. But doing so leads with other spokes; it could be damaged by discrim- to a structure of layers of discrimination and ination in other spoke markets; and its ability to potential conflict. A hub economy that is already compete in all markets against firms that are based dominant in economic terms can easily be per- in the hub might be reduced (Wonnacott 1996). It ceived as trying to acquire political dominance by can respond in three ways: biasing the rules of the new trading system in its favor (Wonnacott 1996). 1. It could organize equivalent agreements with other spoke countries. But doing so may be Assessing the Alternatives costly, and the risk is that a series of such agree- ments, negotiated one after the other, could sim- What type of regional arrangements will best serve ply add to the layers of discrimination as each the East Asia region? First, they should not become a pair deals with its own set of difficult issues. "stumbling bloc" to multilateral liberalization. Sec- 2. It could make one bloc with all the other spokes. ond, they should focus less on easing market access Realistically, however, because the original and more on facilitating trade measures and achiev- membership of the set of spokes was the result ing cooperation on external issues of common inter- of pressures from interest groups in the hub, it est and mutual benefit, such as some of the negotiat- may be difficult for this group to agree subse- ing issues in the WTO. Third, they should contribute quently and simultaneously on how to deal with positively to the multilateral trade liberalization a now larger set of contentious issues. process--for example, by including standstill provi- 3. It could unilaterally cut tariffs to the rest of the sions on further barriers to trade and investment; by world. Depending on the extent of these cuts, implementing a simultaneous program of reduction such an initiative could offset the investment of barriers to nonmembers; by allowing the most lib- diversion effects. The spoke agreement would eral rules of origin possible; and by avoiding the then be part of a transition to free trade, but an spaghetti bowl outcome for other issues. expensive one. In hindsight, a giant leap to free trade is preferable to a couple of small steps. ASEAN and China Hub and spoke mechanisms can lead to greater Adjustments in industrial trade and investment resistance to multilateral liberalization. Snape resulting from China's accession to the WTO are 50 East Asia Integrates likely to occupy the East Asian economies for the way it can remain competitive in the global market next decade. ASEAN member countries can count is to relocate production plants in China, as it has on both increased export competition with China been doing for some time. This trend will now and new opportunities within China. This section accelerate and could have serious implications for briefly explores some of the issues this scenario ASEAN. Singapore may find it difficult to compete raises for a preferential agreement between ASEAN with high-tech exports produced in China. and China. Like other countries, ASEAN members will also At the outset it is important to emphasize that have access to the more open and growing Chinese among the ASEAN economies, three of the poor- market, where import tariffs on ASEAN manufac- est--Vietnam, Lao PDR, and Cambodia--are not tured products will continue to come down, from yet WTO Members and, as a result, do not have 15 to 10 percent over the next five years. Quotas MFN status to export to China. It is a priority to and quantitative restrictions will be removed and ensure that these three members are integrated into replaced by tariff-rate quotas. These developments the WTO system as soon as possible to avoid unbal- will be important for ASEAN agricultural products anced market access outcomes. such as palm oil, rice, and sugar. Other nontariff Now that China has acceded to the WTO, the barriers and investment-related measures, such as new pressure from competition will further local content and trade-balancing requirements, sharpen China's competitiveness and its ability to also will be removed, immediately or gradually. win market shares, including in the domestic mar- Under its accession commitment, China will open kets of ASEAN member countries. This enhanced up its services sector over the next five years. Espe- competitiveness is already affecting labor-intensive cially important for ASEAN economies will proba- products such as textiles, garments, electronics/ bly be professional services, tourism, and the possi- electrical appliances, footwear, and toys. Further- bility of some professional labor migration. more, China will now have the same MFN treat- China's demand for imports from ASEAN is ment as all other ASEAN member countries. With likely to increase. Products likely to benefit include this status, it becomes a more stable supplier--a oil and gas, wood, rubber, food and other agricul- desirable feature from the point of view of import- ture-based products, as well as some manufactured ing firms. products such as electrical machinery. It is up to The planned removal of quotas on textiles and ASEAN member economies to ensure their com- garments, if fully implemented by 2005, will also petitiveness in supplying these products to China. mean that the ASEAN economies will have to com- Other than greater market access, the hope is pete openly with China in third-country markets, that China's WTO accession will result in greater and it is likely that ASEAN, and other developing transparency and certainty in laws and regulations Asian economies in South Asia, will lose market and their implementation. Surveys of businesspeo- share to China. In the U.S. market for textiles and ple in Malaysia and Singapore have revealed some clothing, for example, ASEAN countries have up to of the common problems experienced by ASEAN now managed to maintain and sometimes slightly and other investors in doing business in China. increase their share, aided by the quota allocations They cite unpredictable laws, uncertain product in this market.25 But in Japan's more openly com- standards, a weak legal infrastructure, insufficient petitive market, they have been losing ground to trade facilitation measures, inefficient bureaucracy, China; 62 percent of Japan's imports of textiles and low quality of work, poor protection of intellectual apparel now come from China and only 8 percent property, financial market restrictions, and poor come from ASEAN.26 enforcement legislation. China is also increasingly competitive in high- China's accession will eventually benefit ASEAN tech products. There are predictions that by 2006 as a whole. China's restructuring is providing an chip manufacturing in Shanghai will be as big as impetus for the establishment of new regional pro- that in Taiwan (China), positioning China as an duction networks, initially in electronics, which important competitor to both Taiwan and Singa- promote more productive firms. For now, these pro- pore. Taiwan is fully aware that, politics aside, it duction networks are oriented toward export to must integrate economically with China. The only developed countries, but the growing markets New Regionalism: Options for East Asia 51 within East Asia provide a potential complementary encouraged Chinese companies to invest in and source of demand.27 The Framework Agreement on contract for major engineering and construction China-ASEAN Comprehensive Economic Coopera- projects (Far East Economic Review, March 28, tion, with its "Early Harvest" provisions, provides 2002). This go-abroad policy appears to be another avenue through which the poorer countries designed to expose Chinese firms to international in the region, including non-WTO Members,28 can business practices as well as to achieve resource benefit from trade opportunities in the region. security, given this high-growth economy's demands for fuel, minerals, and other resources. For example, the Chinese state-owned offshore oil ASEAN-China Economic Relations company, CNOC, recently acquired the Spanish oil ASEAN and China have strengthened their mutual company Repsol-YPF for its Indonesian oil and gas trade and investment ties in the last decade, pro- assets.29 Other gas purchase deals between Indone- pelled by the dynamic growth in China, and in sia and China are being negotiated. But because of ASEAN before the financial crisis; by the liberaliza- the sluggish growth of most economies in the tion undertaken by individual countries; and by group, ASEAN in general remains a less attractive their geographic proximity. destination for Chinese investment than Latin ASEAN as a region is China's fifth largest trading America, the United States, or Europe. Thus far, it partner, after the United States, Japan, the EU, and accounts for 20 percent of China's outward invest- Hong Kong. ASEAN's share of China's trade is still ment. quite small, at around 8 percent, even though it has ASEAN and China have important relationships been growing by 20 percent a year during the last in services, especially tourism, finance, and decade. Between 1991 and 2000 ASEAN increased telecommunications. China's growing prosperity its share of China's exports from 5.7 to 6.9 percent means that an increasing number of Chinese and of China's imports from 6 to 9.9 percent. tourists are visiting ASEAN countries. Indeed, 2.2 ASEAN exports to China still consist mainly of million visited ASEAN countries in 2000, particu- resource- or agriculture-based products (minerals, larly Malaysia, Singapore, Thailand, and Vietnam. pulp, wood, vegetable oil, rice, and sugar). How- For their part, ASEAN tourists are visiting China in ever, a growing share consists of machinery and increasing numbers. One drawback to investment electrical components for assembly into final goods and the attractiveness of tourist destinations in in China. These exports are linked to trade-related ASEAN members is the discriminatory treatment investments and are part of the multinational of ethnic Chinese minorities in some of these regional production structure as well as the countries, especially Indonesia. This problem needs regional pattern of sourcing (ASEAN-China Expert to be overcome and properly managed because of Group on Economic Cooperation 2001; Yusuf, its sensitive nature. Altaf, and Nabeshima 2003). Intraindustry trade also takes place in textiles and fibers. Proposed ASEAN-China Free Trade Area: Investment relationships between ASEAN and The Goals and Means to Achieve Them China also have been growing stronger. Invest- ments have flowed from ASEAN countries into In 2001 China and ASEAN agreed to establish the China in the last decade, with varying degrees of ASEAN-China Free Trade Area within 10 years. The success. Singapore was one of the first Southeast stated motivation behind this initiative is to take Asian countries to enter when China opened up to advantage of complementarities and build on exist- foreign investment in the late 1970s; reflecting ing strengths, to make the region collectively more ancestral links, the investments went mainly to efficient and competitive, and to attract investment. small businesses in Guangdong and Fujian. By the The aim is to use the enhanced efficiency to compete end of 2001, Singapore was the fifth largest investor in third-country markets, as well as to provide mem- in China, with a cumulative realized investment of bers with preferential entry to each other's markets $18.6 billion in that country. (China being the largest). The experience of the Recently, China itself has begun a vigorous out- ASEAN Free Trade Area and other RTAs that have ward investment drive, and the government has been introduced recently suggests that any kind of 52 East Asia Integrates ASEAN-China free trade area must go beyond liber- to import palm oil); and issues of testing, stan- alization of cross-border barriers. The proposal pre- dards, and labeling requirements, investment pro- pared by the expert group is in fact comprehensive motion and protection, visa facilitation, and infra- and covers trade and investment liberalization in structure development. The agreement may also goods and services, trade and investment facilitation, provide opportunities to build up the capacity of capacity building and technical assistance, and coop- ASEAN's less developed members. eration in various areas (ASEAN-China Expert But the idea is not yet widely accepted by Group on Economic Cooperation 2001). ASEAN member countries, and some of them have Certainly the agreement promises big economic indeed expressed concern about their ability to benefits. Estimates based on the study by the expert compete with China and about whether a free trade group indicate that the ASEAN-China agreement area would be mutually beneficial. Thailand and would augment trade on both sides by 50 percent Singapore seem to be the most supportive of the and increase GDP by 0.9 percent for ASEAN and by idea, with Indonesia and the Philippines taking a 3 percent for China. China's exports to ASEAN moderate position, Malaysia adopting a careful countries would be 55 percent greater with the stance, and the lower-income economies within agreement than without it, and ASEAN countries' this group--Cambodia, Lao PDR, Myanmar, and exports to China would be 48 percent greater.30 Vietnam--expressing the most concern about their For ASEAN, the agreement would provide first- ability to compete and mutually benefit from such mover advantages in the Chinese market before it is an arrangement. China, for its part, has an interest opened on an MFN basis (Box 3.2). As well as pro- in the reduction of ASEAN tariffs and trade barri- vide opportunities for the "Early Harvest," it is ers on products of special export interest, including hoped the agreement will address various nontariff motorbikes and vehicle components. barriers of concern to ASEAN members, such as As He Fan (2002) suggests, it may be that China quotas on palm oil and other agricultural products; wants to assuage ASEAN countries' fear of the China's complex import procedures (whereby, for "China threat" by opening up its markets to them. example, only state-owned enterprises are allowed Other than the relatively comprehensive economic BOX 3.2 "Early Harvest" "Early Harvest," which refers to provisions of the · Measures enhancing market access opportu- Framework Agreement on China-ASEAN Com- nities for specific products or services of inter- prehensive Economic Cooperation, would liber- est to ASEAN and China, such as agricultural alize tariffs in priority sectors of interest and and tropical products, textiles and clothing, implement other trade and investment facilita- machinery and electronic products, footwear, tion measures that are deemed to generate oils and fats, foodstuffs, forestry and aquacul- immediate benefits to the ASEAN and Chinese ture products, and energy (the list of prod- business communities. These measures could ucts and services will be determined by include: mutual consultation) · Development and technical assistance to · Extension of MFN treatment of China's acces- build capacity among countries, particularly sion commitments to non-WTO members of for the new members of ASEAN, in order to ASEAN in compliance with WTO rules improve their competitiveness · Any other measures delivering immediate · Trade and investment facilitation measures mutual benefits. · Trade policy dialogue · Business sector dialogue Source: SEOM-MOFTEC Inputs on Forging · Facilitation of visa arrangements for business- Closer ASEAN-China Economic Relationship, people Second SEOM-MOFTEC Meeting, Brunei Darus- · Standards and conformity assessment salam, October 28, 2001. New Regionalism: Options for East Asia 53 integration proposed, such an agreement and Other than restructuring and maintaining their cooperation would enhance the sense of commu- competitive edge, individual ASEAN countries nity within the region, reduce tensions, and con- should continue on the current track of unilateral tribute to stability. It may also enhance members' liberalization and reforms, while exploring regional cooperation on the external front. For China and cooperation such as that being proposed through ASEAN alike, the process of getting to know each the ASEAN-China initiative. other is important, and it is hoped that a trade agreement will provide a basis for developing a Conclusions: The Way Forward cooperative stance on external issues of common concern such as the global financial architecture. In Regional economic cooperation arrangements in the near term, given that ASEAN's share of China's East Asia raise complex issues, and decisionmakers trade is still small, the political impact of the need to understand and assess the options carefully, ASEAN-China free trade area may be more impor- especially in view of the limited resources and tant than the economic impact. capacity of some of the region's economies. Progress in implementing the agreement is likely East Asian economies seek increased economic to be slow,even though it is known that the wider and efficiency and competitiveness rather than eco- deeper the coverage, the greater will be the benefits. nomic integration in the manner of Europe, and The agreement will probably be confined initially to they can pursue these goals through enacting uni- eliminating tariffs on goods. Sensitive sectors such as lateral reforms and through ensuring an open, agriculture will likely be excluded, and progress will rules-based multilateral trading system. be slow in other areas such as the removal of nontar- The role of regional trade arrangements is to iff barriers, liberalization of services, investment, and facilitate and build toward this outcome. The way other New Age RTA issues such as treatment of work- forward in the first instance is to ensure that unilat- ers, environmental standards, and the movement of eral reforms are continued, that commitments to labor. But if the ASEAN-China Free Trade Area is to the multilateral process are faithfully met, and that be a building block for deeper economic integration economies are preparing themselves to maximize within the region, the framers of the RTA would be the benefits from future developments in the multi- well advised to focus not just on the nitty-gritty of lateral trading system.32 In the process of accession item-by-item tariff lines and exclusions, but also on to the WTO, for example, China had to concede on ways to increase competition and efficiency and real many sectors and issues. In the forthcoming negoti- economic benefits.31 In this regard, Elek (2000) sug- ations, China will now be able to press for market gests focusing on trade facilitation through, for access for products and services of importance to example, simplifying customs procedures and har- its economy, and for discipline and fairness in the monizing standards. rules of the game as well. To become viable suppliers to the Chinese mar- Progress on involving a large group of ket and to compete successfully with China in their economies can be expected to occur through func- own and export markets, ASEAN member countries tional cooperation in the financial and monetary need to anticipate and prepare for the structural area. The tasks to be carried out seem likely to be changes that will happen in China. Meanwhile, their confined to surveillance, technical assistance and own comparative and competitive advantages need capacity building, information sharing, and limited to be continually strengthened. Companies in swap arrangements. Joint macroeconomic policy ASEAN must focus on specialization and product coordination or currency union seems unlikely in differentiation, including in terms of quality. Each the near future. member country should devise a comprehensive Given the limited experience with regionalism program to achieve such goals as well as an action in East Asia to date, progress on an inclusive plan that anticipates the institutions, human regional trade agreement is likely to be slow. The resources, and infrastructure needed to support the ASEAN Free Trade Area lacks scope and depth, as program. (In fact, such a program is likely to be a noted earlier, and the ASEAN-China Free Trade more helpful response to the China challenge than Area being negotiated is likely to suffer from the is a regional trade arrangement per se.) same problems because it is based on AFTA. But it 54 East Asia Integrates is possible that ASEAN-China cooperation can Endnotes extend to forging common positions for interna- 1. ASEAN has 10 members: Brunei, Cambodia, Indonesia, tional negotiations, such as in the WTO. Lao People's Democratic Republic, Malaysia, Myanmar, the Both in the formal regional arrangements such Philippines, Singapore, Thailand, and Vietnam. 2. All dollar amounts are current U.S. dollars. This GDP is as AFTA and in the informal regional processes half of Japan's, but if current growth rates are maintained, such as APEC the goal up to now has been more to it will catch up with Japan's in the next 5­10 years. In pur- stimulate and support unilateral reforms and con- chasing power parity (PPP) terms, China's GDP now tribute to multilateral liberalization than to achieve exceeds that of Japan (World Bank 2002). 3. Until a few years ago, apart from the ASEAN Free Trade Area greater regional integration or intraregional trade (AFTA), East Asia had no formal regional cooperation agree- and investment. In such arrangements in East Asia, ments. A proposal by Malaysian prime minister Mahathir it may be that the process of regional cooperation is Mohamad in December 1990 to form an East Asian eco- nomic grouping elicited strong protests from U.S. Secretary more important to the participants than outcomes. of State James A. Baker III, and support of the proposal from If this is so, the agreements should be evaluated not Japan or China was not forthcoming. There was in general a just on the basis of whether they achieve free trade strong sentiment against institutionalizing regional coopera- tion or regionalism. See also Munakata (2001). and investment, but also according to the process 4. The texts of most regional trade agreements (RTAs) can be whereby each country actually makes progress in found on the Web site of the RTA secretariat (e.g., the unilateral reforms.33 And it may be that the infor- ASEAN Secretariat at www.aseansec.org or the European mal process needs to be nurtured more systemati- Union Secretariat at europa.eu.int) or of the relevant administering government department in one or more cally so that it will evolve into a more institutional member countries (e.g., the Singaporean Ministry of Trade one. Whether this means creating an East Asian sec- and Industry at www.mti.gov.sg or the New Zealand Min- retariat needs further debate and deliberation. istry of Foreign Affairs and Trade at www.mfat.govt.nz). The texts of all agreements in the Americas are available on Is ASEAN/AFTA the natural locus for regional the Web site of the Organization of American States cooperation? Some analysts believe that it is not in (www.sice.oas.org). Useful reviews of developments under a strong position to serve this important role particular RTAs also are available. See, for example, Pelk- mans (1997), Thanadsillapakul (2001), Lloyd (2002), because of weaknesses within ASEAN itself, which Mahani (2002), and Low (2003). The Asian Development have resulted partly from lack of leadership and Bank (2002: Part III) reviews RTAs in the Asia-Pacific area, slow progress on new areas of cooperation (Soesas- and Estevadeordal (2002) provides an excellent compari- tro 2001). It has been suggested that to strengthen son of some features of RTAs in the Americas. See also the WTO's World Trade Report (2003: 46­66) on recent devel- ASEAN the implementation of AFTA must be opments in the realm of RTAs. accelerated and the ASEAN economies must 5. Kaminsky and Reinhart (1999) have shown that output rethink their vision of regionalism. losses are much more significant when a currency crisis is accompanied by major financial sector problems. To conclude, broader liberalization on a multilat- 6. The distinct period for each group of economies is also the eral basis will lead to greater net benefits, but East period in which flows of foreign direct investment, espe- Asia is likely to realize net gains if it pursues comple- cially intraregional FDI, accelerated into that group. 7. This section provides only an introductory overview of the mentary regional approaches simultaneously with various regional and multilateral trade arrangements in multilateralism. One further benefit from stronger East Asia, and the following questions may need further regional cooperation could be a more effective research: How large is intra-ASEAN Free Trade Area stance at the WTO on issues of common interest-- (AFTA) trade relative to AFTA's total trade? How much intra-AFTA trade is excluded because of "sensitive sectors"? for example, in achieving further discipline on How would these numbers change if China were added? antidumping. China would play a critical role in this How much variance is there among AFTA and ASEAN regard because of its growing economic dominance members in how much of their trade is covered? The answers to these questions will influence how different in the East Asian region, if not the world, and thus countries and groups approach others and their approach the type of regional cooperation arrangement in to liberalizing in other geographic areas. which it chooses to participate will have a lasting 8. There is evidence that some AFTA members have fallen behind in their reform schedules. One example is delays in impact on the course of events and pace of develop- including the automotive and petrochemical sectors in the ment in the region. Economic development in AFTA by Malaysia and the Philippines, respectively. China can only strengthen the perception of East 9. The main reason is that because of China's WTO accession Asia as a "good neighborhood" and place to be these three have separate customs territories, and so China cannot give preferential treatment to Hong Kong or Macao within the global community. unless they are members of a free trade agreement. New Regionalism: Options for East Asia 55 10. Eichengreen (2001) finds that unlike in the EU countries gration, reduced transaction costs, and higher levels of and those under the North American Free Trade Agree- trade and investment that may result from exchange rate ment (NAFTA), in East Asia monetary and financial coop- stability within ASEAN. The costs come from the loss of eration has tended to precede cooperation in trade. monetary autonomy involved, which limits the macroeco- 11. Such as the role of China in the Korean Peninsula situation nomic policy options available to deal with unexpected and Japan's alleged support for Taiwan. macroeconomic shocks (Bayoumi and Mauro (2001). See 12. Agreement was also reached at the November 2000 meeting also Lloyd (2002). on reviewing the possibility of an East Asian economic 22. There is some evidence that this has worked effectively in zone. The fact that the ASEAN-China economic coopera- some cases in the Western Hemisphere, but in other cases tion has forged ahead faster than East Asia-wide coopera- unilateral reforms have been delayed as economies held tion (or a similar earlier idea for cooperation between back moves to open up until they entered into RTA and Japan and ASEAN) indicates, some would say, the current WTO negotiations. lack of vision and strategy shown by Japanese leaders about 23. Freund uses the same segmented oligopolistic market Japan's leadership in the region (www.Asahi.com, Novem- model as Andriamananjara (1999), but now governments ber 2001). The main issue in Japan is resistance to opening maximize welfare--that is, the sum of producer and con- up its agriculture sector. sumer surplus, plus tariff revenue. The outcome of unilat- 13. For details, see Yusuf and others (2003: Chap. 5). eral policymaking is that each government sets the optimal 14. Henning (2002) and Kuroda and Kawai (2003) provide tariffs on imports from other economies, taking those details on recent bilateral swap agreements under the CMI economies' tariffs as given. In this setting, when bilateral and discussions of related issues. agreements are possible, each country wants to sign an 15. Asia's official reserves account for over half of global agreement with every other country. reserves. See "APEC Declining Relevance" (2003). 24. Suppose there are three countries: A, B, and C. Country A 16. See Kuroda and Kawai (2003). Manupipatpong (2002) concludes separate agreements with B and C, but B and C reviews the development of regional surveillance and self- do not have an agreement with each other. Country A is the help mechanisms through the ASEAN Surveillance Process hub, and B and C are the spokes. Lloyd and Crosby (2002), and the Chiang Mai initiative. using this example, point out that the entity at the hub 17. Much depends on the assumptions made and the methods could itself be a regional trading agreement. used to measure trade creation and diversion, and whether 25. James, Ray, and Minor (2002). Before its WTO accession, the effects measured are dynamic or static. Some authors China was not eligible to take advantage of growth in the find a positive net trade creation effect (Salazar-Xirinachs U.S. quotas. 2001), while others argue that in some cases there has been 26. For example, from 1996 to 2001 China's share of Japan's a serious net trade diversion effect, such as in North Amer- market for cotton knit apparel increased from 47.3 to 77.3 ica (Panagariya 2000). The balance between positive and percent, and for manmade fiber knit apparel it increased negative effects also may differ between members and non- from 59.1 percent to 80.4 percent. Japan does not impose members of an RTA. For example, dynamic benefits might bilateral quotas, and thus its market reflects more open outweigh positive effects for members, but nonmembers competition. might still be adversely affected by trade diversion as a 27. China's domestic market is the largest of these, but recent result of their exclusion.Yeats (1998) found net trade diver- trends show Korea, Thailand, and other economies grow- sion in the case of MERCOSUR (Mercado Común del ing on the strength of domestic consumption demand. Sur/Southern Common Market Agreement), where a con- 28. Vietnam, Lao PDR, and Cambodia are not yet WTO Mem- tributing factor seems to have been the high margins of bers. preference. However, for some pairings the cost in terms of 29. This acquisition, for $584 million, is the biggest foreign trade diversion may be relatively small when trade barriers acquisition of Indonesian oil and gas assets in the last are already low. The preferential trade agreement route decade (Far East Economic Review, March 28, 2002). then looks like a low-risk option. 30. He Fan (2002: 14), quoting Zhou Keren, the Chinese vice 18. CGE models have the advantage of providing more options minister of foreign trade and economic cooperation. for good choices of base scenarios and for capturing more 31. According to Chirathivat (2002), the ASEAN-China FTA of the detail of preferential policies. The models are becom- has "contributed to the rethinking of East Asia, not for only ing increasingly sophisticated and able to incorporate scale a geographical concept, but more strongly as an institu- effects, imperfect competition, and capital accumulation. tional arrangement." However, both ASEAN and China Panagariya (2000), however, identifies some problems in remain highly dependent on outside markets "rather than a the modeling approaches. Appendix C of "The Angkor self-fulfilling grouping among themselves" to further stim- Agenda" (available from www.aseansec.org/aem/angkor_ ulate their economic growth. agenda.pdf) also lists some empirical papers. 32. For both the Maastricht Treaty and North American Free 19. They use the APG-cubed model (18 countries and 6 sec- Trade Agreement (NAFTA), crucial unilateral domestic tors), which permits the identification of dynamic gains decisions that reflected the pressures of rising openness and allows for allocative efficiency effects, terms of trade preceded the negotiations. For the Maastricht Treaty, changes, and capital accumulation for goods and for ser- French and Italian financial market liberalization was a vices, as well as endogenous productivity effects. prerequisite, and for NAFTA, Mexican trade liberalization 20. These gains are nearly three times as great as those found was a necessary precondition (Milner 1998). by an earlier study that excluded services liberalization and 33. For example, APEC had the purported role of helping to the productivity effect. break the deadlock in the Uruguay Round negotiations in 21. On the basis of the theory of optimum currency areas, the 1993. benefits emanate mainly from the greater economic inte- 56 East Asia Integrates References Freund, Caroline. 2000. "Spaghetti Regionalism." International Finance Discussion Paper No. 680. Federal Reserve Board, The word processed describes informally reproduced works that September. Available at www.federalreserve.gov/pubs/ifdp/. may not be commonly available through libraries. Gandal, N., and O. Shy. 2001. "Standardization Policy and Inter- national Trade." Journal of International Economics 53: Anderson, Kym, and Richard Snape. 1994. "European and 363­83. American Regionalism: Effects on and Options for Asia." He Fan. 2002. "Regional Economic Cooperation: China's Per- Journal of the Japanese and the International Economies 8 (4): spective." Background Paper for EAS Country Unit. World 454­77. Bank, Washington, D.C., June. Andriamananjara, Soamiely. 1999 "On the Size and Number of Henning, Randall C. 2002. "East Asian Financial Cooperation." Regional Integration Arrangements: A Political Economy Policy Analyses in International Economics 68 (September). Model." Policy Research Working Paper 2117. World Bank, Institute for International Economics, Washington, D.C. Washington, D.C. James, William E., David J. Ray, and Peter J. Minor. 2002. "APEC Declining Relevance." 2003. Oxford Analytica Brief. June "Indonesia's Textile and Apparel Industry: Meeting the Chal- 18. lenges of the Changing International Trading Environment." ASEAN (Association of Southeast Asian Nations)-China Expert Report for Ministry of Industry and Trade, Republic of Group on Economic Cooperation. 2001. Forging Closer Indonesia, USAID/ECG, Jakarta, July. ASEAN-China Economic Relations in the 21st Century. Kaminsky, Graciela, and Carmen Reinhart. 1999. "The Twin Report submitted to the ASEAN SEOM-MOFTEC, October. Crises: The Causes of Banking and Balance of Payments Asian Development Bank. 2002. "Preferential Trade Agreements Problems." American Economic Review 89 (3): 473­500. in Asia and the Pacific." Asian Development Outlook. Avail- Krueger, Anne O. 1995. "Free Trade Agreements versus Customs able at www.adb.org/Documents/Books/ADO/2002/ Unions." NBER Working Paper No. 5084. National Bureau of pta0000.asp. Economic Research, Cambridge, Mass., April. Baldwin. R. 1997. "The Causes of Regionalism." World Economy Kuroda, Haruhiko, and Masahiro Kawai. 2003. "Strengthening 20 (7). Regional Financial Cooperation in East Asia." Pacific Eco- Bayoumi, Tamim, and Paolo Mauro. 2001. "The Suitability of nomic Papers. Australia-Japan Research Centre, Australian ASEAN for a Regional Currency Arrangement." World Econ- National University, Canberra, May. omy 24 (7): 933­54. Lloyd, Peter. 2002. "New Bilateralism in East Asia." World Econ- Chia Siow Yue. 2002. "East Asian Regionalism." Paper presented omy 25 (9): 1279­96. at East Asian Cooperation: Progress and Future Agenda, Lloyd, Peter, and Mark Crosby. 2002. "China's Options for Institute of Asia-Pacific Studies (Chinese Academy of Social Regional Integration and Cooperation: Conceptual and Sciences, CASS) and Research Center for APEC and East Theoretical Framework." Background Paper for EAS Coun- Asian Cooperation (CASS), Beijing, August 22­23. try Unit. World Bank, Washington, D.C., June. Chirathivat, Suthiphand. 2002. "ASEAN-China Free Trade Area: Low, Linda. 2003. Multilateralism, Regionalism, Bilateral and Background, Implications and Future Development." Jour- Crossregional Free Trade Arrangements: All Paved with nal of Asian Economics 13: 671­86. Good Intentions for ASEAN. Asian Economic Journal 17 (1): Clarete, Ramon, Christopher Edmonds, and Jessica Seddon Wal- 65­86. lack. 2003."Asian Regionalism and Its Effects on Trade in the Mahani, Zainal-Abidin. 2002. "ASEAN Integration: At Risk of 1980s and 1990s." Journal of Asian Economics 14: 91­129. Going in Different Directions." World Economy 25 (9): Crawford, Jo-Ann, and Sam Laird. 2000. "Regional Trade Agree- 1263­77. ments and the WTO." Paper prepared for meeting of North Manupipatpong, Worapot. 2002. "The ASEAN Surveillance American Economic and Finance Association, Boston, Janu- Process and the East Asian Monetary Fund." ASEAN Eco- ary 6­9. nomic Bulletin 19 (1): 111­22. Davis, Lee, Warwick McKibbin, and Andrew Stoeckel. 2000. McKinnon, Ronald. 2001."After the Crisis, the East Asian Dollar "Economic Benefits from an AFTA-CER Free Trade Area: Standard Resurrected." In Joseph Stiglitz and Shahid Yusuf, Year 2000 Study." Centre for International Economics, Can- eds. Rethinking the East Asian Miracle. Washington, D.C.: berra and Sydney, June. World Bank and Oxford University Press. Eichengreen, Barry. 2001."Hanging Together? On Monetary and Milner, Helen. 1998. "Regional Economic Cooperation, Global Financial Cooperation in Asia." University of California, Markets and Domestic Politics: A Comparison of NAFTA Berkeley. Processed. and the Maastricht Treaty." In William D. Coleman and Elek, Andrew. 2000. "Sub-Regional Trading Arrangements Geoffrey Underhill, eds., Regional and Global Economic Inte- among APEC Economies: Managing Diversity in the Asia gration: Europe, Asia and the Americas. London and New Pacific." Pacific Economic Papers No. 309. Australia-Japan York: Routledge Publications. Research Centre, Australian National University, November. Munakata, Naoko. 2001."Focus on the Benefits--Not Threats-- Estevadeordal, A. 2002. "Traditional Market Access Issues in of Regional Economic Integration." PacNet Newsletter, RTAs: An Unfinished Agenda in the Americas?" Paper pre- November 16. Centre for Strategic and International Stud- sented at WTO Seminar on Regionalism and the WTO, ies, Washington, D.C. Geneva, April 26. Mundell, Robert. 2003. "Prospects for an Asian Currency Area." Findlay, Christopher. 2001. "Old Issues in New Regionalism." Journal of Asian Economics 14: 1­10. Pacific Economic Papers 311. Australia-Japan Research Cen- Panagariya, Arvind. 2000."Preferential Trade Liberalization: The tre, Australian National University, January. Traditional Theory and New Developments." Journal of Eco- Frankel, J. 1997."Regional Trading Blocs in the World Economic nomic Literature 38 (June): 287­331. System." Institute for International Economics, Washington, Pelkmans, J. 1997. European Integration: Methods and Economic D.C. Analysis. New York: Longman. New Regionalism: Options for East Asia 57 Salazar-Xirinachs, José M. 2001. "Implications of Proliferating Wonnacott, Ronald. 1996."Trade and Investment in a Hub-and- Sub-Regional Trade Agreements: Lessons from the Latin Spoke System versus a Free Trade Area." World Economy 19 American Experience." Pacific Economic Cooperation (3): 237­52. Council (PECC), Trade Policy Forum, Seminar on Regional World Bank. 2002. World Development Indicators 2002. Wash- Trade Agreements, Bangkok, June12­13. ington, D.C.: World Bank. Snape, Richard. 1996a. "Trade Discrimination--Yesterday's WTO (World Trade Organization). 2000a. "Synopsis of `Sys- Problem?" Economic Record 72 (December): 381­96. temic' Issues Related to Regional Trade Agreements." ---------. 1996b. "Which Regional Trade Agreement?" In Bijit WT/REG/W/37. Note by the Secretariat, March 2. Bora and Christopher Findlay, eds. Regional Integration and ------. 2000b. "Committee on Regional Trade Agreements, the Asia Pacific. Melbourne: Oxford University Press. Mapping of Regional Trade Agreements." WT/REG/W/41. Snape, Richard, Jan Adams, and D. Morgan. 1993. Regional Trade Note by the Secretariat, October 11. Agreements: Implications and Options for Australia. Report ------. 2002. "Regional Trade Integration under Transforma- for the Australian Department of Foreign Affairs and Trade, tion." Paper prepared for the World Trade Organization's AGPS, Canberra. seminar Regionalism and the WTO, Geneva, April 26. Soesastro, Hadi. 2001."Whither ASEAN Plus Three?" Paper pre- ------. 2003. World Trade Report. Geneva, August. sented to Pacific Economic Cooperation Council (PECC) Xu, X., and L. Song. 2000. "Export Similarity and the Pattern of Trade Policy Forum, Seminar on Regional Trading Arrange- East Asian Development." In P. Lloyd and X, Zhang, eds., ments, Bangkok, June 12­13. China in the Global Economy. Cheltenham, UK: Edward Elgar. Soloaga, I., and A. Winters. 1999. "Regionalism in the 1990s: Yeats, Alexander. 1998. "Does Mercosur's Trade Performance What Effect on Trade?" Discussion Paper No. 2183. Centre Raise Concerns about the Effects of Regional Trade Agree- for Economic Policy Research (CEPR), London. ments?" World Bank Economic Review 12 (1):1­28. Srinivasan, T. N., John Whalley, and Ian Wooton. 1993."Measur- Yusuf, Shahid, M. Anjum Altaf, and Kaoru Nabeshima, eds. ing the Effects of Regionalism on Trade and Welfare." In 2003. Global Production Networking and Technological Kym Anderson and Richard Blackhurst, eds. Regional Inte- Change in East Asia. Washington, D.C.: World Bank and gration and the Global Trading System. London: Harvester Oxford University Press. Wheatsheaf for the GATT Secretariat. Yusuf, Shahid, with M. Anjum Altaf, Barry Eichengreen, Sudar- Thanadsillapakul, L. 2001. "Open Regionalism and Deeper Inte- shan Gooptu, Kaoru Nabeshima, Charles Kenny, Dwight H. gration: The Implementation of ASEAN Investment Area Perkins, and Marc Shotten. 2003. Innovative East Asia: The and ASEAN Free Trade Area." Processed. Future of Growth. Washington, D.C: World Bank and Oxford University Press. 4 Market Access Barriers and Poverty in Developing East Asia Bijit Bora with additional material by Paul Brenton and Takako Ikezuki A central element of any poverty alleviation strat- market access barriers in conjunction with issues egy is to improve market access for exports in such as downstream integration and coordination which the poor have a comparative advantage. This between producers and retailers.3 The value chain chapter analyzes market access barriers to agricul- between the production of commodities by the tural exports produced by the poor in East Asia. poor and the final product that reaches the markets The effects of trade liberalization on the poor are of developed and developing countries is central to the subject of intense debate,1 but there is general improving market access for products of the poor support for freeing market access for products of (see Chapter 9 in this volume for approaches to export interest to the poor. The discussion takes as such improvements). a starting point the premise that increased exports of products of interest to the poor will help to reduce poverty.2 Products and Markets of Export Interest to the Poor This first section of this chapter identifies the key export products and major markets of East Reducing the trade barriers in agriculture is partic- Asian exporters likely to be of interest to the poor, ularly important for reducing poverty in East Asia.4 with a focus on agricultural and agroprocessing Although agriculture is no longer a major source of products. The next section analyzes market access exports and income for many economies of East barriers for those products. A concluding section Asia,5 it is the main livelihood for poor households suggests priorities. in both low-income and middle-income countries Identifying and removing barriers to interna- in the region. According to statistics compiled by tional trade are, however, only part of the process of the World Bank, agriculture represents the main alleviating poverty. Policymakers need to consider source of income for over 90 percent of poor The views in this paper are expressed in a personal capacity and should not in any way be associated with the World Trade Organization, World Bank, or their member states. I am grateful to Zheng Wang for her excellent research assis- tance and to Aki Kuwahara, Daniel Morales, and Jurgen Richtering for extracting and processing the tariff data. The paper also benefited from comments from Florian Alburo and other participants in the seminar held at the Institute for Southeast Asian Studies, Singapore. 59 60 East Asia Integrates households in Cambodia and Vietnam, and over and from improved market access. But the majority two-thirds of poor households in Indonesia, the of the gains, about three-quarters, would come Philippines, and Thailand.6 from productivity improvements as farmers gain Determining the exposure of the poor to inter- unrestricted access to export markets and specialize national markets is a first step in identifying prod- in higher value added crops. China would be one of ucts of export interest to the poor. Research using a the main beneficiaries, and, indeed, China has framework developed by Winters shows that their taken a position of unilaterally opening its markets exposure is largely indirect (see McCulloch, Win- for agriculture, thereby obtaining some of the ters, and Cirera 2001). Poor households tend to potential benefits for itself. But as shown in Chap- produce primary products such as fruits and veg- ter 1, China is likely to become a major importer of etables, fish and fish products, and commodities oilseeds, sugar, beverages, tobacco, plant fibers, and such as jute and coffee. Those in these households other food products in which regional East Asian also work for wages in low-skill jobs such as in tex- economies have strong comparative advantage. tiles, clothing, footwear, and travel goods. In each of Thus some of the gains from agricultural liberaliza- these cases, output is sold to wholesalers, who then tion can be generated through regional arrange- make international transactions. This chapter ments, even if global negotiations do not make sig- focuses on agricultural and agroprocessed exports. nificant progress. Analytical modeling suggests that the gains from agricultural liberalization are reasonably large, as Agricultural and Agroprocessed Exports shown in Figure 4.1. This figure reveals that the East Asia region stands to benefit substantially from Within the East Asia region, there is significant agricultural liberalization in high-income coun- diversity of products through which the poor are tries, and perhaps three times as much from agri- exposed to international markets. Using an cultural liberalization in low- and middle-income approach described in the technical appendix to countries. Full global agricultural liberalization this chapter,8 we examined the structure of exports, would bring benefits to East Asia of almost US$300 which suggests some important priorities despite billion,7 an amount comparable to the gains from this diversity. further liberalization in manufactured goods. Some The agricultural exports of the East Asian devel- of the gains would come from a better allocation of oping countries are quite concentrated.9 The top resources associated with agricultural liberalization four products account for between 30 percent and FIGURE 4.1 Real Income Gains in Developing East Asia from Agricultural Liberalization Real income gains from agriculture Real income gains from agriculture liberalization: high-income country liberalization: low- and middle-income liberalization country liberalization (1997 $US billions, compared with (1997 $US billions, compared with baseline income in 2015) baseline income in 2015) 76 64 213 148 20 80 15 60 10 40 5 20 0 0 East Asia and China Other East Asia East Asia and China Other East Asia Pacific Pacific Simulations with fixed productivity Simulations with fixed productivity Simulations with endogeneous productivity Simulations with endogeneous productivity Source: World Bank (2002: Table 6.1). Market Access Barriers and Poverty in Developing East Asia 61 90 percent of the total agricultural exports of each fruits and vegetables and then other products.10 For country. In Indonesia, for example, four products the poorer developing countries, fruits and vegeta- account for more than half of agricultural exports: bles and coffee appear to be relatively more impor- palm and coconut oil account for 40 percent, and tant.11 Among the agroprocessed products coffee and cocoa beans add another 14 percent. exported by the least-developed countries in East Myanmar represents an extreme case, where fresh Asia, those such as fish and fish products are vegetables account for more than half of agricul- important, as well as leather and rubber products. tural exports and nearly a quarter of total exports. Rice is Myanmar's next most important export, accounting for 20 percent. Similar concentration Markets for the Key Agroproducts effects can be found for Cambodia, Lao People's The major markets for the key agroproducts of the Democratic Republic (Lao PDR), and Vietnam. least-developed countries in East Asia are summa- Table 4.1 identifies the top products for develop- rized in Table 4.2.12 Not surprisingly, the main fea- ing East Asia based on the number of times partic- ture of the results is the concentration of exports to ular types of products appear on the list of top agri- Association of Southeast Asian Nations (ASEAN) cultural exports for the individual countries. The members, China, the European Union (EU), Japan, most frequently listed products are spices plus and the United States. Myanmar is the exception, cereal and other food preparations, followed by with India and Pakistan in its list of destinations. TABLE 4.1 Frequency of HS4 Lines in Top 30 Exports of Developing East Asia by Value Number of MTN category appearances in for agriculture East Asian top Harmonized 30 agricultural MTN system nomenclature exports category Description (HS 1996) 86 15 Spices, cereal, and other 0407­10, 0904­10, 1101­04, food preparations 1107­09, Ch. 19, 2102­06, 2209 83 12 Fruit and vegetables Ch. 07, Ch. 08, 1105­06, 2001­08 76 23 Other agricultural products Ch. 05 (except 0509), 0604, 1209­10, 1212­14, 1802, 230110, 2302­03, 2307­ 09, 290543­45, 3301, 3501­05, 380910, 382460, 4101­03, 4301, 5001­ 03, 5101­03, 5201­03, 5301­02 40 20 Beverages and spirits 2009, 2201­08 37 13 Coffee, tea, maté, cocoa, 0901­03, Ch. 18 (except 1802), and preparations 2101 37 18 Oilseeds, fats and oils, 1201­08, Ch. 15 (except 1504), and their products 2304­06, 382360 29 17 Animals and products thereof Ch. 01, Ch. 02, 1601­02 27 19 Cut flowers, plants, 0601­03, 1211, Ch. 13, Ch. 14 vegetable materials, etc. 21 22 Tobacco Ch. 24 18 14 Sugars and sugar confectionery Ch. 17 15 16 Grains Ch. 10 11 21 Dairy products 0401­06 Notes: Ch. = harmonized system chapter; HS4 = harmonized system to four digits. MTN = Multilateral Trade Negotiations. Source: World Trade Organization. of a Fed. n.e.s. Rep. ASEAN 5th (0.8) EU Lanka (10.4) (2.5) China (8.3) EU (7.7) (6.1) Japan (1.6) (6.5) EU (4.4) (9.1) (4.6) USA (3.8) USA (8.2) Pakistan Sri Areas, Korea, Other Russian of a a total) Rep. n.e.s. of 4th (0.8) EU (3) (11.1) Japan USA (11.6) (10) EU (10.7) (4.5) Japan (8.1) EU Australia ASEAN (11.3) China (11.2) ASEAN (7.1) (6.9) China (12.3) cent Korea, Areas, (per a markets 3rd emenY (1.4) EU EU EU ASEAN (12.3) USA (4.7) India (13.5) ASEAN (11.2) India (12.2) (5.5) Pakistan (13.3) USA (18.3) (19) USA (10) China (11.0) (15.1) export a a a a Fed. Largest China ASEAN ASEAN EU 2nd (6.2) (16.3) (13) ASEAN (20.7) USA (20.5) China (12.8) (13.2) (23.7) China (19.4) Japan (20) China (17.1) Japan (14.6) Japan (19.6) Russia Other Other Other specified. not a a a areas ASEAN ASEAN ASEAN ASEAN Ist EU (90.3) Japan (28.9) China (71.7) (23) China (41.3) (27.2) China, (75.0) India (23.7) Japan (34) USA (29.4) Japan (48.6) (40.6) (20) other Other Other Other Other all = n.e.s. Exports Areas, 2 total 0.04 4.79 6.52 0.34 5.01 14.83 44.38 0.68 3.93 0.92 2.05 8.85 exports/ exports Macao. Agricultural Agricultural (China), 1.3 0.5 1 1 0.004 0.04 1.4 2.2 0.2 0.6 0.2 0.7 2.3 Asian exports/ imports aiwanT Agricultural agricultural East (China), for value Kong (US$) 1,007,143 data. 62,917,847 199,419,463 Markets Export 11,946,201,904 4,046,168,687 4,048,359,304 1,628,656,355 4,918,276,068 1,016,346,249 1,496,655,871 1,591,817,483 2,829,032,672 6,088,003,412 Hong Key China, Comtrade Kong 4.2 of UN: ce Includes Reporter Brunei Darussalam China Hong (China) Indonesia Japan Malaysia Mongolia Myanmar aiwanT (China) Philippines Korea, Rep. Singapore Thailand a. Sour ABLET 62 Market Access Barriers and Poverty in Developing East Asia 63 Regional agricultural markets are large. Notably, sures where possible. This is not to suggest that China (with Hong Kong and Taiwan) is the first or support payments in importing countries are second largest market for all economies with the unimportant to developing country exporters, but exception of Brunei, Singapore (where China is rather that these payments should be considered in third largest), Indonesia (fifth largest), and the relation to the export potential of specific coun- Philippines (fourth largest). ASEAN is also a large tries. According to Hoekman, Ng, and Olarreaga and important market, with the rest of ASEAN (2002), domestic subsidy payments are concen- being the first or second largest market for all the trated in a few sectors such as meat, dairy products, ASEAN economies. For the northern Asian and cereals (Table 4.3). economies of Japan, Republic of Korea, and China, Most products important to the poor of East ASEAN is the third or fourth largest market. This Asia, such as fruits, vegetables, and spices, together implies that many of the mutual gains from agri- account for only 7 percent of domestic support cultural liberalization depend on the policies of payments. The major exception is rice. Within cere- countries in the East Asia region. als, rice accounts for nearly half of all direct sup- There also is a clear distinction between markets port commitments, totaling $46 billion driven by in which preferential access through the ASEAN the $36 billion attributable to Japan (see Box 4.1). Free Trade Agreement (AFTA) is important, and Because of the dominance of East Asia in rice pro- those in which most-favored-nation (MFN) rates duction and trade, many of the mutual gains from apply. Some countries have nonreciprocal prefer- agricultural liberalization depend on the policies of ences through the Generalized System of Prefer- countries in the East Asia region. Coordinated ences (GSP). Cambodia, as a least-developed coun- regional action could contribute to an increase in try, is eligible for a specific set of preferences. the share of rice traded and thereby to an increase in the reliability and stability of rice markets, which have been at the core of concerns of East Asian pol- Market Access Barriers icymakers. This section examines the implications of some of the market access policies of importing countries in Tariff Barriers sectors of central interest to East Asian developing countries. In agriculture, governments use a variety Because several different types of measures are of policies beyond tariffs that not only protect their known collectively as tariffs, it is useful to clarify markets but also distort them so that efficient pro- the measures analyzed here. First, bound and ducers are not the main suppliers of products applied MFN tariffs differ. Bound tariff rates, which (ABARE 2000). are negotiated in the World Trade Organization (WTO), are the ceilings that apply to tariffs. Applied tariffs are the actual duties applied at cus- Domestic Support toms on an MFN basis.14 They can be raised in a Hoekman, Ng, and Olarreaga (2002) make an manner consistent with a country's WTO obliga- important point that guides the analysis here. tions so long as the increase is not above the bound Comparing the impact of a 50 percent cut in rate. This chapter looks at the applied rates. Second, domestic support by an importing, not exporting, it is important to identify the prevalence of specific country with a 50 percent cut in tariffs, they find tariffs as well as ad valorem duties. Data presented that developing countries as a group would receive in this section show the prevalence of specific tariffs a larger benefit from the cut in tariffs.13 For one on the exports of developing Asian countries. thing, domestic support payments are concentrated Table 4.4 provides an overview of the simple in few products in which developing countries average of tariff rates applied on products of interest directly compete, and tariff cuts have a more direct to the poor in East Asian countries.15 These figures effect on world market prices than do cuts in have been calculated using only ad valorem lines. domestic support programs. This section therefore Many countries do not report ad valorem equiva- focuses on the different types of border measures lents for specific tariffs or tariff-quota schemes, so and the identification of relevant nontariff mea- these figures may largely reflect the barriers posed 64 East Asia Integrates TABLE 4.3 Commitments and Average Direct Domestic Support Levels, 1995­98 Direct support As % (millions of US$) of total HS2 Product Commitment 1995­98 Commitment 1995­98 01 Live animals 250 63 0.1 0.1 02 Meat and edible meat offal 60,155 14,907 22.3 18.5 04 Dairy products; birds' eggs; honey 39,372 11,557 14.6 14.3 06 Live trees and other plants; bulbs, cut flowers 0 14 0.0 0.0 07 Edible vegetables and roots and tubers 10,326 3,975 3.8 4.9 08 Edible fruit and nuts; melons 7,879 3,474 2.9 4.3 09 Coffee, tea, maté, and spices 1272 50 0.5 0.1 10 Cereals 104,109 27,953 38.5 34.6 11 Milled products; malt; starches 421 142 0.2 0.2 12 Oilseed, oleaginous fruits 8,577 447 3.2 0.6 13 Lac; gums, resins, other vegetables 0 0 0.0 0.0 15 Animal/vegetable fats and oils and products 1,899 1,050 0.7 1.3 17 Sugars and sugar confectionery 12,370 5,304 4.6 6.6 18 Cocoa and cocoa preparations 16 0 0.0 0.0 20 Prep. of vegetable, fruit, nuts products 892 529 0.3 0.7 21 Miscellaneous edible preparations 0 0 0.0 0.0 22 Beverages, spirits, and vinegar 4,306 1,172 1.6 1.5 23 Residues and waste from food industry 382 192 0.1 0.2 24 Tobacco and manufactured tobacco products 2,662 735 1.0 0.9 50 Silk 416 14 0.2 0.0 51 Wool, fine/coarse animal hair nest 124 17 0.0 0.0 52 Cotton 3,411 655 1.3 0.8 53 Other vegetable textile fibers and yarns 34 71 0.0 0.1 98 Nonproduct-specific 11,276 8,392 4.2 10.4 Total agricultural products above 270,151 80,714 100.0 100.0 Notes: Direct domestic support is defined as the sum of World Trade Organization DS4 to DS9 categories. HS2 = harmonized system to two digits. Source: Hoekman, Ng, and Olarreaga (2002). Based on WTO Secretariat (2000). by ad valorem tariffs only.16 The table does show age tariff rates than developed countries. Second, reasonably high numbers in the key agricultural cat- industrial countries apply above-average tariffs on egories (categories 12­23)--in particular, fruits and agricultural products of particular interest to devel- vegetables (category 12); coffee, tea, and cocoa (cat- oping countries. These products tend to face tariffs egory 13); and cut flowers (category 19). Average that are double the average rates in the United tariffs are also high on agroprocessing products of States, Japan, and the EU. traditional export interest to the poor: leather, rub- Specific tariffs are quite prevalent. Table 4.5 ber, footwear, and travel goods (category 3) and fish summarizes the structure of the trade policies of and fish products (category 11).17 The data here are the EU, Japan, and United States affecting market based on MFN trade barriers and do not take into access for East Asia as a whole.18 The data here are account the preferences that East Asian countries based on MFN trade barriers and do not take into are granted in these markets. account the preferences that East Asian countries This table confirms two general points about the are granted in these markets. Yet they suggest that, landscape of protection. First, developing countries for individual countries, specific tariffs can affect a in general impose considerably higher overall aver- very large proportion of agricultural exports, Market Access Barriers and Poverty in Developing East Asia 65 BOX 4.1 Global Trade in Rice Over 90 percent of rice production and con- Taiwan (China). Protection in other countries of sumption occurs in Asia. Much of Asian rice is East Asia is more moderate. The greatest degree subject to monsoon climates, resulting in of protection is in medium and short grain rice, uncertain rice yields and rice supplies. Only 6.8 and thus very few rice-exporting countries pro- percent of global rice production is traded, a duce this type. Most global rice trade is in long smaller share than other major grains and grain rice, which is characterized by tariff escala- oilseeds. The price variability of rice is also the tion, notably in the European Union and Central largest of all food crops, with a coefficient of and South America. This pattern of protection variation of 47 compared with 34­35 for wheat depresses world prices for high-quality, milled and maize. However, its price variability is long grain rice and discriminates against the much less than that of other commodities such milling sectors of exporting nations such as Thai- as sugar and oil with a coefficient of variation land and Vietnam. over 70. Protectionist policies have contributed to the The major types of distortion in world rice thinner global markets by encouraging greater markets are import tariffs and tariff-rate quotas reliance on domestic production for domestic (TRQs) in key importing countries and price sup- consumption. At the same time, the variability in ports in key exporting countries. In 2000 the global markets has been a motivation for heavier global trade weighted tariff on all rice was 43.3 protection. If collective action were taken in East percent. Global trade weighted average rice tar- Asia to address transition and vulnerability con- iffs for japonica rice markets were much higher cerns, it is estimated that trade liberalization than those on indica rice--217 percent com- would result in a 73 percent increase in global pared with 21 percent--because of the TRQ and trade of short and medium grain rice, which in quotas in the major japonica rice­importing turn would contribute to the stability and relia- economies of Japan, the Republic of Korea, and bility of global markets. notably to the EU and the United States. These tar- mon in Japan, affecting only about 10 percent of iffs contribute to a lack of transparency about the agricultural exports from East Asia. magnitude of the overall extent of protection. The incidence of specific duties likely falls partic- In agriculture, more than 94 percent of East ularly heavily on the least-developed countries Asian countries' exports to Japan are subject to exporting relatively low-value products, where the duties at the border (ad valorem or a specific tariff). computed ad valorem equivalents of specific duties Almost three-quarters of their agricultural exports are subsequently often high.20 The reason is that the to the EU and just over 46 percent of those to the methodology for conversion depends on the per United States are subject to duties. Specific tariffs unit value. In many cases, the per unit value of are prevalent in the United States, affecting more products from the poor is lower than those from than one-quarter of all East Asian exports of agri- higher-income producers, resulting in a high ad val- cultural goods and more than half of the East Asian orem rate. Thus paradoxically, low-quality goods, agricultural exports that are subject to duties. In the and cheap goods from low-income countries, are EU, specific tariffs affect about one-quarter of total subjected to higher rates of border taxation for their agricultural exports from East Asia and more than exports to Europe and the United States. one-third of those that are subject to MFN duties. For example, almost two-thirds of Thailand's agri- Tariff Escalation cultural exports to the EU are subject to specific duties, and 45 percent of total agricultural exports High tariffs are only part of the market access pic- from the Philippines to the United States are subject ture for developing countries. Tariff escalation, to specific duties. Within the East Asia region, how- whereby tariffs imposed on processed products are ever, specific duties are very rarely used as a means relatively higher than those on raw materials, of trade protection.19 Specific duties are less com- reduces the incentive for producer countries to USA 0.8 9.4 4.4 2.2 3.4 3.2 1.2 1.9 1.9 2.1 1.1 7.8 2.6 6.2 3.1 2.2 3.4 9.1 cent) 1.5 6.1 2.3 3.0 5.5 1.4 2.3 1.7 2.8 1.1 2.7 1.4 4.0 3.7 4.4 3.1 (per 12.2 11.5 Canada EU 2.2 8.5 4.2 2.5 4.9 4.1 1.7 2.5 2.0 2.6 9.8 5.8 5.0 5.4 5.3 3.2 11.2 11.4 Partners 9.2 13.8 25.4 26.3 12.5 10.4 23.6 13.2 10.0 15.0 57.6 58.9 06 46.3 42.5 50.3 82 radingT Thailand 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 Singa pore Major of and 5.9 8.0 6.1 7.3 5.5 6.4 6.1 5.9 6.8 10.1 16.2 55.6 55.3 20.1 24.7 14.3 Korea, Rep. 111.8 192.5 8.9 7.7 5.9 4.2 8.2 3.5 5.0 5.5 5.5 9.0 9 6.3 Countries Philip- pines 12.6 10.4 18.9 18.9 18.5 27.1 Asian 6.5 5.3 3.0 2.3 3.9 1.6 4.2 4.3 6.4 8.1 mar 10.6 13.1 41 5.6 8.1 0.9 1.7 11.4 Myan- sia 9.3 3.6 3.7 6.7 8.8 5.1 2.4 2.9 9 2.8 2.6 0 0.5 1.7 Selected,y 10.9 13.5 14.0 18.5 Malay- a 1.3 7.6 6.7 1.4 2.6 0.0 0.0 0.2 0.8 1.1 5.9 8.4 1 7.8 2.1 Japan 11.6 10.1 12.5 [23.8] Categor 7.9 8.5 6.6 2.3 7.7 6.0 5.0 5 4.9 3.8 5.2 2 4.6 4 Indo- nesia 14.0 10.7 12.1 10.3 MTN by 4.6 9.4 5.9 6.0 3.7 4.9 5.3 4.2 4.8 2.8 8.2 aiwan,T 11.6 27.1 28.7 13.8 27.3 20.2 26.3 (China) Rates HK 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 ariffT (China) 9.8 China 14.1 26.8 17.7 11.4 23.3 14.4 16.1 12.1 18.0 21.5 22.6 26.1 27.9 31.4 54.4 20.7 31.1 Applied 4.1 0.6 3.0 0.0 0.4 6.3 0.5 5.0 0.0 0 0.9 0 0 0 0 0 Brunei 13.7 14.2 MFN , y , and oils verage A footwear, y and supplies articles specified and paper clothing machiner products maté, products and goods and preparations sugar y equipment vegetables preparations fats products pulp, and rubber, machiner products stones metals fish tea, and and cereal, and Simple and food Description furniture travel elsewhere and their ood, 4.4 W and extilesT Leather and Metals Chemicals photographic ransportT Nonelectric Electric Mineral precious precious Manufactured not Fish Fruits Coffee, cocoa, Sugars confectioner Spices, other Grains Animals thereof Oilseeds, and y 1 2 3 4 5 6 7 8 9 MTN Cate- gor 10 11 12 13 14 15 16 17 18 ABLET 66 1.2 1.8 0.8 13.5 204.2 0.7 4.4 7.4 7.3 0.8 2.4 7.7 1.3 11.3 39.7 0 38.5 35.8 29.1 0 0 0 0 0 (2001). tawowyS 28.1 29.1 72.2 33.2 10.1 using 3.2 5 8.4 3.2 10.7 calculated 4.5 3.3 3.1 24.2 25 equivalents 0 9.4 3.6 0.7 Organization. valorem 1.3 14.7 62 4.2 1.1 radeT ad orld with 5.7 5 4.4 W 80 10.7 value 9.2 3.7 28.1 18.6 25.4 indicate Negotiations, 0 0 0 0 0 radeT Brackets rice. 12.4 50.6 40.3 56.7 12.3 on Multilateral = high 0 0 0 0 0.1 are MTN which nation; products spirits quotas, plants, materials; and most-favored agricultural tariff-rate IDB. flowers, etc. products = y Cut vegetable lacs, Beverages Dair obaccoT Other MFN WTO: ce Excluding 19 20 21 22 23 Notes: a. Sour 67 68 East Asia Integrates TABLE 4.5 Trade Policy Structure of European Union, Japan, and United States for Imports from East Asian Countries (percent) Share of Share of products Share of Share of dutiable imports subject to specific duty- free dutiable subject to duties in imports imports specific tariffs total imports Agriculture EU 25.65 74.35 33.26 24.72 Japan 5.81 94.19 10.88 10.25 United States 53.32 46.68 56.30 26.28 Manufactures EU 34.21 65.79 0.75 0.49 Japan 59.59 40.41 10.56 4.27 United States 46.14 53.86 13.31 7.17 Note: For details, see the appendix tables to this chapter at www.worldbank.org/eaptrade. Source: UNCTAD TRAINS database. process products before exporting them. This has Thus East Asian exporters appear to face signifi- been a contentious issue in trade negotiations. cant tariff escalation in their main overseas mar- To put the tariff escalation issue in perspective, kets. Within East Asia, however, it is clear that as the consider coffee in Vietnam. By processing its coffee, ASEAN Free Trade Area is fully implemented, the Vietnam could capture a higher proportion of total problem of high tariffs on regional trade and tariff value added in the production chain. But it is diffi- escalation among ASEAN countries will be consid- cult to perceive how Vietnamese producers, despite erably reduced. their phenomenal success in coffee production, would be able to integrate downstream.As indicated Preferential Rates earlier in this chapter, the world market for coffee is highly concentrated, with processors exercising Although this chapter focuses on multilateral mar- monopsonistic powers. Simply lowering an import ket access, nonreciprocal preferential arrangements tariff on processed coffee would ease market access, are discussed briefly in this section because for cer- but it would not be sufficient to induce investment tain exports from certain countries these prefer- that ensures competitive exports of processed coffee ences may make the multilateral market access bar- from the coffee-producing country. riers less relevant.23 However, evidence suggests For this study, tariff escalation was analyzed in that the limited coverage and rules of origin in var- five products of interest to developing countries: ious schemes limit their impact on barriers to coffee and tea, tomatoes, dried vegetables, crus- products important to the poor of East Asia. taceans, and leather products (Table 4.6). For cof- Because the U.S. Generalized System of Prefer- fee, Canada does not discriminate between ences (GSP) scheme effectively excludes the agri- unprocessed and processed products, whereas both cultural exports of most poor East Asian countries, the EU and the United States do to quite a large most of their exports enter the U.S. market at MFN degree, with a difference in duties of about 10 per- rates. As for the EU, the least-developed countries' cent. This difference translates into much higher access for agricultural products is formally more effective protection of coffee processing. The MFN open, because under the Everything but Arms duties applied by developing Asian countries on (EBA) agreement these countries have duty- and coffee are much higher and differ significantly quota-free access to the EU market. Yet their access between processed and unprocessed products.21 to the EU market does appear to be constrained by For most of the other four products, the pattern is restrictive rules of origin. Agricultural products, similar.22 with some exceptions, are covered by the EU's EBA (5) (5) iet-V (50 (5) (5) (5) nam 50 50 5 50 25 30 30 20 30 50 component (5) (5) (5) (5) (5) (5) (5) (5) 5 (5) Thai- land 40 47 60 30 60 60 60 40 40 valorem ad the a aiwanT 5.3 22.5 10 24 12 20 36.5 30.0 2.5 5 (China) . (5) (5) (5) (4) example, (4) (3) (5) (5) (3) (5) system Philip- pines 51 35 10 4 11.6 10 5 4 10.8 15 For commitments) A asonal. (4) 0 4 (3) (5) (2) se (AFT Myan mar 15 51 15 15 5 10 7.5 2 harmonized made. = also Area HS been are (2) (0) (0) (0) radeT countries sia 0 0 0 5 (2.5) (2.5) 0 Area; has Malay- 11.6 5.8 8 20 30 duties radeT The Free ASEAN (0) (0) (0) (0) (0) Lao PDR 20 30 (0) 20 40 40 40(0) 5 10 10 10 Free commitment prices. ASEAN no the (0) (5) (0) (5) (0) (5) ASEAN (0) (0) (5) = that and Indo- nesia 5 5 5(0) 5 5 5 5.0 16 0 15 A lower AFT the indicate season. 30 35 7 35 7 35 15 35 35 24 Markets Cam- bodia the Nations; increases on Asian 25 50 13 25 7.2 25 parentheses that China 30.2 25 8.7 25 Selected No duty in c depending and Southeast USA 0.2 12.2 11.9 1.2 6.9 7.5 7.1 3.4 10 of specific seasonal a Specific parentheses. kilogram, Products (MFN) in and per are Japan 12.0 18.7 0.0 12.6 11.2 15.9 3.8 5.9 42 7.2 Association = cents countries y b Selected 3.9 EU 0 component 9.5 Entr 3.2 15 7.7 11.07 19.04 0­5.5 ASEAN members to for prices 14.4 A Quad 2.8 AFT valorem da 0.0 0 0.0 2.5 6.0 5 nation; by cent. Cana- 11.5 4.35 3.3 18 from ad per an made Escalation 14.4 ranges (0901) (0702) (0713) (0306) (4104) to ariffT y code) (2101) (2002) (1106) (1605) (6403) most-favored duty = comprises 8.8 IDB. digit MFN commitments tariff from specific WTO 4.6 categor ce: Unprocessed Processed Unprocessed Processed Unprocessed Processed Unprocessed Processed Unprocessed Processed The The HS6 (four- Agriculture Coffee omatoesT egetablesV The Nonagricultural products Crustaceans Leather Notes: a. b. ranges c. Sour ABLET 69 70 East Asia Integrates scheme, but the reduction on specific duties is only and safety regulations, notably maximum pesticide 30 percent. Furthermore, when the MFN duty has residue levels; difficulties with understanding and both an ad valorem and a specific component, the administering standards; lack of technical assis- specific component is not reduced. tance; and, in some cases, seasonal tariffs. For The GSP is highly relevant to the exports of the example, health and safety standards are a major least-developed countries and Vietnam to the EU. barrier for fresh fruit, vegetables, fish, sweet bis- Almost 100 percent of Cambodia's EU exports, 93 cuits, and other foodstuffs and drinks. Organic percent of the Lao PDR's EU exports, and 83 per- products face high costs in complying with EU cent of Vietnam's EU exports are eligible for prefer- standards and definitions and problems of certifi- ential access to the EU. For other East Asian coun- cation. tries, the GSP is less important, and poor households in these countries do not benefit. For Conclusions example, less than 30 percent of China's exports to the EU are eligible for preferences, and only 15 per- Agriculture is the predominant source of income cent of Malaysia's exports are eligible.24 for poor households in East Asia. In products of An important feature of many nonreciprocal special importance to these poor households-- preference schemes is their low level of utilization agricultural and agroprocessing products--impor- for products actually exported by developing coun- tant barriers remain in developed and developing tries. For example, although almost all of Cambo- country markets. dia's exports to the EU were eligible for preferences, In contrast to the move toward increased trans- Cambodia took advantage of these preferences for parency and reduced protection in trade policy for only 36 percent of those exports.25 For the other manufactures, agriculture and agroprocessing East Asian developing countries, the take-up of remain more heavily protected, distorting global preferences never exceeds two-thirds of the value of markets and affecting production patterns and the exports eligible for preferences. A major reason is efficient adoption of technology in poor countries. the prevalence of stringent rules of origin. Poor For this reason, agricultural liberalization offers countries, with underdeveloped institutions, find it great potential for welfare gains. And because poor hard to comply with the complex bureaucratic households are heavily engaged in agricultural requirements needed to prove the origin of their activity, they would benefit most from greater mar- exports and so take advantage of tariff reductions. ket access in agricultural products. For East Asia, liberalization in both developed and developing country markets is important, the latter even more Nontariff Measures so than the former. And there is greater potential Data on nontariff measures (NTMs) are very diffi- for realizing these gains through regional arrange- cult to collect, but they remain an important bar- ments, especially between ASEAN and China, given rier in agroproducts. Unfortunately, despite the the political complexity surrounding multilateral important role that nontariff measures play in the agricultural liberalization and given the existing market access of agricultural products, very little marketing links within the region in agricultural work has been done to quantify their impact. One products. possible methodology for assessing the extent of Fortunately, East Asian farmers do not partici- NTMs is the use of frequency data. Such data are pate much in those markets most distorted by the available, but they are limited in their ability to domestic support programs used in high-income quantify the trade distortive element of NTMs (see countries, with the notable exception of rice. But Bora 2002). Qualitative data are more helpful, but each major market has significant barriers, which those available are not comprehensive. Case studies vary according to the product and market. First, the are the best way to deal with these kinds of issues. EU and the United States levy above-average ad va- The qualitative data that are available, such as lorem tariffs on agricultural products and use spe- the analysis by Cerrex (2002) reproduced as Table cific tariffs to a large extent, thereby reducing trans- 4.7, show that for the products of concern to the parency and creating higher effective protection poor countries of East Asia, the key issues are health against low-value goods, often from poorer coun- Market Access Barriers and Poverty in Developing East Asia 71 TABLE 4.7 Nontariff Measures Affecting Products of Interest to Developing East Asia Economies Product HS number Particular problems Fresh fruit Health/safety: pesticide residue levels; difficulties in complying with EU standards (i.e., expense, lack of technical knowledge, difficulty with interpretation of directives). Common Agriculture Policy (CAP) levies; quota/license requirements; seasonal preferences; continuing high residual tariffs. Generally no tariff preference for GSP suppliers. Processed fruit/ Except 20.07 Same as above, especially high residual tariffs. No fruit-vegetable juices to 20.09 concession for GSP. Jams and jellies Except 20.07 Preference given, although residual duty remains very high. Natural honey 04.09 No preference for GSP; high residual duties. Rice 10.06 Quotas and licensing and levies: very complicated tariff structure. Vegetables Health/safety regulations: maximum pesticide residue levels; difficulties with understanding and administering standards; lack of technical assistance. Seasonal preferences for some products; no preference generally for tomatoes. Organic products Misc. High costs of complying with EU standards and definitions; problems with certification; lack of subsidies from national governments; costs of meeting phytosani- tary standards. Mixture of tariff preferences (e.g., no GSP preferences for chicken or pork); minimal GSP preference for sausages. Walnuts Except 08.02 Tariff quotas may apply. Fish Except Ch. 3 High residual tariffs often granting GSPs no preference; and 16 problems of origin rules; cost of meeting and administering health regulations; difficulties and cost of testing procedures. Chocolates Origin rules and sugar levies. Sweet biscuits 19.05 Difficulties meeting health/phytosanitary standards; sugar levies; quotas. Fresh flowers Orchids and 06.03 Difficulties in meeting environment and social rose stems standards; CAP levies and duties; quota/license requirements. Wine Except 22.04 Subject to continuing high duties; specific excise duties; removal of GSP preference for certain countries. Cocoa Except 18 A mixture of the effect of sugar levies. Origin rules for chocolate, quotas, and the number of preferences available (up to 28) make this the most complicated chapter in the tariff. Foodstuffs and drinks, electrical and engineering goods Passim Standards. Wheat gluten 11.09 Quotas may apply; no preference for GSP countries. Refined palm oil 15.11 CAP levies and duties may apply; quota restrictions. Frozen vegetables Chs. 7 and 20 Need for financial assistance for freezing capacity. Essential oils and 33.01 and 33.03 Operation of excise duties to the disadvantage of perfumes cheaper products. Notes: This list is not meant to be a fully comprehensive one of the products or of the problems. The information given is as of mid-2001. HS = harmonized system; GSP = Generalized System of Preferences; Ch. = harmonized system of chapter. Source: Cerrex (2002). 72 East Asia Integrates tries. Second, domestic support and tariff-quota to address barriers specifically for products of schemes remain an important element of protec- interest to developing countries. Such steps would tion in Japan, notably for rice. Third, tariff escala- include reducing tariff barriers, converting non­ad tion on agroprocessed goods in the EU, North valorem lines to ad valorem rates, reducing domes- American, and Japanese markets results in high tic support, and disciplining export subsidies. The effective protection rates and discourages diversifi- type of result that emerges from the negotiations, cation into processing activities as well as the cap- however, depends on the effective participation of ture of more value added by producing countries. developing countries in the negotiations. Fourth, standards are increasingly a major barrier, Yet another possible avenue for pursuing with East Asian exporters facing difficulties in enhanced market access is through the ASEAN Free meeting health and safety standards, for example. Trade Area process and other regional arrange- The results of the analysis in this chapter suggest ments. The current implementation schedule has that reducing trade barriers to agriculture-based resulted in fewer barriers in the more advanced products important to the poor of East Asia should ASEAN members. This is an important step be a central trade policy objective. Because no sin- because evidence suggests that many products are gle trade protection element dominates in barriers produced on a regional basis. A more liberal to these products--and because each trade protec- regional trading landscape will encourage further tion element is important in at least one major processing and specialization in a regional context, market--the approach toward trade policy needs to which could enhance opportunities for exporting be broad. Thus the targets for this trade policy the types of products discussed here. should cover all of the following: Endnotes · Reducing the above-average tariff rates · Reducing the extent of tariff escalation and pro- 1. See Rodriguez and Rodrik (1999) for a general discussion of these issues, as well as Rodrik (2001). viding additional routes for product diversifica- 2. Much has also been written about the need for developing tion and export growth in developing countries countries to move from labor-intensive and low-technol- · Reexamining health and safety regulations and ogy exports into higher technology and higher value added procedures for administering them products. Indeed, such a transition is crucial from a general development perspective. But because agricultural prod- · Improving the transparency of tariff schedules ucts and labor-intensive products remain of key interest to by converting non­ad valorem rates to ad va- poor households, this chapter focuses on these products. lorem rates. 3. For example, consider coffee in Vietnam. Since 1990 Viet- nam's coffee production has increased more than tenfold-- mainly from small farms. Farmers sell their coffee cherries Deciding on the arena in which to pursue these to wholesalers, but all they can do is try to ensure that their objectives is another issue. The chapter has stressed crop meets minimum quality standards. Four companies purchase half of the world's coffee production, and fewer that many of the poorer countries have adopted than 20 percent of the world's population drinks 65 percent nonreciprocal market access as a central compo- of the coffee. Traders who mediate between farmers and the nent of their export strategy. However, these major buyers are responsible for the rest of the production schemes remain beyond the scope of multilateral process. Some coffee retailers buy some supplies directly from farmers, but the coordination costs of dealing with negotiations for reasons that are beyond the scope millions of farmers are high. The U.S. retail chain Star- of this chapter. Nevertheless, an opportunity for bucks, for example, buys only about 10 percent of its total improvements in market access conditions clearly coffee purchases directly from farmers. This discussion is based on Stern (2002). exists through further clarification and simplifica- 4. See Bacchetta and Bora (2002) and Bora (2002) for a gen- tion of rule of origin procedures, clearer and more eral discussion of this topic. streamlined applications of standards, and, in the 5. Agriculture's share of official exports is only significant for Myanmar, as shown in Table 4.2 (44 percent). Agriculture's case of some developed markets, a broadening of share of GDP exceeds 25 percent in only a few countries the coverage of such schemes. such as Mongolia, Lao People's Democratic Republic (Lao An opportunity also exists to address some of PDR), and Cambodia. the market access barriers discussed in this chapter 6. World Bank, Poverty Projections Toolkit, 2002 estimate. 7. All dollar amounts are 1997 U.S. dollars. through the current round of multilateral negotia- 8. The technical appendix to this chapter can be found at tions. Negotiators have been given a clear mandate www.worldbank.org/eaptrade. Market Access Barriers and Poverty in Developing East Asia 73 9. The top 30 agricultural exports of each least-developed intermediary step toward multilateral liberalization. In country in East Asia were identified for this study. Lists are industrial markets, the issue is whether preferences remove available from the author upon request. or limit the impact of the above-average tariffs and the 10. It should be noted that the frequency count could overstate extensive use of specific tariffs on products of particular the importance of a particular category. For example, relevance to the poorest countries. countries such as Malaysia have fairly diverse exports, but 24. Further details of preferential access arrangements for poor export value is heavily concentrated in a few products. East Asian countries to the EU and the United States are 11. See last column of Appendix Table 4.1 of this chapter at given in the technical appendix to this chapter at www.worldbank.org/eaptrade. www.worldbank.org/eaptrade. 12. The major markets were identified for each of the key four- 25. Brenton (2003) shows that, even with duty-free access on digit products on the basis of aggregated overall agricul- paper, Cambodia faced an average tariff equivalent to 7.7 tural exports. ASEAN served as one group, given the simi- percent on its exports to the EU. larity of members' protection profiles in the ASEAN Free Trade Area. In that way, further attention could be focused on non­ASEAN markets such as Canada, Japan, the Euro- References pean Union, and the United States. 13. The reasons for this are the high incidence of domestic sup- The word processed describes informally reproduced works that port for products that are of relatively low importance to may not be commonly available through libraries. developing countries and the fact that tariffs affect world prices more strongly than do subsidies. This is not to say ABARE. 2000. The Impact of Agricultural Trade Liberalization on that support payments in importing countries are unim- Developing Countries. Canberra: Australian Bureau of Agri- portant to developing country exporters, but rather that cultural and Resource Economics. one needs to consider these payments in relation to the Bacchetta, M., and B. Bora. 2002. "Industrial Tariffs and the exports of specific countries. Doha Development Agenda." World Trade Organization, 14. A distinction can also be drawn between the applied rate of Geneva. Processed. duty as in a nation's tariff schedule and the rate of duty that Bora, Bijit. 2002. "LDC Market Access Issues and the Doha is applied at the customs point. Development Agenda." World Trade Organization, Geneva. 15. See the technical appendix to this chapter at www.world Processed. bank.org/eaptrade. Brenton, Paul. 2003."Integrating the Least Developed Countries 16. See the technical appendix to this chapter at www.world into the World Trading System: The Current Impact of EU bank.org/eaptrade. Preferences under Everything But Arms." Policy Research 17. China's tariff profile does not take into account its final Working Paper, World Bank and forthcoming in Journal of WTO accession commitments, which have yet to be fully World Trade. implemented. Cerrex. 2002. "Making EU Trade Agreements Work." Study pre- 18. For details, see the appendix tables 4.4 and 4.5 to this chap- pared for the UK Department for International Develop- ter at www.worldbank.org/eaptrade. ment. Processed. 19. See Appendix Table 4.3 to this chapter at www.world Hoekman, Bernard, F. Ng, and M. Olarreaga. 2002. "Reducing bank.org/eaptrade. Agricultural Tariffs versus Domestic Support: What's More 20. For example, in the EU market in 2001 the specific duty of Important for Developing Countries." World Bank, Wash- 128 euros per 1,000 kilograms on imports of broken rice ington, D.C. Processed. translated into an ad valorem tariff of 24.75 percent for McCulloch, Neil, L. Alan Winters, and Xavier Cirera. 2001. Trade imports from China and 28.92 percent for imports from Liberalization and Poverty: A Handbook. London: Centre for Vietnam. In the U.S. market, specific tariffs on certain nuts Economic Policy Research. exported from East Asian countries can translate into ad Rodriguez, Francisco, and Dani Rodrik. 1999. "Trade Policy and valorem equivalents in excess of 12 percent. Economic Growth: A Skeptic's Guide to the Cross-National 21. Lao PDR is an exception, imposing a lower duty on Evidence." NBER Working Paper No. 7081. National Bureau processed coffee than on unprocessed. of Economic Research, Cambridge, Mass. 22. In certain cases, the tariffs on processed agricultural prod- Rodrik, Dani. 2001. "The Global Governance of Trade: As If ucts can be very complex. For example, in the EU processed Development Really Mattered." United Nations Develop- food products such as biscuits are subject not only to a tar- ment Programme, New York, October. Processed. iff on the product itself (of about 9 percent) but also to spe- Stawowy, V. 2001."A Method for Calculating Ad Valorem Equiv- cific tariffs on both the milk and sugar content of the prod- alents." United Nations Conference on Trade and Develop- uct. Such tariffs can represent very high levels of ment (UNCTAD), Geneva. Processed. protection. Stern, Nicholas. 2002. "Crisis in a Coffee Cup." Fortune. Decem- 23. For developing Asian countries, two important issues sur- ber 9. round preferential rates: the nonreciprocal preferences they World Bank. 2002. Global Economic Prospects and Developing are granted in the industrial country markets and the recip- Countries: Making Trade Work for Poor. Washington, D.C. rocal granting of preferences in the context of regional WTO (World Trade Organization) Secretariat. 2000. Domestic integration. As discussed elsewhere in this volume, an Support. WTO document G/AG/NG/S/1, Secretariat Back- important question for these countries is whether ground Paper 13. World Trade Organization, Geneva, improved market access at the regional level could be an April. Part II Development Orientation for A behind-the- border agenda 5 Trade and Logistics: An East Asian Perspective Robin Carruthers Jitendra N. Bajpai David Hummels Why focus on logistics? The reason is simple. the arguments presented in this chapter show that Reducing the cost and improving the quality of ports are only one aspect of the connection logistics and transport systems improve interna- between logistics and trade growth. A look at the tional market access and lead directly to increased total cost of getting products from producers to trade--and through this to higher incomes and a markets reveals that land transport to ports wider scope for significant reductions in poverty. accounts for a higher proportion of the cost than East Asia's progress on logistics has failed to keep processing within the port or the maritime voy- pace with its growth in trade. Developing countries age itself, and improvements in land access offer in other regions are now catching up, so faster the greatest scope for increasing trade competi- progress on logistics development will be crucial to tiveness. sustaining East Asia's competitive advantages. High Countries that have moved beyond exporting logistics costs for East Asian countries stem from basic agricultural and mining commodities find poor transport infrastructure, underdeveloped that logistics requirements have become more transport and logistics services, and slow and costly onerous, not less onerous. Manufacturing firms, bureaucratic procedures for dealing with both especially those integrated into global production exported and imported goods. The balance among chains, seek not only low transport costs but also these three factors varies among countries, but in fulfillment of a host of sophisticated logistical each country a complementary approach to needs: short transit times, reliable delivery sched- addressing all of them will be needed to produce a ules, careful handling of goods in cold storage sustainable improvement in competitiveness. chains, certification of product quality, and security Recent studies have indicated the importance from theft. Basic transport infrastructure does not of efficient ports (in terms of both operational meet the logistics needs of manufacturing firms. efficiency and document facilitation) for trade Thus the requisite policy agenda extends broadly to competitiveness (Wilson and others 2002), but stimulating the evolution of transport services, The authors are grateful to Dr. Jose Tongzan for his comments at the seminar at the Institute for Southeast Asia Studies, Singapore. 77 78 East Asia Integrates promulgating product standards, licensing im- found that countries closer to world markets enjoy ports, and encouraging foreign investment. higher levels of trade, and that a 1 percent rise in This chapter reviews the logistics issues facing the ratio of trade to gross domestic product (GDP) East Asia and proposes a policy agenda to address increases income per person by at least 0.5 percent. them. We first briefly outline the macroeconomic Redding and Venables (2002) estimate that more connections between logistics and trade, and, than 70 percent of the variation in per capita because the economies of the region differ funda- income across countries can be explained by the mentally in their levels of development, the extent of geography of market and supplier access. Better openness and composition of trade. We then discuss access to coasts alone raises incomes by 20 percent. the logistics needs specific to each of three broad As for income differences within countries, groupings of economies. This discussion is followed internal and effectively landlocked regions have by a description of the benefits that flow from systematically lower levels of income than coastal improvements in logistics, and the channels through regions. Comparing China's regions, Wei and Yi which these benefits are produced. We then review (2001) show that trade levels, trade growth, and the current situation in different aspects of logistics income growth rates all drop as one moves inland across the region and offer policy recommendations. from coastal areas.1 The evidence on inland regions makes an especially strong case for the importance of access to international markets, because within- Trade and Logistics Nexus country differences control for institutional charac- The literature offers substantial evidence linking teristics that cross-country regressions cannot. improvements in transport and logistics directly to improvements in export performance. The effects Country Group Perspective are especially strong when importers have access to multiple suppliers of highly substitutable com- A useful way to organize thinking about logistics in modities. Comparing sales by manufacturers of East Asia is to place economies on a graph of trade similar products, Hummels (1999) estimates that openness and accessibility (Figure 5.1). Those lying exporters with 1 percent lower shipping costs will above the horizontal axis score high on measures of enjoy a 5­8 percent higher market share. Limao openness. The economies to the right of the vertical and Venables (2001) estimate that differences in axis are accessible to world markets in the sense of infrastructure quality account for 40 percent of the having superior logistics and low transport costs.2 variation in transport costs for coastal countries The combination of these two measures groups and up to 60 percent for landlocked countries. economies on the basis of their current logistics Fink, Mattoo and Neagu (2002) estimate that liber- status, as well as on the basis of the value of alizing the provision of port services and regulating improved openness and logistics services in the the exercise of market power in shipping could future--that is, economies with fewer political bar- reduce shipping costs by nearly a third. riers to trade can enjoy greater returns to logistics A World Bank study (Wilson and others 2002) investments than those whose tariff structures shows that member countries of the Asia-Pacific would prevent much trade growth even with Economic Cooperation (APEC) forum differ sub- world-class infrastructure. Similarly, accession to stantially in the quality of their logistics and trade the World Trade Organization (WTO) may be of facilitation across a broad range of measures, limited value if logistics services are too weak to including ports infrastructure, customs clearance, support trade growth. regulatory administration, and e-business use. Figure 5.1 suggests two additional correlates: per They find that these differences are related signifi- capita incomes and the commodity structure of cantly to differences in trade performance, and trade. The economies in the upper-right quadrant conclude that substantial growth in trade within enjoy higher incomes than those in the lower left, the APEC bloc could be accomplished by bringing and their exports are high-technology manufac- lagging countries up to median performance levels. tures rather than resource-based commodities. Furthermore, improving access to international Causality probably runs both ways. Economies like markets raises incomes. Frankel and Romer (1999) those of Singapore and Hong Kong (China) have Trade and Logistics: An East Asia Perspective 79 FIGURE 5.1 Potential Contribution of Transport to Economic Growth in East Asia Openness 2.50 2.00 Singapore Taiwan 1.50 Hong Kong Japan Korea 1.00 Malaysia Thailand NE 0.50 Thailand Mongolia China China Inland Vietnam -1.00Philippines 0.00 Accessibility Philippines ­15.00 ­10.00 ­5.00 0.00 Mindanao5.00 10.00 15.00 Vietnam Inland Indonesia ­0.50 Indonesia Sulawesi Cambodia­1.00 Laos PNG ­1.50 ­2.00 Source: Carruthers and Bajpai (2002). TABLE 5.1 Availability of Transport Infrastructure No. of airports with paved Total roads Percent of runways longer (km/ roads Railways than 1,523 m/ 1,000 km2) paved (km/1,000 km2) 1,000 km2 Cambodia 78.6 11.6 3.3 22.1 China 146.3 28.3 7.1 27.4 Indonesia 178.5 46.3 3.4 32.8 Lao PDR 59.1 24.0 0.0 25.3 Korea, Rep. of 888.9 74.7 31.7 375.7 Malaysia 196.1 75.3 5.5 63.7 Mongolia 46.1 2.2 1.2 4.5 Philippines 606.5 19.8 3.0 113.3 Thailand 125.7 97.5 7.9 75.9 Vietnam 283.1 25.1 9.5 45.5 Argentina 78.6 29.5 12.3 32.8 Brazil 234.0 9.3 3.6 21.3 Mexico 167.9 29.7 9.3 64.8 USA 695.4 90.1 23.2 190.7 France 1,621.8 100.0 58.1 138.1 Poland 1,218.6 65.6 74.9 236.7 Note: The World Development Report 1994 (World Bank 1994) provided a broad assessment of the basic infra- structure of 132 countries. The consultants for this report expanded and updated the data. UN ESCAP provides an even more comprehensive and current transport database for Asia at www.unescap.org/tctd/data/index2.asp. Source: Policy Research Corporation (2003). 80 East Asia Integrates grown rich in part because their past investments in Pearl River Delta on China's southern coast. The superior logistics have facilitated trade. Meanwhile, major success factors for these and other newly Mongolia, Lao People's Democratic Republic (Lao emerging ports are their transaction cost and time PDR), and Cambodia still suffer from poor basic cost advantages. Other advantages include their road access (Table 5.1). Similarly, while sophisti- efficient auxiliary services (particularly customs cated logistics facilitate the move up the manufac- and freight forwarding), less congested road and turing quality ladder, demands from advanced rail access, and better links to national and interna- manufacturers may push the private sector to tional transport networks. Relative to Singapore, improve logistics. Figure 5.1 also presents data for the competitive advantage of Tanjung Pelepas will inland regions of China, Thailand, and Vietnam as increase as Malaysia expands its rail network and well as for the islands of Mindanao in the Philip- links the port to Cambodia, Thailand, and Viet- pines and Sulawesi in Indonesia, because these nam. Likewise, the growth in traffic through ports areas have very different access to international in southern China is expected to accelerate with the markets compared with the coastal metropolitan dismantling of restrictions on the direct imports of areas or principal islands in the same countries. goods from the United States and Taiwan. Using Figure 5.1, we divided the economies of the East Asia region into three groups: Group 2: Outward-Oriented, Accessible 1. Outward-oriented, highly accessible: Hong The countries that fall in this category--China, Kong, Rep. of Korea, Singapore, Taiwan (China) Indonesia, Malaysia, the Philippines, and Thai- 2. Outward-oriented, accessible: China, Indonesia, land--face daunting problems in addressing the Malaysia, the Philippines, Thailand logistics impediments to higher pro-poor eco- 3. Less open and accessible: Cambodia, Lao PDR, nomic growth. They are still in the early stages of Mongolia, Papua New Guinea, Samoa, Vietnam. designing national policies and institutional struc- tures that encourage rather than inhibit the growth of the multimodal transport services that they need Group 1: Outward-Oriented, Highly Accessible for increased external and domestic trade. Many In the higher-income economies--Hong Kong, parts of these countries have inadequate and ineffi- Korea, Singapore, and Taiwan--exports have cient port and shipping facilities, transport ser- shifted toward high-technology manufactures and vices, and land infrastructure. Poor rural roads services. Logistics costs are at very competitive lev- cause long transport times and high transport els, transport volumes are high, and multimodal costs, which in turn hamper the development of linkages are well developed. Given the strength of domestic markets and lead to substantial interre- global competition, however, these economies need gional price differences, particularly within China, to keep their logistics costs competitive and Indonesia, and the Philippines. Local authorities improve quality in line with the evolution in tech- and other agencies charge quasi-legal and illegal nology and with recently introduced security road tolls, often in a desperate attempt to raise rev- requirements. enue, but always to the detriment of market access. Hong Kong and Singapore have specialized in In the countries of this group, the form of con- providing logistics services as transshipment hubs tracts used for international trade and transport dis- for neighboring states. Singapore has been the main courages multimodal use. It is common for the pur- transshipment stop for much of Southeast Asia, chaser of exports to contract free on board (FOB) in and Hong Kong has specialized in direct shipment the port of the exporting country, and the seller to and from neighboring Guangdong Province, the (producer) is responsible for contracting landside manufacturing region that produces nearly 40 per- delivery to the specified port. Because no one agent cent of China's exports. But these two megaports assumes responsibility for the whole logistics chain, now face competition from emerging regional the full advantages of multimodal transport are not ports: Singapore from Tanjung Pelepas in Malaysia realized. Third-party logistics (3PL), in which an and Laem Chabang in Thailand, and Hong Kong industry contracts out its logistics functions to spe- from the rapidly expanding container ports of the cialized suppliers of logistics services, are uncom- Trade and Logistics: An East Asia Perspective 81 mon in Group 2 except in Thailand. In some cases, Highlands Highway, the prime road corridor used however, such as large international agroproduct for the export of minerals and coffee, is inadequate companies in the Philippines, large exporters have for container movement by trucks; it suffers from developed their own well-integrated logistics. Trade poor maintenance and frequent closures, causing documentation is a particularly difficult hurdle for enormous losses in trade. international trade with an inland origin or destina- Government policies in the realm of logistics tion. Customs and health authorities are very reluc- tend to lack stability, consistency, and transparency, tant to allow inland clearance. creating many additional problems for the process- Interviews with freight forwarders in the interior ing, storage, transport, grading, marketing, and sale of China suggest that inland transport costs can of products. Policy coordination between central account for about two-thirds of the total transport and provincial authorities is poor. Many countries costs from Chinese producers to overseas markets. in this group suffer from highly distorted prices as a In the context of WTO requirements, customs fees remnant of their previously centrally planned will be reduced. China's recent liberalization process economies. National standards on product quality, has led to an influx of foreign shipping services (for- packaging, storage, and transport conditions are eign carriers transported almost 60 percent of often rudimentary, and farmers lack management, export trade cargo in 2002). In some cases, ship- marketing, and logistics capabilities. In Vietnam, ment costs for exports are subsidized by imports.3 overall port, cargo handling, trucking, and rail tar- iffs may not be excessive, but relative inefficiencies impose high costs in terms of inventory and the Group 3: Less Open and Accessible unproductive times of vessels and road vehicles. These mostly low-income countries--Cambodia, The intermodal transport systems of most of Lao PDR, Mongolia, Papua New Guinea, Samoa, these countries are poorly integrated, with no Vietnam--are either former centrally planned streamlined procedures to support the seamless socialist economies or small island states, depen- movement of containers between coastal and dent on a small number of commodities and inland areas. Many countries lack container freight tourism. Their regimes, recent history, and geo- stations, yards, and trucks in their inland regions. graphic position are largely responsible for the Container tracking capability is particularly poor, small role that international trade plays in their with shippers often unaware of their containers' national economies. Except for Vietnam, they have whereabouts. The potential impact of improved low population density, small domestic markets, multimodal transport is well illustrated in a recent and primarily subsistence agrarian economies. study on the time and costs of container move- Many logistical constraints hamper the economic ments from Lao PDR to Europe. The study showed growth of these countries, the most fundamental that alternative multimodal routes could reduce the being lack of adequate infrastructure. Roads are present high door-to-door cost from Lao PDR to often closed and services suspended, and the state of Europe (through Danang and Singapore) by almost infrastructure requires the use of small, inefficient one-third (Banomyong and Beresford 2000). vehicles and vessels, which have high operating costs. Border procedures are also cumbersome and In Cambodia, for example, truck rates for distances time-consuming in these countries. One reason is over 100 kilometers vary from about $.034­$.116 excessive and ad hoc regulatory requirements. In per ton-kilometer, depending on road conditions. In Cambodia, which experiences enormous competi- Lao PDR, almost 40 percent of villages are more than tive pressures from its neighbors, customs clearance 6 kilometers from a main road, half are inaccessible takes 8 days for imports and 10­14 days for exports. during the rainy season, one-quarter of the district Informal payments are needed to handle vessels centers lack year-round road access, and a quarter of after 5 p.m. in its main port, Sihanoukville. Port provincial and local roads cannot be used during the charges at Sihanoukville are the highest in the rainy season. This situation not only makes it diffi- region, almost four times those at the comparably cult to get outputs to markets, but also increases the sized port of Songkhla in Thailand (Cambodia import costs of essential consumer products and Ministry of Commerce 2001). For landlocked inputs such as fertilizers. In Papua New Guinea, the countries, land border crossings invariably involve 82 East Asia Integrates delays and costs arising from the inefficient appli- that these remote areas become nearly autarkic. cation of national regulations, difficulties in cargo Remoteness is not limited to landlocked regions. In and document clearance, unnecessarily long stor- the Philippines, Mindanao has few connections age, high insurance premiums, and long waiting with population centers, and containerized cargo times. For example, Mongolian rail freight arriving services are limited. Inland transport costs of mov- at the border with China often waits for days for ing goods from some remote regions of China to Chinese rail wagons so that it can complete its jour- external markets are large multiples of the inland ney to the port of Tianjin. Because of a lack of qual- transport costs at the other end (Table 5.2). ified staff at the relevant ministries or government Logistics difficulties associated with remoteness departments, documentation is often inadequate. extend beyond the frequency and cost of transport. Border delays lead to extra storage charges, which For example, connections are poor between the cannot always be properly controlled. Detailed cus- main cities and many of the outlying counties and toms statistics are often collected at a high cost to villages in the Chinese provinces of Hunan and traders, but data that might be of use to them are Yunnan. In such interior regions, it is difficult to not compiled, published, or analyzed. maintain road access and harder still to maintain supporting facilities such as warehouses, cold stor- age rooms, intermodal terminals, and information Remote Regions within Countries structures to control flows. As a consequence, cargo Many East Asian countries have vast remote areas consolidation is limited, harvests cannot be stored, with poor connections to other domestic markets, and perishable products do not survive shipment. as well as to international sea and air gateways. The Inland regions are squeezed on two sides--by problem begins with low population density and higher prices for their imported consumption geographic remoteness, but it is worsened by the items and by lower net revenues (and therefore lack of basic transport infrastructure. Of course, lower wages) for their exported output.4 In Lao where markets are distant and trade volumes are PDR, for example, regions that enjoy more days of low, it is difficult to justify building and maintaining road access have much lower rates of poverty. Fig- even basic infrastructure. But this situation creates a ure 5.2 charts poverty rates across Lao provinces vicious cycle, resulting in transport costs so high against road access in the rainy season. FIGURE 5.2 Poverty and Accessibility in Lao PDR Percentage of population in poverty 80 70 60 50 40 30 20 25 30 35 40 45 50 55 60 65 70 75 Rainy season road access (%) Source: D. Hummels's calculations, based on Arnold (2003). Trade and Logistics: An East Asia Perspective 83 Similar arguments apply across countries. A nounced in countries such as Cambodia, Lao PDR, country with high logistics costs is, in a broad eco- and Vietnam, where agricultural production is a nomic sense, far from international markets in that large portion of national output and where interior it pays high prices for its imports and receives low regions are poorly connected or inaccessible. prices for its exports.5 Good internal logistics facilitate commodity shipments from surplus to shortfall regions, smoothing price variations between them and pro- Benefits from Improved Logistics viding a kind of insurance against shortage and What are the specific channels through which glut. Better market access appears to dampen price investments in logistics can improve welfare in volatility for a broad range of products. According developing economies? This section reviews exam- to Engel and Rogers (1996), the volatility in goods ples and evidence taken both from academic stud- prices between city pairs rises with the distance ies and from a series of recent country studies by between the cities, and is especially large for city the World Bank. pairs across national borders. Essentially, arbitrage is necessary to narrow price differentials across locations, but such an objective is much harder to Reduced "Wedge" between Consumer and achieve when logistics are poor. Producer Prices Reducing the cost of moving goods between mar- Reduced Inventory Costs kets reduces the prices paid by consumers and increases the prices received by producers. On the Better transport and logistics systems not only consumer side, this effect can be seen most clearly lower the costs of delivery, but also make the timing by examining the price of goods at the port relative of delivery more reliable. Because producers cannot to the price of goods inland. Producers will not ship manufacture goods without the inputs they need goods inland unless the price they receive, net of and retailers cannot sell goods they do not have in shipping, is at least as high as the price at the port. stock, firms must hold large inventories of goods if As a result, inland consumers bear the full burden delivery times are uncertain. Gausch and Kogan of shipping costs for any good that is not produced (2001) found that inventory holdings in manufac- locally. This burden can be substantial. Data from turing are two to five times higher in developing Mongolia, for example, reveal substantial price dif- countries than in the United States, and they esti- ferences between Ulaanbaatar and outlying regions, mated that halving inventories could reduce unit with as much as 67 percent of the higher costs in production costs by 20 percent. outlying regions attributable to transport. One reason for long and uncertain delivery A similar logic applies to exports. International times is poor infrastructure--roads may be impass- markets will not pay more for goods produced in able and railways nonexistent, or roads may inflict inland regions, and so the ex factory or ex farm damage on trucks that involves high costs and long prices are reduced by the full inland logistics costs delays. Another reason is peak load congestion, a of access to the port. severe problem in areas where population and manufacturing intensity have grown faster than infrastructure capacity. In Vietnam's Ho Chi Minh Insurance against Regional Price Fluctuations City, for example, the government bans trucks Agricultural output sometimes varies because of heavier than 2 metric tons in the inner city between unpredictable weather--one region may experi- 6 a.m. and 6 p.m. to combat growing traffic conges- ence drought and food shortages, while another tion. Delivery trucks sit idle during that period, enjoys sufficient rainfall and ample harvests. If the forcing manufacturers to maintain large invento- regions are separated by poor logistics and trans- ries all day. By contrast, under modern just-in-time port systems, consumers in the shortfall region will techniques, deliveries are continuous, and the face very high food prices, and producers in the transport capital stock is in constant use. surplus region will face a market glut. Regional But leaner production techniques require a sub- variations in agricultural prices are especially pro- stantial flow of information. Manufacturing firms 84 East Asia Integrates TABLE 5.2 Composition of Logistics Costs of Container Transport from Inland China (Chongqing) to U.S. West Coast Activity US$ per TEU Percent of total cost Land access to port 2,300 63 Port handling 200 5 Maritime transport 750 21 Port handling 150 4 Land access to final destination 250 7 Total 3,650 100 Note: A TEU is a 20-foot equivalent unit, the size of a standard container. Source: Carruthers and Bajpai (2002). can run with small inventories of inputs only if that many firms can share are significant. Second, they are certain where and when the next shipment by lowering the marginal costs of serving markets, will arrive--a situation that requires sophisticated firms can increase sales and spread entry costs over electronic data interchange. Table 5.5 shows that in more units. many countries in the East Asia region systems for The diversification argument also applies across this purpose are far from competitive.6 markets, because reliance on a single export desti- nation leaves firms subject to significant risks from business cycles. The market destinations for East More Developed Markets Asia's exports are mainly the industrial countries, Well-developed logistics increase consumers' and intraregional trade is limited. This situation choices and producers' sources of supply, and bring stems in part from the pattern of logistics costs and more markets within the reach of producers. the institutional barriers to land-based trade that According to Evenett and Venables (2002), 40 per- make many countries in the region effectively cent of the growth in trade in East Asia arises from closer to industrial countries than their own geo- offering new product lines and extending exports graphic neighbors. Thus, although the proposed of existing product lines to new trading partners. conversion of the Association of Southeast Asian Product diversification benefits both agricul- Nations (ASEAN) to a free trade area would do tural- and industrial-based economies. For both, much to remove tariff barriers, the signatory coun- specialization in a narrow range of products can be tries will have to make a corresponding change in dangerous because of price volatility. Recent studies their implementation of customs and other restric- have shown that most of the differences in trade tions on trade, supported by a reduction in logistics levels between small and large economies can be costs, if they are to realize the full benefits of tariff attributed to differences in the range of goods reductions. traded, and that much of the growth in imports that results from lowering trading costs arises from expanding the set of products available.7 Romer Moving Up the Value Chain (1994) shows that the welfare benefits from As countries move up the value chain from resource expanded product variety can dwarf those from extraction to sophisticated manufacturing and standard calculations of the gains from trade. And higher value added agriculture, they must develop manufacturing firms are consumers as well. Feen- their logistics capabilities accordingly. As the value- stra and others (1999) find that expanding the vari- to-weight ratio of a country's exports rises, logistics ety of inputs in Taiwan and Korea is associated with and transport costs will fall, but only if the nature of productivity growth in manufacturing. logistics and transport services stays unchanged. In Better logistics allow expanded variety in two practice, the more sophisticated exports generally ways. First, they may directly lower the fixed costs impose much greater demands on logistics and of expansion. The scale advantages to providing transport services. For example, iron ore and bulk transport hubs, warehousing, and logistics services grains may be heavy and difficult to move over land Trade and Logistics: An East Asia Perspective 85 for long distances, but their shipment is otherwise tainer usage represents both a revolution in mar- uncomplicated. High-value agriculture (flowers, itime technology and a significant logistics chal- fruits, and seafood) requires careful handling, time- lenge to economies in the region. liness, and product standardization. Electronics Although regional container ports are becoming manufacturing requires all this, as well as tightly more efficient at handling containers, they are not integrated supply chains. keeping pace with the rapidly growing demand for berth space. Although the capacity of the container fleet on the East Asian routes increased by more Regional Transport and Logistics than 20 percent a year between 1980 and 2000, the Issues capacity of container berths to handle those ships Despite two decades of improvement, East Asia has increased by less than 8 percent a year. Countries significant scope for further reducing its transport are responding to the shortage by adding new and logistics costs.8 The rest of this section reviews berths, converting general cargo berths to container problems and opportunities in specific areas, as a handling, and developing new ports (Vitasa and prelude to the policy recommendations offered Seprato 1999). later in this chapter. Expanded container capacity requires the avail- ability of additional land for container storage, road and rail links, and associated services. These Maritime Issues issues are discussed later in this section. Ninety percent of the world's trade in manufac- Higher East Asian trade volumes have led liner tured goods is now carried by containers, and the services to introduce large container ships that use of containerized shipping has increased require deeper access channels, which often can be throughout East Asia (Table 5.3). During the 1990s, provided only through voluminous dredging. As a total container movements increased by nearly 10 result, ports in river estuaries (such as Bangkok, percent a year, with the fastest growth occurring at Haiphong, Saigon, and Shanghai) may become less the ports of China and new ports in Malaysia and competitive than coastal deepwater ports (such as Thailand (Frenkel 1998). The rapid growth in con- Laem Chabang and Hong Kong). Some existing TABLE 5.3 Container Movements at Selected East Asian Ports, 1995­2001 Annual growth Port 1995 1996 1997 1998 1999 2000 2001 (%) Hong Kong (China) 12,550 13,460 14,567 14,582 16,211 17,800 17,900 6.1 Singapore 11,846 12,944 14,135 15,136 15,945 17,040 15,520 4.6 Shanghai 1,196 1,305 2,527 3,066 4,206 5,613 6,310 31.9 Port Klang 1,134 1,410 1,685 1,820 2,550 3,206 3,759 22.1 Laem Chabang -- 729 1,036 1,425 1,756 2,195 2,424 27.1 Qingdao 603 810 1,031 1,213 1,540 2,100 2,639 27.9 Tianjin 702 822 935 1,018 1,302 1,708 2,010 19.2 Guangzhou 515 558 687 848 1,179 1,430 1,628 21.1 Taichung, Taiwan 447 695 842 880 1,107 1,130 1,069 15.6 Total (East Asia) 54,433 57,836 65,119 68,155 75,155 83,422 94,267 9.6 Hong Kong and Singa- pore (%) 45 46 44 44 43 42 35 -- Not available. Source: Containerization International, March 2002. 86 East Asia Integrates estuarial ports are already looking for new develop- gression to the earlier stage of second-party ments on the coast to overcome this disadvantage. logistics--in which companies unify their inter- Because the scope for further reducing costs by nal transport and warehousing functions and increasing vessel size is limited, the next development create their own internal logistics depart- is likely to be more direct services from what are now ments--is still under way. feeder ports. With higher volumes and more efficient · Develop effective communications systems. Such smaller vessels, such a step could overcome the high systems will allow freight forwarders to take cost penalty of transfers in the hub ports. This trend advantage of the shipping alternatives available is already reflected in the slower growth rates of the and inform clients about the status and location two regional megaports, shown in Table 5.3. of their freight. Some of the least accessible countries have Multimodal Transport improved the quality and scope of their informa- Much of East Asia uses containers only for the mar- tion systems (Table 5.5), but few have been able to itime part of trips, loading and unloading them in develop freight forwarding agencies that perform as the ports rather than at the origin and destination of well as those in the more accessible and trade-open their cargo. This approach, however, eliminates the countries. The efficiency of logistics services corre- main cost savings advantages of container use. The lates closely with the efficiency of other trade facili- countries most successful at encouraging door-to- tation procedures. Countries that have poorly inte- door movement of containers using multimodal grated trade services, such as customs requirements transport will be the ones best equipped to compete that are badly coordinated with those of taxation and to bring trade benefits to their more remote and health agencies, also tend to have poorly devel- regions. To achieve this integration, they will need to: oped logistics services. Evidence from some East Asian countries, particularly Thailand, suggests · Match inland infrastructure with maritime infra- that the presence of a strong logistics industry can structure. One important reason that containers be effective in bringing about efficiencies in these do not move inland from container ports is that other trade facilitation procedures. road and rail infrastructure lack the appropriate Some evidence also suggests that even those carrying capacity and vehicle dimensions for the countries with the least developed logistics systems transport of loaded containers. have recognized the need to catch up. According to · Simplify trade documentation. Examples include a recent report on logistics in China,9 the develop- the use of through waybills and single invoices ment of hypermarkets and other large retail cus- for all modes. In customs clearance times, those tomers with significant bargaining power is creat- of the East Asian economies are quite similar to ing a demand for better logistics services that was those of other developing countries (Table 5.4), not obvious even a few years ago. At the same time, but, taken together, developing countries are sig- the Chinese government is encouraging companies nificantly slower than developed countries. to outsource their logistics services, and as part of Another regulatory change--allowing containers the agreement on WTO accession, it is in the to be cleared for tariffs, customs, health, and tax- process of rapidly opening its logistics services to ation charges at inland locations away from the foreign competitors and participants. ports--would help to reduce port congestion, but could raise additional security concerns. Ports and Land Access · Develop an efficient freight forwarding industry. As noted earlier, third-party logistics is not a The high costs of land access to ports, reinforced by well-advanced concept in East Asia. In industrial the effects of production agglomeration, have countries, almost a third of logistics turnover is caused an excessive concentration of export-related contracted to 3PL providers, but even in several activities in port cities and essentially have industrialized East Asian economies, barely 10 restricted the benefits of trade growth to the areas percent of trade-related transport services is immediately surrounding ports. In China, for provided in this way. In some countries, pro- example, more than 90 percent of foreign direct Trade and Logistics: An East Asia Perspective 87 TABLE 5.4 Customs Clearance Times Average no. of days for customs clearance Country Air Sea LCL Sea FCL France 1 4 2 Germany 1 1 1 Greece 1 1 1 Netherlands 1 2 2 Spain 2 2 2 Sweden 1 2 2 USA 2 3 3 Average, sample of developed countries 1.3 2.1 1.9 China 4 30 5 Hong Kong (China) 2 4 3 Indonesia 3 4 4 Malaysia 4 4 4 Philippines 4 5 3 Singapore 2 3 3 Taiwan (China) 4 10 7 Thailand 5 5 5 Vietnam 5 7 7 Average, sample of East Asian counties 3.7 8.0 4.6 Argentina 7 15 12 Brazil 10 10 10 India 8 10 12 Mexico 4 7 4 Mozambique 5 8 8 Russia 10 12 15 Zimbabwe 4 5 5 Average, sample of other developing countries 6.9 9.6 9.4 Note: LCL = less than container load; FCL = full container load. Source: International Exhibition Logistics Associates (www.iela.org). investment in export-oriented activities has gone to Traffic congestion costs, in particular, can be enor- the four main coastal provinces (Fujian, Guang- mous. A recent study in Bangkok estimated that dong, Jiangsu, and Shanghai). Similarly, the multi- moving port-related activities out of the downtown plier effect of the textile export boom in Cambodia area would result in a 10 percent reduction in peak- has been limited largely to areas that have easy hour trips and would entail benefits of up to access to the deepwater port at Sihanoukville. US$400 million10 annually (UN ESCAP 2000). If the benefits of trade are to be more widely Reductions in port access costs depend on having distributed, the penalties of inaccessibility have to available adequate infrastructure, the appropriate be addressed. Such action could stimulate trade- vehicles, and the logistics technology that allows induced growth in currently inaccessible regions, both to be used efficiently (Box 5.1.) and, if successful, could slow the growth of trade- As container ports expand, municipalities find it induced urban congestion and pollution in port increasingly difficult to accommodate both the cities. added space requirements and the road congestion The social costs of the concentration of manu- that results from the high volumes of truck traffic facturing activities in port cities can be significant. servicing the ports. In some instances, the only fea- 88 East Asia Integrates TABLE 5.5 Electronic Data Interchange and Transport E-Commerce in Selected Countries Electronic Port Traders in Customs Traders in All parties trade in operators port EDI agency customs EDI electronically transport Country MIS system MIS system linked services Japan X X X X Singapore X X X X X X Korea, Rep. of X X X X X X Thailand X X X X X X Philippines X X X X Indonesia X X X X Vietnam None Lao PDR None Cambodia None Note: MIS = management information system; EDI = electronic data interchange. Source: Based on UN ESCAP (2002). BOX 5.1 Successful Integration of Ports and Land Transport Networks: Republic of Korea The ports of the Republic of Korea benefit from Korea (www.pusanconsulting.co.kr/En-about one of East Asia's most developed land access Pusan.htm). networks. Major road and rail links run to the The Yangsan inland container terminal has ports of Pusan and Kwangyang from Korea's been constructed to relieve port-generated traf- major manufacturing regions. Kwangyang, fic congestion and environmental problems which is alongside a major steel mill and indus- resulting from the massive transport movements trial complex, is now in its second stage of devel- that the port generates. Another inland con- opment, with a potential capacity of 2.4 million tainer terminal is being developed in the center TEU (20-foot equivalent unit). Pusan has devel- of the Korean peninsula to serve the growing oped a new port area away from the downtown industrial zones on the west coast and in the area. This location has reduced traffic conges- center of the country. Together with the ports, tion and air pollution, and improved logistics these terminals are part of a logistics system efficiency has made the new port easier to reach based on an advanced electronic data inter- from the city's industrial areas and the rest of change and information service. sible response is to build a new port. A less costly ciated with the existing port must be reduced even response is to move port activities out of down- further, and this need is part of the justification town urban areas, while retaining the employment (together with the limited depth of the access chan- and business activities that the ports attract. For nel) for the development of a new deepwater port example, one of the principal reasons for develop- away from the urban area. Based on two offshore ing the Waigaoqiao container terminal and Luojing islands, the new port will require new 30-kilometer coal terminal outside Shanghai was to move port- access roads, including a bridge, and is conserva- based traffic out of the urban area. This goal was tively estimated to cost up to $4 billion just for the achieved between 1996 and 1998. The port traffic first stage. handled at terminals within the city was reduced by When the depth of the maritime access channel about 13 million metric tons, and that in the new is not a constraint on growth, the urban congestion terminals outside the city was increased by about problems of port growth can often be solved by the same amount. However, the urban traffic asso- moving the nonmaritime port activities (mostly Trade and Logistics: An East Asia Perspective 89 value-adding production and packaging services) that can offer better services are growing faster than inland, closer to the industries that the port serves, the larger ones. and building rail links to avoid generating extra Airlines and traditional freight forwarders both road traffic (Rafferty 2003). compete and cooperate with each other to provide air transport-based freight services. The progress of air freight forwarding and air freight logistics in Air Freight Hong Kong, Japan, Korea, and Singapore compares Air freight accounts for only about 1 percent of East favorably with that in the United States and Europe. Asia's international trade by volume, but more than Some countries in the region still depend on a few 35 percent by value. More than for other regions, air multinational air carriers for efficient air freight freight is important for East Asia. For one thing, the logistics. distance from the region's major markets--the United States and Europe--makes quick delivery of Policy Recommendations sea freight impossible. And, for another, a high pro- portion of the region's manufactured exports require Against this background, governments in East Asia timely delivery. (Because these manufactures have need to take action to improve trade-related logis- high value-to-weight ratios, the ad valorem cost of tics on several fronts if they are to increase their their air transport is comparatively low.) trade competitiveness. A ranking of cargo volumes by airport reveals the importance of air freight to East Asia. Of air- Domestic Integration ports outside the United States, those in East Asia took 10 of the top 30 places for air freight volume For the less open and less accessible group of coun- in both 2000 and 2001, and accounted for more tries, and for landlocked regions and remote than 30 percent of air freight at the top 30 airports regions in China, Indonesia, the Philippines, and worldwide (Table 5.6). Thailand, the development of more tightly inte- Good air freight facilities are important in grated domestic markets and logistics systems is a attracting fast-growing, high value added indus- high priority. Besides promoting an appropriate tries. Competition is, therefore, growing between mix of modes--roads, waterways, and rail--insti- airports in the region to act as a hub for major tutional actions must be taken to extend better logistics companies. The smaller, newer airports transport services to remote areas and to establish TABLE 5.6 Ranking of Major Freight Airports in East Asia, 2001 Amount of Global cargo handled ranking Airport Code (metric tons) 3 Hong Kong (China) HKG 2,099,605 5 Tokyo NRT 1,680,938 8 Singapore SIN 1,529,930 15 Inchon ICN 1,196,845 16 Taipei TPE 1,189,874 18 Osaka KIX 871,161 19 Bangkok BKK 842,588 23 Tokyo HND 725,124 27 Seoul SEL 598,620 28 Beijing PEK 586,704 Total East Asia 33,210,120 Percent of total top 30 34.1 Source: Airports Council International, 2002. 90 East Asia Integrates better conditions for market development through, firms seeking seamless freight movement across for example, postharvest services, cargo consolida- modes. A transparent, uniform regulatory and legal tion through farmer or business associations, infor- regime for private sector participation, safety, envi- mation on prices and market demand, access to ronment, traffic rules, vehicle weight, and dimension credits, and human skills. is a prerequisite for an effective transport ministry. An obvious solution is regulatory consolidation so that businesses face consistent rules. Private Sector Improved transport infrastructure in trade cor- Transport of the outputs of very simple extractive ridors is an important part of a regional poverty industries may not require advanced logistics abili- reduction strategy that seeks to attract foreign ties, but high-value products call for services such as direct investment and trade-based growth away freight forwarding, third-party logistics, warehous- from port cities. Although infrastructure develop- ing, storage, packaging, e-business, and trucking.11 ment in competing modes can stimulate inter- Logistics needs of this kind tend to be better modal competition, it can be costly in the early served by the private sector than by the public sec- stages of trade corridor expansion. Government tor. The governments of many countries may be should promote an integrated planning framework well advised to halt providing logistics services for developing these corridors, with an established directly and turn instead to creating an environ- hierarchy of modal interfaces (inland terminals, ment conducive to competition and private invest- container stations, cargo clearance facilities for cus- ment. Such an approach may entail legalizing and toms, health inspections, and tax payments). Coor- deregulating freight forwarders and allowing new dination between the agencies responsible for dif- entrants, including the international companies ferent modes is essential here. that can be a major source of the capital, technol- A further problem is that access to publicly ogy, and new management practices needed to administered transport may not be allocated develop sophisticated services. Even areas tradi- according to an efficient market-based pricing tionally managed by government, such as port model. In central China, for example, some bulk management and operations, may benefit from pri- commodities are allowed queuing priority on rail vate services.12 Countries throughout East Asia are lines. Such a practice reverses the priorities that likely to reap a very high payoff from a well-devel- would exist if queuing were determined by market oped private logistics industry. prices. Often some public or private monopolies14 Given the fixed costs of entry, private operators are unable to provide satisfactory services despite are reluctant to provide services where trade vol- their high service charges. umes are low. One solution supported by the World Countries already within the WTO, such as Bank is to provide initial public funding for the China, and those likely to join in the near future, development of facilities where the economic bene- such as Cambodia, Vietnam, and Lao PDR, need to fits indicate earlier development than the financial review their transport regulations and policies. returns would support. The initial public invest- This review would be aimed at bringing them into ment in facilities such as inland container terminals line with WTO and General Agreement on Trade in in potentially high-growth regions distant from Services (GATS) rules on nondiscrimination ports can be recovered through later concession between foreign and local service providers, trans- revenues or outright sales.13 parency of rules and regulations, and elimination of quantitative restrictions. Regulatory Environment Cross-Border Facilitation. Cross-border facilitation Transport. Lack of regulatory coordination across is a key intervention in lowering overall export and transport modes is a common problem in East Asia. import transaction costs. It promises high returns One regulatory agency monitors ports; another for all East Asian countries, although less for those monitors roads; and a third monitors rail. And each that are relatively advanced. It can be best achieved may have different plans, standards, and reporting by harmonizing and simplifying customs proce- requirements. This situation presents problems for dures, sharing information, modernizing informa- Trade and Logistics: An East Asia Perspective 91 tion and communications technology and customs tions, and training for customs and security staff. administration, and establishing transparent transit The solutions that countries select are likely to vary rules and postentry compliance audits. Plans for according to country conditions (human skills and improvements must take into account the interests resources, technology, quality of facilities), risk pro- of related public and private communities--includ- files, and types of trade transactions. The World ing customs, freight forwarders, shippers, port oper- Customs Organization Task Force on Security and ators, shipping lines, insurers, and bankers. Trade Facilitation is expected to develop appropri- ate processes and identify the related needs for Urban Land Use and Management. Urban gov- capacity building. ernments need to exercise market-responsive land Because security and facilitation are two sides of use policies for locating logistics infrastructure, the same coin, an integral approach to supply chain especially ports. This is perhaps more difficult than management would be appropriate, although now it sounds because of unpriced externalities. Firms the focus is on limiting the security risks of imports move production into clusters because there they by improving a few discrete points in supply chains enjoy positive productivity externalities--informa- (customs). Security initiatives are now targeting tion spillovers, access to intermediate inputs, and containers as well as air cargo and passenger move- specialized human capital--but they do not take ments.15 For security initiatives to succeed, the rele- into account the effect that clustering has on vant technologies and information systems must be already severe congestion. Government efforts to shared among government agencies, shippers, and relieve congestion or relocate facilities must take corporations. into account the trade-offs between the positive externalities (agglomeration economy) and nega- Regional Cooperation tive externalities (congestion, pollution) of the businesses or facility to be displaced, such as a port Such sharing is best achieved through a regional or container depot. Unfortunately, these externali- dialogue--a dialogue that is also central to the har- ties, especially the positive spillovers, remain diffi- monization of national, regional, and international cult to measure. rules and policies related to trade facilitation. Cooperation is especially important in assisting landlocked countries, such as Mongolia and Laos Security PDR, and those with long land borders, mostly in The September 11, 2001, terrorist attacks on the the Mekong River region. United States heightened the importance of effi- Ongoing regional and subregional initiatives cient and secure trade facilitation in supporting (ASEAN, APEC, Mekong River Commission, and trade-led growth. The increased transport costs Greater Mekong subregion) provide regular oppor- related to security have already begun to affect tunities to cooperate on country-specific invest- trade volumes and the competitiveness of ports. ments and reforms to address the regional needs.16 According to an estimate by the Organisation for The World Bank's country-specific programs--and Economic Co-operation and Development more specifically the Global Trade Facilitation Part- (OECD), trading costs excluding the increased nership17 among more than 100 international, pub- inventory and other "behind the border" costs have lic, and private agencies--can also assist countries risen by about 1­3 percent, resulting in a loss of through investments and technical assistance on about $75 billion in aggregate welfare. trade facilitation issues, while serving as a major The increased costs and challenges of trade facil- source for knowledge sharing in global best prac- itation will depress the level of trade of countries tices and capacity building.18 that cannot manage them successfully. All the inter- national agencies involved directly in international Endnotes trade are helping exporting countries to achieve higher security as expeditiously as possible. They 1. Gallup, Sachs, and Mellinger (1999) point out that "core" coastal regions worldwide contain 10 percent of the world's are providing funding for the procurement of population but contribute 35 percent of the gross world equipment, technical assistance in drafting regula- product. 92 East Asia Integrates 2. The measures of openness are based on those indicated in 15. See Reddy (2002). Traditional security methods result in the Global Competitiveness Report 2001­2002 (World Eco- processes that increase the cycle time for activities, such as nomic Forum 2002). Values for countries not included in queuing to go through metal detectors and physical inspec- that report have been added using World Bank staff assess- tions. Reddy argues that the preferred approach from both ments of the three indices (see Carruthers and Bajpai the security and efficiency perspectives is for goods and raw 2002). All values have been normalized for the countries of materials to be secured after inspection at the point of ori- East Asia. The measures of accessibility are based on the gin. If this approach can be implemented for the 1 percent cost of transporting a standard container from the metro- of shippers that account for more than 60 percent of all politan region of the largest port to Hamburg, Germany. container movements at the point of origin, then national For inland regions, the land transport cost to the metropol- security agencies can focus on a more thorough examina- itan region has been added. tion of the remaining 40 percent of containers. The overall 3. For example, in Thailand the inbound rate for a 20-foot result should be to minimize the impact on the flow of container is about $380 compared with $168 for an out- international trade. A key to this approach is the security of bound container. containers once they have left their country of origin. Some 4. Transport cost incidence is a primary candidate for technologies already available allow a container to be explaining why incomes drop as international market tracked and monitored at all times between its origin and access drops, as shown by Wei and Yi (2001) for China and destination. by Frankel and Romer (1999) and Redding and Venables 16. Including for cross-border transport links, improved cus- (2002) for the situation worldwide. toms processing, and harmonization of rules dealing with 5. A manufacturer in Thailand claimed to be closer in time customs, vehicle standards, border crossings by vehicles and cost to the United States than to Vietnam because of registered in neighboring or third countries, movement of the poor land access facilities from Thailand to Vietnam. hazardous material, insurance coverage, safety and treat- 6. Certainty in delivery also requires that shipments be secure ment of transit traffic. from theft and pilferage. This problem is likely to be most 17. The Global Facilitation Partnership for Transport and Trade serious for products that are of high value but sufficiently (GFP) seeks to pull together all interested parties, public homogeneous that gray markets exist for resale. Firms' and private, national and international, that want to help strategies to avoid theft can result in higher transport costs achieve significant improvements in transport and trade and uncertainty in delivery times. For example, manufac- facilitation in World Bank member countries. The partners turers of integrated circuits in the Philippines reduce will together agree to design and undertake specific pro- hijacking of shipments by using irregular and therefore grams toward meeting this objective, making use of their costly transport schedules. respective comparative advantages in the subject matter in a 7. See Hummels and Klenow (2001) for cross-sectional com- coordinated fashion (http://wbln0018.worldbank.org/twu/ parisons of small and large economies, and Hillberry and gfp.nsf/). McDaniel (2002) for an examination of how trade liberal- 18. For example, by means of distance learning and toolkits on ization affects trade growth. customs modernization, supply chain management, and 8. An assessment of the issues at the country and specific trade port modernization. flow levels is provided in country studies undertaken for the World Bank in 2002. See Arnold and Villareal (2002); Arnold, Banomyong, and Ritthironk (2002); Infrastructure References Consulting Ltd. (2002); International Trade Institute of Sin- gapore (2002); and Nomura Research Institute (2002). The word processed describes informally reproduced works that 9. See Shaw and Wang (2002). Other articles in this series of may not be commonly available through libraries. the McKinsey Quarterly also confirm the growing impor- tance of multimodal transport arrangements. Arnold, John. 2003. "Logistics Development and Trade Facilita- 10. All dollar amounts are current U.S. dollars. tion in Lao PDR." Working Paper No. 3, East Asia Transport 11. Domestic providers of such services may not be up to inter- Unit. World Bank, Washington, D.C. national standards. For example, because of concerns about Arnold, John, and Theresa Villareal. 2002. "Philippine Logistics the quality of local freight forwarders, Japanese manufac- Study." Study prepared for the World Bank, Washington, D.C. turers in Vietnam insist on working with freight forwarders Arnold, John, Ruth Banomyong, and Nipawis Ritthironk. 2002. that are joint venture partners with familiar (Japanese) "Logistics Development and Trade Facilitation in Lao PDR." logistics providers. Part of the reason may be a need for the Study prepared for the World Bank, Washington, D.C. superior technology that the foreign logistics providers can Banomyong, R., and A. K. C. Beresford. 2000. "Multi-Modal offer, and part may be the long-standing relationship Transport Systems: The Case of the Laotian Garment Indus- between the manufacturer and logistics provider. Freight try." Trois-Rivieres, France. Processed. forwarders that fail to provide reliable delivery times and Cambodia Ministry of Commerce. 2001."Integration and Com- careful handling jeopardize contracts not only in Vietnam petitiveness Study--Part B." Pilot study prepared under but also back in Japan. Integrated Framework for Trade-Related Technical Assis- 12. See Fink, Mattoo, and Neagu (2002) for the Latin American tance. experience with port privatization. Carruthers, Robin, and Jitendra N. Bajpai. 2002. "Trends in 13. This approach is being applied in the China Container Trade and Logistics: An East Asian Perspective." Working Transport Project supported by the World Bank. Paper No. 2, Transport Sector Unit. World Bank, Washing- 14. Such as the Philippines Ports Authority, the private ship- ton, D.C. ping cartels Vinamarine and Vinalines in Vietnam, and the Engel, Charles, and John. H. Rogers. 1996. "How Wide Is the traders' cartel for freight forwarding in Lao PDR. Border?" American Economic Review 86 (5): 1112­25. Trade and Logistics: An East Asia Perspective 93 Evenett, Simon J., and Anthony J. Venables. 2002. "The Geo- Policy Research Corporation. 2003. "East Asia Transport and graphic Spread of Trade: Evidence from Twenty-Four Devel- Logistics Strategy: Measuring Progress." Working Paper oping Countries." European Research Workshop in Interna- No. 9, East Asia Transport Unit. World Bank, Washington, tional Trade (ERWIT), Munich, June 14, 2002. D.C. Feenstra, Robert C., Dorsati Madani, Tzu-Han Yang, and Chi- Rafferty, Laura. 2003. "East Asia Ports in their Urban Context." Yuan Liang. 1999. "Testing Endogenous Growth in South Working Paper No. 7, East Asia Transport Unit. World Bank, Korea and Taiwan." Journal of Development Economics 60 Washington, D.C. (2): 317­41. Redding, Stephen, and Anthony, J. Venables. 2002. "Economic Fink, Carsten, Aaditya Mattoo, and Ileana Cristina Neagu. 2002. Geography and International Inequality." Discussion Paper "Trade in International Maritime Services: How Much Does No. 2568. Centre for Economic and Policy Research (CEPR), Policy Matter?" World Bank Economic Review 16: .81­108. London. Frankel, Jeffrey, and David Romer. 1999. "Does Trade Cause Reddy, Ram. 2002. "Friction over Security Gaps." October 8. Growth?" American Economic Review 89 (3): 1­16. Available at www.intelligententerprise.com/. Frenkel, E. G. 1998."China's Maritime Developments." Maritime Romer, Paul. 1994. "New Goods, Old Theory, and the Welfare Policy Management 25 (3): 235­49. Costs of Trade Restrictions." Journal of Development Eco- Gallup, John Luke, and Jeffrey Sachs with Andrew D. Mellinger. nomics 43: 5­38. 1999. "Geography and Economic Growth." International Shaw, Stephen, and Feng Wang. 2002."Moving Goods in China." Regional Science Review 22 (2): 179­232. McKinsey Quarterly. No. 2. Gausch, Luis J., and J. Kogan. 2001. "Inventory in Developing UN ESCAP (United Nations Economic and Social Commission Countries: Level and Determinants, a Red Flag on Competi- for Asia and the Pacific). 2000. "State of the Environment in tiveness and Growth." World Bank, Washington, D.C. Asia and the Pacific." ST/ESCAP/2087. Bangkok. Hillberry, Russell, and Christine McDaniel. 2002. "A Decompo- ------. 2002."Review of Developments in Transport and Com- sition of North American Trade Growth since NAFTA." munications in the ESCAP Region, 1996­2001." Bangkok. International Economic Review (May/June): 1­6. Vitasa, H.R., and N. Seprato. 1999. "Maritime Sector Develop- Hummels, David. 1999. "Towards a Geography of Trade Costs." ments in ASEAN Countries." Paper presented at United Graduate School of Business, University of Chicago. Processed. Nations Development Programme (UNDP) Conference on Hummels, David, and Peter Klenow. 2001. "The Variety and Trade and Development, Jakarta, October. Quality of a Nation's Trade." NBER Working Paper No. 8712. Wei, Shang-Jin, and Yi Wu. 2001. "Globalization and Inequal- National Bureau of Economic Research, Cambridge, Mass. ity: Evidence from China." NBER Working Paper No. Infrastructure Consulting Ltd.2002."Mongolia Trade Competitive- 8611. National Bureau of Economic Research, Cam- ness."Study prepared for the World Bank, Washington, D.C. bridge, Mass. International Trade Institute of Singapore. 2002. "Logistics Wilson, John S., Catherine Mann, Yuen Pau Woo, Nizar Assanie, Development, Trade Facilitation and Its Impact on Poverty and Inbom Choi. 2002. "Trade Facilitation: A Development Reduction in China's Lagging Provinces." Study prepared for Perspective in the Asia Pacific Region." Working paper pre- the World Bank, Washington, D.C. sented to APEC (Asia-Pacific Economic Cooperation). Limao, Nuno and Anthony Venables. 2001."Infrastructure, Geo- World Bank, Washington, D.C. graphical Disadvantage, Transport Costs and Trade." World World Bank. 1994. World Development Report 1994: Infrastruc- Bank Economic Review 15: 51­479. ture for Development. Washington, D.C.: Oxford University Nomura Research Institute. 2002. "Vietnam Logistics Develop- Press and World Bank. ment and Trade Facilitation." Study prepared for the World World Economic Forum. 2002. Global Competitiveness Report Bank, Washington, D.C. 2001­2002. Geneva. 6 Protecting Industrial Inventions, Authors' Rights, and Traditional Knowledge: Relevance, Lessons, and Unresolved Issues Manjula Luthria Keith E. Maskus In recent years, the increasingly intensive use of tives for innovation and the diffusion of new infor- knowledge and innovation in international trade mation into consumption and competition. The flows and foreign direct investment has brought the interests of innovative businesses need to be deli- issue of intellectual property rights (IPRs) into the cately balanced with those of the users of new mainstream in global trade discussions.1 The debate products and technologies.2 It is possible to estab- leading up to the Agreement on Trade-Related lish intellectual property standards that are too Aspects of Intellectual Property Rights (TRIPS), weak to support local innovation and product adopted by Members of the World Trade Organiza- introduction, but it is also possible to adopt stan- tion (WTO) in 1995, was a contentious one between dards that are excessively protectionist, limiting the developed countries, which were lobbying for access that consumers and rivals have to new tech- stronger intellectual property rights, and developing nologies. Both of these problems can limit a coun- countries, which largely considered such across-the- try's long-term economic growth prospects. It is board upgrading to be premature, given their posi- evident that the interests of countries that are pri- tions as users rather than producers of technological marily technology importers and adapters could innovations. Subsequent and ongoing negotiations differ significantly from those of the more industri- at the global and bilateral levels are attempting to alized developing economies, which are moving address some of these tensions. rapidly into technology development and exports. Intellectual property rights are a form of domes- To assist new adopters of IPR protection, this tic business regulation that sets background incen- chapter discusses national case studies of patent This chapter draws on background reports prepared by Jaeyong Song, Yonshei University, Seoul, Korea, and Timothy Swanson, University College, London, UK. The authors gratefully acknowledge helpful comments from Myrna Austria and other participants in the seminar held at the Institute for Southeast Asian Studies of the National University of Sin- gapore and in the Fourth Asia Development Forum in Seoul as well as seminars held at the Thailand Development Research Institute in Bangkok and the World Bank office in Jakarta. 95 96 East Asia Integrates and copyright protection and also looks closely at Did Korea's Stronger Patent Laws the issues involved in the protection of traditional Contribute to Its Success in knowledge. The idea is to examine how these three Foreign Patenting? forms of intellectual property protection are being implemented in countries of widely different devel- Governments often strengthen IPRs in the hope opment levels and technological capacities. We that by establishing exclusive rights to use and sell then attempt to draw basic conclusions about the newly developed technologies they will promote development aspects of different forms of IPRs investments in knowledge creation and business across countries. innovation, including by foreign investors who will The first case study is retrospective: it considers introduce advanced technologies. A rise in innova- the impacts of past patent reforms in the Republic tive activity, or in the amount of productive research of Korea, an advanced developing economy. Patents and development (R&D), signals an acceleration of have been studied in the literature on the econom- technological change and a positive outlook for pro- ics of innovation for nearly four decades, and the ductivity growth. If increases in patenting, and in repercussions of patent protection are generally the innovation that gives rise to them, can be attrib- well understood. Patents are usually of relevance uted entirely to changes in IPR laws, there are for countries importing or creating cutting-edge important implications for the design of public pol- technologies and innovations, and they are becom- icy. Specifically, policymakers need to better under- ing relevant more broadly because the TRIPs agree- stand the complementary roles of broader public ment mandates stronger patent protection in all its policies that promote innovation and of private cor- signatory nations. porate sector strategies in facilitating technological The second case study is contemporary: it sum- catch-up. To shed light on these issues, we examine marizes the results of a survey of copyright-sensi- the determinants of Korea's dramatic success in tive firms in Indonesia, a lower-income developing patenting in the United States. country that is in the process of upgrading its pro- The Republic of Korea is a premier example of a tection. Copyrights have been studied much less nation that has achieved rapid structural transfor- than patents, but they have recently taken center mation and industrialization in recent decades. As stage as economic incentives in areas such as soft- such, it illustrates an economy that was poised to ware and the entertainment industries.3 Copyrights achieve dynamic gains from stronger protection at affecting the software, publishing, film, and music the time of its major policy regime shift. In a series industries are relevant to a broad range of middle- of new laws enacted from 1985 to 1992, the govern- to low-income nations. ment expanded patent coverage, increased the The final case study is prospective: it analyzes scope of claims, lengthened the duration of protec- the difficulties poor countries will face in develop- tion, and made other changes. Although there was ing appropriate regimes for protecting traditional significant opposition to these policies, which knowledge. This knowledge is concentrated in seemed to many to have been imposed externally lower-income nations, and its protection is gener- via pressure from the United States, some industrial ally expected to have direct effects on poverty concerns had already established competence in reduction.4 As some industrial countries attempt research and product development.5 Thus Korea is to patent the products they have developed based a good candidate for a case study of a country with on traditional knowledge, developing nations per- an industrial structure and resource base that ceive themselves as doubly disadvantaged--their would be able to react positively overall to stronger indigenous resources are being used without com- domestic patent protection. Our interest is in ascer- pensation and the resulting products are coming to taining whether the patent regime change itself them at a higher price because of the associated might be responsible for rising rates of innovation patents. Protection of traditional knowledge is a as measured by Korean patent applications in a relatively new issue, and the policy dialogue would major foreign market, the United States. benefit from more analytical work to improve Before beginning this analysis, we note that understanding of the conceptual issues underlying other factors entered into Korea's ability to expand this debate. its innovation stream. For example, an important Protecting Industrial Inventions, Authors' Rights, and Traditional Knowledge 97 element was Korea's success in educating and train- by Korea may have increased the propensity of ing rapidly rising numbers of technical personnel inventors to seek patent protection. for deployment in industry during this period. Between 1980 and 1996, the number of researchers The supply side factors are per million inhabitants in the country rose from 484 to 2,193.6 Such factors are discussed further in · Industrial upgrading push. In this view, upgrad- the next section. ing of Korean industry shifted the output mix to more patent-prone, high-technology industries. · Corporate management push. Large Korean busi- Factors Contributing to the Rise in ness groups (chaebols) placed ever more empha- Korean Patenting sis on innovative capabilities, and they improved The number of U.S. patents filed by Korean firms R&D management and productivity. This view has risen dramatically in recent years (Figure 6.1). postulates that changes in the corporate man- Patenting activity might rise because of either an agement of R&D have contributed to increased increase in the propensity to patent or an increase in patenting activity. the ability to patent. The former could be framed as the demand side view and the latter as the supply The rest of this section examines each factor in side view. Four factors might explain Korean patent- turn. The analysis relies on aggregate data on inter- ing success. On the demand side, the factors are national patent applications by technology class and assignee of patents in the United States; patent · Fertile technology pull. In this view, patenting was citation data of Korean firms; and aggregate mea- spurred by a global surge in discovery and inno- sures of research effort. We also use information vation in emerging, "patent-intensive" industries from in-depth, firm-level interviews. such as semiconductors and biotechnology.7 · Friendly court pull. By making courts more Fertile Technology Pull. If a surge in innovation is prone to grant and enforce patents, the stronger driven by breakthroughs in specific technologies, intellectual property protection regime adopted we should expect to see an uneven increase in FIGURE 6.1 Comparison of Patenting Trends, Republic of Korea and Major Developing Countries, 1978­99 (Registration counts in U.S. Patent Office) Patents 4,000 3,500 3,000 2,500 2,000 1,500 1,000 500 0 1978 1980 1982 1984 1986 1988 1990 1992 1994 1996 1998 Korea Taiwan India Brazil Singapore Source: U.S. Patent Office. 98 East Asia Integrates patenting across technologies. When the techno- Is Korea simply enjoying the benefits of a world- logical revolution is more widespread, we should wide technology explosion in semiconductors? A expect a general increase in patenting activity comparison with other industrial and industrializ- worldwide. Korean semiconductor patents, in par- ing countries suggests instead that Korea is an out- ticular, show a growth pattern similar to the general lier (Figure 6.3). increase in patenting activities in the semiconduc- Pair-wise t-tests support this conclusion, show- tor industry worldwide (Figure 6.2). ing that Korea's semiconductor patents have prolif- FIGURE 6.2 Comparison of Korean Semiconductor Patents with Worldwide Semiconductor Patents, 1976­99 (Registration counts in U.S. Patent Office) World Korea 18,000 1,000 16,000 900 14,000 800 700 12,000 600 10,000 500 8,000 400 6,000 300 4,000 200 2,000 100 0 0 1976 1978 1980 1982 1984 1986 1988 1990 1992 1994 1996 1998 World Korea Source: U.S. Patent Office. FIGURE 6.3 Comparison of Korean Semiconductor Patents with Semiconductor Patents of Major Advanced Countries, 1976­99 (Registration counts in U.S. Patent Office) Patents 1,000 900 800 700 600 500 400 300 200 100 0 1976 1978 1980 1982 1984 1986 1988 1990 1992 1994 1996 1998 Korea United Kingdom France Germany Source: I.S. Patent Office. Protecting Industrial Inventions, Authors' Rights, and Traditional Knowledge 99 erated much faster than those of advanced coun- between 1983 and 1991 Korea's patent growth rate tries such as Germany, Japan, and the United States. was significantly faster than that of either Japan or However, there is no statistically significant differ- Germany.10 ence between the growth rates of semiconductor patents in Korea and Taiwan (China).8 Industrial Upgrading Push. Did improvements in Korean industry shift the output mix to more Friendly Court Pull. What was the role of intellec- patent-intensive industries? If so, the industrial tual property protection regimes in patenting activ- composition of Korean patents in both the United ities at home and abroad? We examine important States and Korea should more closely match the changes in IPR protection within Korea to under- industrial composition of Korean exports to the stand whether it increased access to foreign knowl- United States than the industrial composition of edge at home (through foreign patents) and pro- U.S. patents at large. Analysis shows that the indus- vided a testing ground for domestic innovators trial composition of Korean patents in the United before they attempted patenting abroad. If Korean States correlates very highly with that of Korean courts become more patent-friendly, both foreign exports but not with that of overall U.S. patents. and domestic firms should find patenting in Korea This finding strongly suggests that Korean patent- increasingly attractive, and thus patent registration ing in the United States has been influenced more data should be relatively uniform across both for- by Korea's industrial upgrading than by specific eign and native residents. industry-level propensities for patenting.11 Three major changes have affected Korean IPRs The highly significant positive correlation since the 1970s:9 between patenting and exporting suggests that industries that develop strong innovative capabili- · 1980­82: Revision of the patent law in 1980 in ties (reflected in subsequent patenting) tend to compliance with the Paris Convention for the achieve strong revealed comparative advantages Protection of Industrial Property and revision of and thereby to export more. At the same time, the patent law in 1982 in compliance with the export-driven industries tend to invest more in Patent Cooperation Treaty innovative R&D activities and file for patents · 1986: Revision of the patent law to introduce abroad more aggressively, both to build and main- substance (product) patents for pharmaceutical tain their global competitiveness and to protect and chemical materials their newly developed technologies and products in · 1995­97: Revision of the patent law in 1995 in their main export markets. Thus export-driven compliance with TRIPs as well as to reform the industrial upgrading and patenting through build- appeals/trials system. The patent law was revised ing innovative capabilities seem to be mutually again in 1997 to introduce opposition to the reinforcing trends, even though we cannot infer granting of patents after registration. cause and effect here. Trends in foreign and total patent applications Corporate Management Push. Increases in in Korea seem to support the view that IPR changes patenting activity may have been caused by have spurred increases in patenting. Foreign patent increases in R&D and its productivity. To test this applications in Korea had two major inflection proposition, we compared the R&D productivity of points in the 1980s. The first jump took place in Korean firms with that of selected advanced coun- 1983 just after the major changes in Korean IPRs in tries, using the number of U.S. patents per US$1 1980­82. The second jump occurred in 1986­87, billion12 spent by each country on R&D. We just after the introduction of substance (product) allowed for a five-year time lag between R&D patent eligibility for pharmaceuticals and chemi- investments and patent registrations.13 Research cals. In the 1980s, when the major IPR changes productivity was lower in Korea than in Japan and were made in Korea, patent applications by foreign- the United States, but substantially higher than in ers rose faster than those by local residents, Germany, the United Kingdom, or France (Table although patent applications by the latter also 6.1). U.S. firms have a clear home advantage in increased steadily. Paired t-tests confirm that patenting in the United States, and the other four 100 East Asia Integrates TABLE 6.1 Research Productivity of Selected Countries U.S. patent counts R&D expenditures U.S. patent count/ in 1999 (billions of 1994 US$) R&D expenditures Korea, Rep. of 3,529 12.8 275.7 USA 75,014 169.3 443.1 Japan 32,666 75.1 435.0 Germany 8,121 37.3 217.7 France 3,150 26.5 118.9 UK 2,351 21.7 108.3 Sources: Patent counts from our U.S. patent database; R&D expenditures from OECD (1999). countries account for about 70 percent of total U.S. In 1997 the private sector accounted for 77 per- patents of foreign origin. Thus Korea's research cent of Korea's R&D expenditures, and since then, productivity, measured by U.S. patent counts per private corporations' share of R&D expenditures amount of R&D expenditure, seems to be the sec- has continued to rise. In 2000 they increased their ond highest in the world, just behind Japan's.14 R&D expenditures by 20 percent, while government Private corporations, particularly the five largest research institutes and universities increased theirs chaebols, accounted for most of the Korean patents by only 9 percent and 3 percent, respectively. Private granted in the United States in the 1990s.15 The R&D laboratories proliferated, especially in the late increased productivity of R&D in Korea has essen- 1990s, as the government's policy of nurturing tially been driven by the good R&D management high-technology start-up firms began to pay off.18 and productivity of the five largest chaebols, espe- The chaebols were also able to hire the best R&D cially the Samsung Group. The largest chaebols researchers. After they began investing heavily in increased their R&D expenditures dramatically in the high-tech industries, they aggressively sought to the 1980s, and in that decade Korea's investment in lure experienced ethnic Korean engineers from private sector R&D per unit of gross domestic major companies and universities abroad, espe- investment was higher than that of any other coun- cially in the United States (Song, Almeida, and Wu try.16 In 2000 the four largest chaebols--Samsung, 2002). The returned engineers played an important LG, Hyundai, and SK--spent 4.731 trillion won role in speeding up the learning process of Korean (about $4 billion) on R&D, and in 2001 their R&D companies and enhancing the productivity of their spending continued to increase, even in the wake of research. Koreans residing abroad also served in the global recession.17 R&D posts for Korean companies and as coun- Over time, R&D investments have become more selors to the Korean government in helping to concentrated in the top 20 companies. According to shape the direction of policy support in Korea. the Korean Ministry of Science and Technology, the top 20 companies in Korea (all industries) Technology Trajectory accounted for 49.8 percent of R&D spending in 1992 and 51.7 percent in 1993, and that percentage How do Korea's firms compare with the world's climbed steadily to reach 56.5 percent in 1996 and leaders in the technology fields in which Korea has 65.8 percent in 1998. The top five corporate large numbers of patents? And why has Korea's pri- spenders on R&D accounted for 31 percent of vate sector been so successful in one fast-growing Korea's R&D spending in 1993. In manufacturing, technology field, semiconductors, but not in Korea's top 20 spenders on R&D undertook 64 per- another, biotechnology, where world patenting is cent of research spending in 1996. This concentra- rising dramatically? tion is significantly higher than comparable figures for the United States, at 43 percent, and Japan, at 44 Innovation Performance: Korean Firms and percent (Korea Ministry of Science and Technol- Worldwide Leaders. All new claims for patenting ogy, 1998). must contain a reference to a previous invention on Protecting Industrial Inventions, Authors' Rights, and Traditional Knowledge 101 which the current invention builds. This is known general, substantially behind the leading global as a citation. Patent citation trends were examined companies in terms of numbers of both patents to see whether Korean companies have established and self-citation rates.23 The exception is in semi- their own country-specific or firm-specific techno- conductors, where Samsung Electronics has logical paths.19 Among U.S. patents granted to emerged as one of the leading innovators in the Korean private corporations, the ratio of Korean industry worldwide. patent citations to total patent citations has increased steadily, reaching almost 8 percent by Why Semiconductors and Not Pharmaceuticals 1999 (Figure 6.4). The ratio for U.S. patents granted or Chemicals? Compared with its performance in to government research institutes (GRIs) is sub- the semiconductor industry, Korea's innovation stantially lower. This result may indicate that pri- performance in chemicals and pharmaceuticals has vate corporations have developed better technolog- been particularly poor (Figure 6.6). Figures on ical bases or more solid technological paths over revealed comparative advantage confirm the dis- time than government research institutes. parity between the semiconductor and chemical Next, trends in self-citations were traced to mea- industries in Korea.24 sure the path dependence of firms and countries in Interviews with senior researchers and R&D innovations.20 Self-citation occurs when a patent filed managers in these industries revealed that the elec- by an organization (or a country) cites another patent tronics/semiconductor industry is a more engineer- from the same organization (or country).21 Generally, ing-driven, assembly-oriented industry than either the self-citation rate is high for a firm or a country that the chemical or pharmaceutical industries, both of has developed strong innovative capabilities over which are more science-driven. In electronics/semi- time. The self-citation rate in patent applications by conductors, technical progress has been very fast, private corporations has risen steadily over time (Fig- and companies have often experienced major tech- ure 6.5). By the late 1990s, half of the patent citations nological discontinuities that offered good entry by Korean private firms were self-citations. opportunities for latecomers like Korea.25 The patenting trends of representative Korean By contrast, companies in the chemical and companies were compared with those of leading pharmaceutical industries must master cumulative overseas companies in several individual indus- know-how that is acquired mainly by time-con- tries.22 The leading Korean companies are still, in suming trial-and-error learning--a particularly FIGURE 6.4 Citations of Korean Patents as a Proportion of Total Patent Citations, by Types of Patent Applicants, 1986­99 Ratio (percent) 10 9 8 7 6 5 4 3 2 1 0 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 Private corporation Government research institute Source: U.S. Patent Office. 102 East Asia Integrates FIGURE 6.5 Self-Citation Rates of Korean Private Corporations, 1980­99 Ratio (percent) 6 5 4 3 2 1 0 1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 Source: U.S. Patent Office. FIGURE 6.6 Korean Patents Granted in the United States, Selected Industries, 1981­99 Patents 2,500 2,000 1,500 1,000 500 0 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 Year Chemical Electronics Machinery Semiconductor Pharmaceutical Biotechnology Source: U.S. Patent Office. slow process for a country with a relatively short giants for global clinical trials, drug approvals, and period of industrial experience. Even in firms that worldwide distribution. are industry leaders, new drugs take several years to Korea's government and companies placed their develop, entail huge costs, and come with a high bets on electronics and information technology, chance of failure.26 And--given the heavily regu- which they perceived as easier industries in which to lated nature of the drug approval and distribution catch up technologically. As a matter of industrial processes--even if Korean companies develop policy, both the investments and the number of potentially valuable drugs or new chemical entities, researchers in the chemical, pharmaceutical, and they must rely on the multinational pharmaceutical biotechnology industries have been small compared Protecting Industrial Inventions, Authors' Rights, and Traditional Knowledge 103 with those in the electronics/semiconductor indus- from field interviews and surveys of copyright-sen- try.27 Historical industrial policy in Japan and Korea sitive firms, we draw some general lessons for the have favored engineering-based industries as design of copyright policy.28 opposed to science-based ones, presumably because From the standpoint of economic development, of the greater uncertainties associated with the latter the main reasons for a developing country to adopt type of research. An additional factor may be that and enforce stronger copyright laws are to encour- Korea's close proximity to and trade relationships age creative activity by local artists and firms and to with Japan fostered the expansion of Korea's elec- support the transformation of that activity into tronics sectors through the use of Japan's OEM products for the domestic and export markets. (original equipment manufacturer) technologies. Stronger copyright regimes can generate significant income streams for creative people and firms. Yet at the same time these new rights can be expected to Conclusions raise the prices of copyrighted goods by reducing Korea's success in innovation has required much the supply of pirated versions. The higher prices more than a tightening of intellectual property will hurt consumers, although they benefit from rights. Industrial upgrading, exports, a big push in having higher-quality products and support ser- R&D from the chaebols, and selective targeting of vices on the market. Stronger copyright regimes the electronics/semiconductor industry have all also may also hinder the progress of education and played a role. scientific and technical advance. Copyright Korean firms have become much more innova- strengthening will likely cut the sales of imitation tive in the short span of a decade, but the self-cita- products over time, and, to the extent that imitators tion patterns of patents registered by Korean firms and their employees are domestic, the impact may reveal that most have yet to attain the status of be higher unemployment and adjustment costs in global leaders in technology. The focus thus far in the labor market. Korea has been on inventing processes and not on To the extent that workers in the pirating firms developing generic technologies. As the innovation are poor, stronger copyrights can worsen poverty, process becomes more sophisticated, it may at least in the short term. In many developing become more difficult for Korea to compete suc- nations, fear of the short-term effects of stronger cessfully in broader-based inventions. Thus the copyright laws on employment and prices is wide- next stage of its R&D effort will likely have to focus spread. For stronger copyrights to reduce poverty, on inventing standard-setting technologies that can they must open new economic opportunities for earn substantial royalties in international markets. poor artists and innovators, including those in the To comply with the TRIPs agreement, many software, film, music, and print publishing indus- countries will have to introduce stronger patent tries, to develop, record, and sell their products. protection. Given Korea's experience, other coun- Furthermore, mechanisms must be in place to tries wishing to boost innovative activity among ensure that these artists and innovators realize their private firms should not expect too much higher incomes from the commercialization of from simply adopting a new patent regime, and their works. they may wish to design packages of complemen- On this subject, Indonesia is an interesting tary "behind-the-border" policies. Korea has used a country to study for two reasons. First, the citizens variety of instruments for this purpose, many of of this lower-income economy may have strong which could be emulated by developing countries. preferences for unimpeded abilities to purchase unauthorized copies of software and entertainment products. Second, Indonesia has a wealth of cul- Will Stronger Copyright Protection tural interests, particularly in music, and a growing Encourage the Development of Copyright-Sensitive Businesses in base of intermediate skills. Thus considerable scope Indonesia? exists for developing local copyright sectors, which would prefer somewhat stronger protection. A con- This section examines the effects of copyright law flict may therefore exist in perceived interests, rais- on business development in Indonesia. Using data ing interesting policy trade-offs. 104 East Asia Integrates Legal Structure and Enforcement governments may have other priorities or inter- ests in weak enforcement. On paper, Indonesia has a strong copyright regime, and one Indonesian official has declared that, of the Despite these limitations, the government is countries in Southeast Asia, Indonesia has the making some progress in combating piracy. A new strongest degree of compliance with the WTO law requires registration of the pressing machinery TRIPs agreement. Indeed, Indonesia's 1997 copy- for compact discs, digital video discs, and video right law was virtually in compliance with that CDs--all of which is imported. The number of agreement. It was replaced by a new copyright law copyright enforcement cases increased from 44 in in early 2003, which established a regime of neigh- 1999 to 109 in 2001 and to 1,999 by early June boring rights (consistent with the Rome Conven- 2002.30 Virtually all such cases involve pirated tion) and significantly increased the civil and crim- movies and music, with enforcement actions inal penalties for infringement.29 against both production facilities and retail shops. As might be expected, the gap between legisla- At the time of this writing, no copyright cases tion and actual protection is significant, because have involved software copying, probably because Indonesia's capacity to enforce the copyright law is of a significant limitation in the law. To sustain a relatively limited for several reasons: claim of infringement, complainants must provide original software code to the courts so that the · The piracy, largely of recorded movies and authorities can compare the code with that of the music, is extensive. The incentives to copy and defendant. Many software companies prefer not to distribute these and similar goods are enor- reveal their machine language in an environment in mous, and piracy has become such a significant which it may be released to rivals. However, in two industry that even a well-funded enforcement cases in 2001 the local agent for a major foreign agency would find it difficult to make much of a software firm sued five firms for having loaded the dent in its scope. firm's products into personal computers without · Funding for the police is extremely low. Accord- authorization. The agent was able to do so under ing to one police official in Jakarta, his depart- Indonesia's consumer fraud statutes and achieved ment's budget for copyright enforcement is judgments against all defendants.31 about $200 per action, but each case costs per- haps $1,000­$1,500. The police therefore accept Effects of Copyright Law on Specific Industries payments from complainants (for example, music recording companies) to defray their This section describes particular copyright-sensi- costs. This practice limits the incentives for legit- tive industries in Indonesia and the changes they imate complaints and is questionable from the might experience in the event of stronger copyright standpoint of honest and efficient enforcement. enforcement.32 Across the board, survey respon- · Very few customs authorities and police person- dents overwhelmingly preferred stronger govern- nel work on IPR issues, because these officials ment efforts to reduce piracy. This finding is not may have higher priorities. The copyright office surprising, because it would remove some costs for itself devotes some effort to making raids and these industries while expanding demand for their raising awareness, but it is seriously under- products. staffed. And, like many developing countries, Indonesia has far fewer qualified judges, prose- Software. In Indonesia, software and databases are cutors, and IPR lawyers than it needs for effec- protected by copyrights providing protection for 50 tive enforcement. Computerized systems are vir- years, and computer programs may be patented. tually nonexistent. According to at least two Indonesian officials, the · Centralized authorities in Jakarta and Bandung government views copyrights as an inducement to find it impossible to undertake enforcement innovation in software development, an industry it activities in Indonesia's widely flung urban and has identified as central to economic growth and rural areas and coastlines. They rely on local recovery. This decision is based on the fact that governments to manage the problem, and those writing code for programs and games is fairly Protecting Industrial Inventions, Authors' Rights, and Traditional Knowledge 105 straightforward and labor-intensive, offering try in relation to Indonesia's economy and popula- prospects for absorbing large numbers of reason- tion, stronger copyright protection should ably well-educated young people into the labor improve the software industry's prospects to some force. Indeed, as several interviewees commented, degree.33 However, the industry is unlikely to grow Indonesia has a good base of semiskilled workers significantly before the economy overcomes its for the software industry. Software is viewed as a current sources of instability and uncertainty and dynamic industry that should experience rising improves its infrastructure for telecommunica- global and domestic demand over the long term, tions services. particularly as the Indonesian economy makes greater use of information technologies. The gov- Recorded Music. Among the copyright-sensitive ernment has established incubator programs that industries in Indonesia, the recorded music indus- provide incentives for software developers. The try probably suffers the most damage from poor programs are too new to assess, but they seem to be copyright enforcement. The industry is largely well structured for this purpose. devoted to local cultural and pop music and the The software industry has considerable scope domestic audience. It is made up mainly of small for expansion in both the domestic and export firms, and, as in software, they are remarkably few markets. Copyright protection will be one element in number for an economy of Indonesia's size. The in realizing this potential. In this relatively young current system yields revenues for a very small industry, firms perceive the threat of software number of musicians. A more sophisticated system, piracy as a significant entry barrier. Other major including professional rights management and col- entry barriers cited are strong competition, short- lection societies, could help to improve this situa- ages of skilled labor, existence of dominant produc- tion considerably. The primary prospect for ers, and concern about weak copyrights in neigh- expanding income opportunities for poor artists boring countries. For firms considering expansion, lies in the expansion the recording companies say the most significant problems are a weak telecom- they would undertake in response to better copy- munications infrastructure and Internet services, right protection; the companies also would invest shortages of skilled labor, and concern about unau- more in artist development. thorized copying. Of the 25 software firms sur- Virtually all the recording companies said the veyed, 24 indicated an intention to expand output government is doing "far too little" to stop piracy, over the next two years. Twenty-two firms cited and claimed unambiguously that their businesses uncertainty about the economy as a "very impor- would grow considerably if copyright enforcement tant" or "somewhat important" barrier to expan- were more transparent and effective. Both large sion. At least over the medium term, better copy- firms, all 11 of the medium-size firms, and 9 of the right enforcement seems to offer only moderate 11 small firms said they would expand, either potential for encouraging additional business activ- modestly or significantly. Ten of the small firms ity in software development in Indonesia. and all the medium-size firms would invest more Virtually all the survey respondents found in developing new songwriters and recording piracy endemic and would like to see stronger artists. Again, given Indonesia's size, there is reason efforts to reduce it. Of the small firms surveyed, to believe that expansion now will pay off consid- only one said it would expand its output and erably in the long run. investment if rights were strengthened. Medium- size enterprises would be more likely to expand Film. Indonesia's film industry is small but grow- their activity and develop new products, as would ing. Film producers with small companies are par- larger enterprises. Presumably, that expansion ticularly concerned about domestic copying and would increase net incomes for software developers could benefit from stronger copyright enforce- and employees in the industry, even as it reduces ment. Like those in the recorded music business, employment opportunities in copying. the great majority of firms said the government is It is possible to be more optimistic about the doing far too little to reduce piracy. However, longer term. Given the country's human capital stronger copyright enforcement is not expected to resources and the small size of the software indus- have as dramatic an impact on film as on music. 106 East Asia Integrates Among the film companies, 9 of 16 said they would it could encourage development of a larger base of attempt to expand modestly or significantly, while technological and creative personnel through edu- 6 of the small firms anticipated no impact or a cation and training incentives. Finally, to the extent small contraction. At the same time, 14 of 16 firms that other barriers to entry in these sectors exist, (including 6 small firms) said they would invest in they could be diminished. more filmmaking capacity in the event of stronger copyrights. Thus, although the signals are some- What Are the Problems in what mixed, there does seem to be a reasonable Protecting Traditional Knowledge indication that the film industry would expand as in Genetic Resources? copyrights become more strongly enforced. Genetic resources, including plants and herbs, and Print Publishing. Indonesia's publishers believe understanding of their properties for nutrition and book piracy is fairly widespread, but their industry medicines are known collectively as biologically is suffering only modest damage and would not based traditional knowledge. Typically, that knowl- expand much as the result of stronger protection. edge derives from generations of collective and They believe that the major problem in starting a experimental use of naturally occurring plants and print publishing business in Indonesia is a shortage soils. This section addresses the problems encoun- of financial backing. tered in converting such knowledge into mar- ketable products. Other forms of traditional knowledge arise in Conclusions music, designs, and folklore. For example, tradi- Indonesia's copyright law is TRIPs-compliant, and tional designs of carpets and clothing are often in some ways it goes beyond the TRIPs require- shared by numerous villages in particular regions of ments. An assessment of whether the extra protec- developing countries. Although many of the prob- tion is really appropriate for Indonesia's economy lems discussed in this section arise in these cultural might be useful.34 For some time into the future, contexts as well, the scientific and biological basis of enforcement is likely to continue to suffer from genetic resources makes them unique and the center weak capacity, chronic underfunding, limited of current controversy. Thus we focus on issues that access to enforcement procedures, and a lack of emerge particularly in the latter area. transparency. Innovations in agriculture and pharmaceuticals As rights are more successfully enforced, copy- are drawing increasingly on traditional or indige- right-sensitive industries will see opportunities for nous knowledge, and the market value of pharma- expansion and more income. These gains would be ceutical derivatives from traditional medicine has felt in the medium term and would have to be been estimated at about $43 billion a year world- assessed against the short-term hardship that wide (IPBN 1995). The trial and error required to would result for the firms and workers now bring about a successful innovation can be drasti- engaged in unauthorized copying and distributing. cally reduced when researchers consult the practi- The extent of this hardship cannot be predicted tioners of traditional knowledge. In medical with confidence, but it seems probable that research, bio-prospectors have increased the suc- employment in copying and selling pirated ver- cess ratio of clinical trials from 1 in 10,000 samples sions of music, movies, and software is significant to 1 in 2,000 (Prakash 1999). Given that roughly 90 in the larger cities. percent of the world's genetic resources and tradi- If copyright enforcement is likely to remain inef- tional knowledge is held in developing countries, fective for the foreseeable future, what might and roughly an equal percentage of the world's Indonesia do to improve the marketing prospects R&D activity takes place in industrial countries, of local artists, software developers, and publishers? there is scope for mutually beneficial bargains First, the government could accelerate its programs between a gene-rich, technology-poor South and a to improve telecommunications and private Inter- gene-poor, technology-rich North. net services, which initially could be used to raise The proper means for establishing the documen- awareness abroad of Indonesian products. Second, tation, ownership, compensation, and exploitation Protecting Industrial Inventions, Authors' Rights, and Traditional Knowledge 107 of traditional knowledge remain unclear. The TRIPs Much of the new knowledge in the biological agreement touches on these issues, but essentially sciences is acquired through observation of natu- leaves them to interpretation or further negotia- ral phenomena, combined with trial and error-- tions. This is an area in which developing nations as, for example, when farmers observe weather- could benefit from advances in thinking and resistance or disease-resistance in certain crops stronger policy options. This section outlines the and then create hybrids that have the desirable analytical and practical issues that need attention to characteristics. Finding new solutions to problems spur progress on protecting traditional knowledge therefore depends on maintaining a diverse gene or compensating its ownership. For this purpose, it pool, and, in turn, the conservation of biodiversity begins by discussing the complexities in research per se is the key to solving the agricultural and and development in the biological sciences. health problems of tomorrow. Because the destruction of biodiversity is irreversible, preser- vation of the stock of genetically rich lands and A Framework for R&D in the Biological Sciences resources is vital. All technological advances take place in three dis- The economic question at the crux of the debate tinct stages. In the primary stage, known as inven- on protecting traditional knowledge is how to tion, a new idea is observed or occurs. In the sec- ensure that both the primary and secondary stages ondary stage, the idea, by being embedded in a of R&D, and their associated capital inputs, receive product or process, is transformed into an innova- enough compensation to generate incentives for tion. In the tertiary stage, the innovation is diffused. their continuing efficient operation. How is natural Each of these stages has features peculiar to the capital to be valued? How are the returns to innova- biological sciences. In the primary stage, observa- tions to be appropriated and apportioned? tion and experimentation, which must rely on a About half of the yield gains in agriculture can combination of physical, human, and natural capi- be ascribed to traditional genetic resources, accord- tal, are needed to capture exogenous information. ing to estimates from models using the standard In the biological sciences, the dependence on natu- Cobb-Douglas production function. But when ral capital (plants and soils) at this stage is unique, models have attempted to quantify firms' willing- and issues related to the valuation of natural capital ness to pay for such resources, shockingly small val- become central. In the secondary stage, traditional ues have emerged.35 The explanation for these knowledge and natural capital are combined into small values lies in the potential redundancy of physical and human capital. Innovations that are solutions in the biological sciences. The more fully embodied in a physical product can be sold, ephemeral a solution promises to be (and thus the and few externalities exist in appropriating the lower the revenue it promises to yield), the smaller returns to investment. But innovations that are dis- the private investment. However, public health and embodied--that is, not contained in a physical food security concerns would dictate that the less product but rather held as knowledge, say, by a durable the solution, the greater the investment farmer or researcher or by a social system--are needed in R&D. Therefore, some form of public subject to such externalities. A large part of biolog- intervention is warranted that rests on the under- ical innovation is disembodied, and, in particular, standing that resources that are not valuable now traditional knowledge is almost always held within may be valuable later. social systems. The tertiary stage, diffusion of the Proper arrangements for the appropriation of new innovation, also varies in character, depending returns and compensation of contributors are pre- on whether the innovation is embodied or disem- conditions for an optimal R&D effort and supply of bodied. When the innovation is embodied, its dif- the factors needed for innovation. Property rights fusion takes place through the marketing of the provide these preconditions. The literature suggests product, albeit at a cost. The diffusion of disem- that the property rights assignment matters very bodied innovations is more complicated. Because much if contracting within the industry is costly, diffusion occurs mainly through the exchange or and that property rights should be placed at the lev- movement of human capital, it often takes place els that most encourage investment in the assets more efficiently in a nonmarket setting. concerned (Swanson and Goeschl 1999). 108 East Asia Integrates Evolution in R&D Management and that investment was needed to maintain and expand that base. Thus the assumption that The international R&D management regime for property right mechanisms should not extend to genetic resources has evolved through three eras: natural forms of capital was brought into ques- tion. Once it was recognized that information in 1. Equal access/common heritage. This first era the biological sector must be optimally pro- emphasized the accumulation of already avail- duced as well as diffused, it became important to able information and its widest possible diffu- move away from the doctrine of common her- sion. The aggregation of traditional knowledge itage and free access. and genetic resources occurred within national 3. Private investment/property rights in base genetic academic and governmental institutions, and resources. The third era of R&D management is almost wholly by virtue of publicly funded taking a harder look at claiming exclusive rights investments. The emphasis was on the public in the base genetic resources. Both physical and good nature of this information, and rapid dif- human capital are now directed at analyzing and fusion was subsidized by means of the universal manipulating the base resources themselves, recognition of the doctrine of free access to col- with the object of developing genetic character- lected resources. istics that are useful in and of themselves. In this 2. Private investment/property rights in end prod- way, the various forms of capital are all com- ucts based on genetic resources. The second era bined (natural, human, and physical), and the arose with the recognition that the traditional end product becomes the understanding or cre- knowledge base could be developed and ation of a genetically generated characteristic expanded with investments of other forms of that is useful in itself. The property right is then capital. Scientific method and physical capital claimed in the genetic trait and its claimed use. could be combined with natural forms of capi- Thus the third era of R&D management has tal (genetic resources) to create more informa- moved the property rights debate down a level, tion and more useful information. Public to that of the genetic trait rather than the bio- investors were probably the first movers, but logical resource.36 they were joined rapidly by private investors whose involvement required the development A Brief History of Global Policies Affecting of institutions for compensating private invest- Genetic Resources ments. In the pharmaceutical industry, these investments could be compensated through Historically, traditional knowledge and genetic existing patent laws. In agriculture, it was neces- resources were treated under the doctrine of the sary to extend IPR law in a new direction--that common heritage of mankind. They were therefore is, toward granting exclusive rights for the use available without restriction. of naturally occurring organisms. This view changed in the 1930s in the United This second era of R&D management States and Europe, where plant breeders first revolved around developing property rights in sought to protect the results of their selective end products based on genetic resources; the fun- breeding programs. In 1961 the International Con- damental notion that naturally occurring vention for the Protection of New Varieties of genetic resources were a common heritage Plants (UPOV) established a harmonized system remained undisturbed. The problem was seen as for international recognition and establishment of one of compensating other forms of capital plant breeders' rights. (human, physical) that were combined with nat- The Convention on Biodiversity (CBD), which ural capital (genetic resources) to produce inno- came into force in 1992, established the doctrine of vative end products, even when those products national sovereignty over genetic resource inputs, were biological and thus capable of natural abrogating the doctrine of common heritage. It also reproduction. established an approach to property rights in It became increasingly clear, however, that genetic resource inputs based on bilateral benefit- the useful genetic resource base was not static, sharing agreements. This bilateral contract Protecting Industrial Inventions, Authors' Rights, and Traditional Knowledge 109 approach to property rights in genetic resources Is There an Optimal Property Rights Institution? has had little impact to date, however, because of the practical difficulties of implementation. Under To be practicable, a system of property rights in the the terms of the CBD, either the state or the local biological sciences must take into account how this community that hosts the resource can refuse diverse (because it depends on so many different access. If it allows an outside party to have access to forms of capital) and globally dispersed R&D sec- its genetic resources, it can require an agreement tor is managed. An intellectual property rights setting out the terms on which access is given. But regime would seek to facilitate R&D management doing so successfully requires the passage of access by providing some form of exclusive marketing legislation for the resources, as well as a means for rights so that the optimal level of investment is gen- deciding among competing rights of ownership erated in various parts of the industry. and establishing the mechanism for consent and Broadly, the options for IPRs are the following: transfer. Few states have adopted and implemented property rights only at the primary stage of R&D; the access legislation required for this task. It is also property rights only at the secondary stage; or prop- clear that establishing such claims and providing erty rights at both the primary and secondary stages. legal consent would be difficult. We examine the pros and cons of each in turn. By contrast, the TRIPs agreement attempts a global standardization of IPR regimes. In particu- 1. Property rights only at the primary stage of R&D. lar, the agreement requires WTO member states to This option would be inadequate. Property right adopt some form of plant breeders' rights, while it institutions could not pertain only to the pri- specifically permits countries to exclude from mary stage of R&D--that is, at the level of the patentability all life forms emanating from essen- products of the natural capital--because it tially biological processes. The wording of the would be impossible to capture the benefits TRIPs agreement requires all member states of the emanating from the other (non-natural) forms WTO to adopt either some form of patent system of capital employed within the sector. The latter for improved plant varieties or a sui generis regime are usually employed at a later stage in the for the protection of the same. Pressure has industry, and the property right pertaining to mounted for the adoption of uniform systems of the first stage would be incapable of managing IPR for plant varieties and other genetic resources the end markets. Also, it is not clear that a prop- (microorganisms) and for countries' mutual recog- erty rights mechanism at this level would dra- nition of the rights they confer for such resources. matically increase the levels of investment in The TRIPs agreement is essentially silent on the reserves. The mechanism selected would need to issues of traditional knowledge. operate at several different levels of knowledge, Thus CBD and the TRIPs agreement are in con- (embodied, disembodied, expressed) in order to flict. CBD emphasizes the importance of sovereign create and share profits among suppliers of tra- rights in indigenous genetic resources and the ditional knowledge. This scenario might be pos- knowledge associated with them. The TRIPs agree- sible, but at the cost of introducing distortions ment requires the adoption of IPR systems for (for example, exclusive dealing arrangements) improved plant varieties, but not for resources that would introduce their own costs. emanating from essentially biological processes. 2. Property rights only at the secondary stage of Reconciliation of these two approaches would R&D. The sector could be managed with a single require, for genetic resources, some manner of property right at the secondary stage of R&D, recognition of rights to innovations, both at the but it would require substantial amounts of con- industry end and potentially for some of the base tracting between the industry developing end- genetic resources used as inputs. Such a regime use products and the providers of traditional would incorporate the rights of states that provide knowledge and genetic resources. In essence, a the natural inputs (as required under the CBD) and single property right at the secondary stage also those that provide the final outputs (as would permit its owner to appropriate the full required under the TRIPs agreement). But how value of the information inherent in the final might such a regime operate? product, but complete contracting would be 110 East Asia Integrates needed to ensure that these rents are distributed practices might be supported if it were possible to efficiently across all components of the vertical provide written documentation of local and tradi- industry. Given the diverse and widely distrib- tional practices, but traditional knowledge has usu- uted nature of the R&D sector, it would be costly ally been passed on orally. to engage in this sort of structured contracting. A third problem concerns the description of tra- 3. Property rights at both the primary and secondary ditional knowledge and ownership claims made stages of R&D. R&D could also be managed with from it in a written application. Specifications for a the placement of a property right mechanism at patent must be written in technical language that both the primary and the secondary stages. Such demarcates its limits and establishes the relation- an approach would create an institution focused ship of the innovation to existing knowledge and on the efficient distribution of the industry's innovations. A particular use of a characteristic of a profits, but at the cost of creating multiple levels plant or animal may be well known to an indige- of monopoly within the same vertical industry. nous community, but it must be stated in a form Multiple constrictions on output (to generate that distinguishes it from other biochemical or rents) would reduce the welfare generated genetic information or usefulness. Much tradi- within the industry. Successive monopolies tional knowledge is disembodied and subject to the would introduce successive distortions within problems of appropriating benefits from disem- the industry, which might result in more welfare bodied information. reduction than enhancement. Finally, to be protected from unauthorized use, traditional knowledge must first be placed in public The choice between the second and third registries. In principle, such registries would estab- options depends on their comparative institutional lish the sources, ownership, and characteristics of costliness. Having a single property right at the sec- the included knowledge. In practice, the operation ondary stage (option 2) would result in substantial of registration systems poses at least two complex contracting costs and thus significant residual inef- questions: ficiency in the distribution of rents. Having multi- ple property rights in the single vertical industry · How would traditional knowledge be registered? (option 3) would address this problem, but it Before any valuation or distribution issues are would result in successive distortions in the market. tackled, traditional knowledge must be docu- mented. Should the burden be on local commu- nities to disclose and register what they know in Practical Concerns the event that this knowledge becomes precious Even if the conceptual and analytical issues just to the scientific community in the future? outlined were somehow addressed, some practical Would communities that do not do so be penal- concerns would remain about the valuation, appro- ized? priation, and distribution of returns from tradi- · How would such registers operate? To establish a tional knowledge. claim of prior possession and therefore preclude First, intellectual property rights are individual- patenting by others, these registers would have istic, but traditional knowledge is normally held to be made public. Putting them in the public collectively and developed by members of an domain could prejudice future protection of the indigenous community over several generations, knowledge documented in them. making it extremely difficult to identify individual inventors of such knowledge. Although the fact that To illustrate these problems and to understand the knowledge is passed down over generations one case in which they have been addressed, con- should, in principle, add to its value, the problem sider recent legislation in the Philippines (Box 6.1). arises that the information may be said to exist This nation, which is composed of many indige- within the public domain. nous peoples, has been a leader in instituting legis- Second, patent applicants must supply evidence lation for access and compensation of indigenous of a single act of discovery. This would be extremely knowledge. Indeed, the country is viewed as a labo- difficult to establish. The patenting of indigenous ratory for evaluating solutions to the property Protecting Industrial Inventions, Authors' Rights, and Traditional Knowledge 111 BOX 6.1 Protection of Indigenous Knowledge in the Philippines The Philippines recently developed and imple- IPRA gave indigenous peoples rights over mented legislation on the protection of biological their ancestral lands as well as rights to use and resources. Two noteworthy attempts have been develop the natural resources found in their Executive Order 247 (EO247) adopted in 1995 ancestral domains. It did so by creating the and the Indigenous Peoples' Act (IPRA) of 1997. National Council for Indigenous Peoples, which EO247 covers the prospecting of all biological is responsible for overseeing the issuance of per- resources in the public and private domains and mits for access to indigenous peoples' lands on requires anyone, whether a national or foreign the basis of prior informed consent. IPRA seeks entity, who wishes to access biological resources to extend the system of controlled access to enter into a formal research agreement with beyond biological and genetic resources to their the government after obtaining the informed "sciences, technologies, and cultural manifesta- consent of the appropriate local community. tions" by recognizing the concept of "commu- Although this approach has increased the nity intellectual property rights." involvement of the indigenous community, the Although IPRA remains in its infancy, it is actual practice has been criticized for being expected to provide impetus for significant more of a procedural checklist than a substan- changes to IPR law in the Philippines. Hurdles tive process in that access requires no more than related to implementation need to be worked notification to the local community. Also, EO247 out--and one of the most challenging ones is to does not elaborate precisely how benefit sharing define and implement community-based IPRs. with the local communities comes about. For details on progress in the Philippines on Notably, this legislation has been used as a basis legislating protection of traditional knowledge, for a model law on access and benefit sharing see www.worldbank.org/eaptrade, which pro- for national systems in Asia by the Association of vides background papers and technical Southeast Asian Nations (ASEAN). annexes. rights problems that are affecting the R&D sector new forum. Complex problems--conceptual and globally. practical--will have to be sorted out before progress can be made in linking traditional knowl- edge issues to WTO agreements or other mecha- Conclusions nisms for managing R&D in the biological sciences. The historical tendency to treat genetic resources In summary, the application of property rights as constant and freely available is an anachronism to traditional knowledge raises important prob- in an era of private property rights for the prod- lems, but there is probably no one regime that ucts emanating from those resources. Thus a better could serve as a solution. In compensating commu- approach to the management of these resources is nities that hold traditional knowledge, other insti- needed, and some public intervention is required. tutional devices should also be considered, such as Given international externalities, this public inter- geographical indications, copyrights, and trade vention should ideally be global in scope. In fact, secrets. In selecting a second-best approach to rights acquired in countries of origin alone will managing R&D in the biological sciences, the com- not mean much unless they are extended to major parative costliness of the various approaches will markets where derived and patented products are have to be the guide. marketed. It is therefore appropriate to discuss the protec- Overall Conclusions and Lessons tion and compensation of traditional knowledge in international forums. The TRIPs agreement estab- The development aspects of IPR regimes differ lishes the principle that some genetic resources across the three forms of intellectual property pro- should be subject to intellectual property rights, tection examined here--patents, copyrights, and but it does not resolve the conflicts between WTO traditional knowledge--for countries of widely dif- member states; it merely transfers the debate to this ferent development levels and technological capac- 112 East Asia Integrates ities. Yet for each form, issues of balance and com- from those resources through R&D activity, which plementary policies are prominent. is heavily concentrated in industrial countries. In Within the more advanced emerging economies East Asia, the Philippines is experimenting in this and parts of China, stronger IPR regimes, success- field with existing and pending legislation on fully enforced, could stimulate innovation in tech- indigenous knowledge. However, rights acquired in nology, provided other conditions are favorable. countries of origin alone will not mean much Retrospective analysis of Korea's strengthened IPR unless they are extended to major markets where regime suggests that it played a role in the country's derived and patented products are marketed. Thus dramatic success in patenting, but so did industrial a broader approach to the management of protec- upgrading, a big push in research and development tion of traditional knowledge is needed, and some from the chaebols, and the government's selective public intervention is required--ideally global in targeting of the electronics/semiconductor indus- scope--to balance the interests of the communities try, where Korea has emerged as a leading innova- holding traditional knowledge, those carrying out tor worldwide. The successes of these innovative the R&D activity, and the ultimate users of these businesses helped to offset the impact of stronger derived products. patent protection on the users of new products and Much of the debate on IPR has focused on the technologies in Korea. Because impending TRIPs- impact of more stringent rights on East Asia as a related obligations mandate stronger patent protec- user of technology created elsewhere, but strength- tion, it will be worthwhile for other, more advanced ened IPR regimes also play a role in local technol- emerging economies to identify such complemen- ogy generation by compensating inventors and cre- tary policies to boost innovative activity among ators. We are mindful of the complexities of design private firms in their economies. for each form of IPR, and for countries at each For a broader range of middle- to low-income stage of development, and the need to identify East Asian economies, copyrights might offer complementary policies. Nonetheless, if managed opportunities for gains, but again balance is criti- carefully, the scope for dynamic gains in innovation cal. Contemporary analysis of Indonesia's copy- should be considerable. right protection suggests that considerable scope exists for developing local copyright sectors--given Endnotes the talents of software developers, musicians, artists, and authors--which would prefer some- 1. A host of instruments are available for protecting different kinds of intellectual property: patents and trade secrets to what stronger protection. Yet these gains would protect industrial inventions; copyrights to protect have to be assessed against the short-term hardship authored works; trademarks to protect industrial logos and faced by the firms and workers who are now symbols; and geographical indications to certify that a engaged in unauthorized copying and distributing. product was made in an area that has specific characteris- tics (such as soil conditions, climate, or design traditions) In addition, enforcement is likely to remain ineffec- underlying the quality of the good. tive for the foreseeable future because of a weak 2. Maskus (2000a) discusses these trade-offs extensively. capacity. In the interim, complementary policies 3. An essential form of intellectual property protection, copy- right gives creative artists and firms the right to control the might be considered to improve the marketing copying and distribution of particular expressions of prospects of local artists, software developers, and music, art, film, and literature. Copyrights and patents publishers, such as improvements in telecommuni- operate differently, primarily because copyrights control the use of an idea's expression, while patents control the use cations and private Internet services which could of the idea itself. In recent years, copyright protection has be used initially to raise awareness abroad of been extended in many countries to such industrially useful Indonesian products. expressions as software, data compilations, performances, Traditional knowledge happens to be concen- television broadcasts, and satellite transmissions, and it also has been applied to the use of electronic copies down- trated in lower-income nations, and its protection loaded from the Internet. For these reasons, copyrights is generally expected to have direct benefits for deserve greater analytical attention than in the past. poverty reduction. Prospective analysis suggests 4. For example, if means can be found for villagers in poor countries to register their collective knowledge, whether of that treating genetic resources as constant and medicinal uses for biological materials or of other forms, freely available is an anachronism in an era of pri- for purposes of licensing its commercial use, they might vate property rights in the products emanating benefit from substantial royalty income. Protecting Industrial Inventions, Authors' Rights, and Traditional Knowledge 113 5. See Maskus (2000a) on the regime changes and World 21. For example, if a patent application granted to Samsung Bank (1993) on the economic structure. Electronics in 1996 cites a patent granted to Samsung Elec- 6. Data from the United Nations Education, Scientific, and tronics in 1990, it is regarded as a self-citation. If the ratio Cultural Organization (UNESCO). Available at http:// of self-citation to total citations is high, it means that a firm portal.unesco.org/uis/TEMPLATE/html/SandTec/Table_III or a country has already established its own technological _1_Asia.html. path or trajectory and will tend to pursue subsequent inno- 7. Patent-intensive industries are those that rely heavily on vations along this path (Song, Almeida, and Wu 2002). patenting to protect proprietary knowledge. Reliance on 22. For detailed results, see the technical appendix to this chap- patents varies by industries because, by definition, patents ter at www.worldbank.org/eaptrade. must reveal what they need to protect. Upon revelation, 23. This observation supports the finding that Korea invents in some knowledge is easier to imitate than other--for exam- areas that have smaller citation peaks, which may imply ple, highly coded knowledge (such as chemical formulas) is newer technologies (Hu and Jaffe 2001). relatively easy to duplicate once it leaks out. 24. Revealed comparative advantages for Korea, Japan, and the 8. For the results of these tests, see Appendix Table 1of this United States were 275, 253, and 133 for the electronics chapter at www.worldbank.org/eaptrade. industry and 21, 45, and 170 for the chemical industry, 9. In essence, these are the changes being requested of most respectively. Revealed comparative advantage is an index of developing countries within the next two to four years. the country's share of patents in those industries divided by 10. We conducted paired t-tests of the growth rates of Japanese the country's share of total patents, computed for the and German patents in Korea and the United States 1981­99 period. between 1983 and 1991. These tests confirm that the patent 25. A good example is the Korean advances in emerging digital growth rate in Korea was significantly higher than that in technologies. Koreans lagged behind Japan and the United the United States for both Japanese patents (t-value of States in the conventional analog-based technologies in the 2.213) and German patents (t-value of 1.835), at the signif- electronics industry. However, when the technological base icance level of p = 0.10. in the electronics industry recently moved to more digital- 11. For the results of this test, see Appendix Table 2 of this based technologies, Korean companies such as Samsung chapter at www.worldbank.org/eaptrade. Electronics and LG Electronics seized golden opportunities 12. All dollar amounts are current U.S. dollars. to catch up with industry leaders in the European Union, 13. Recent increases in patents per unit of R&D in many devel- Japan, and the United States by investing heavily in digital oped countries have also been attributed to an increase in technologies early on. For both companies, a substantial "defensive" patenting--that is, innovations that a firm nor- number of U.S. patents are now for technologies related to mally would not commercialize and protect, but it does so emerging digital appliances, and some of them are more only to prevent others from patenting them later. In this basic product patents than traditional process patents. way, the firm increases the portfolio of patents it owns to 26. About 10 years are usually devoted to the discovery, animal improve its negotiating power in technology purchase deals. testing, clinical trials, and approval of a new drug. 14. This is likely to be the outcome of a combination of factors 27. According to the Korea Industrial Technology Association, such as more patentable inventions per unit of R&D, as well as of December 1999 the number of researchers in the elec- as a higher propensity to patent eligible inventions for the tronics industry was 65,028, or about 38 percent of the strategic reasons just described. See also Kim (1997). researchers in Korea, whereas researchers in the chemical 15. Of the U.S. patents granted to private Korean corporations, industry numbered only 15,807. The ratio of R&D expen- 90 percent were awarded to the five largest chaebols, espe- ditures to sales for Korean electronics companies was 4.54 cially the Samsung Group. percent in 1997, or more than twice the average for manu- 16. It was 32 percent, compared with 16 percent in Taiwan and facturing firms. 9 percent in Japan (Kim 1997). 28. In 2003 extensive interviews of government officials and 17. In 1998 the ratios of R&D expenditures to total sales for the enterprise managers were carried out, and a survey was flagship companies of the largest chaebols were Samsung administered to 70 Indonesian enterprises in these copy- Electronics, 7 percent; Hyundai Electronics, 8 percent; right-sensitive sectors, commissioned by the World Bank. Hyundai Motor, 12 percent; and SK Telecom, 2 percent. 29. The new law also improves procedures for arbitrating and These percentages were generally comparable to or higher resolving disputes, and it clarifies that copyright cases can than those of the global leaders in their respective industries. be appealed in a streamlined fashion to the commercial 18. As of February 2002, Korea had 9,208 private R&D labs, up court. It also increases the authority of the police to under- sixfold from 1,435 a decade earlier (Korea Industrial Tech- take enforcement actions on their own initiative. Perhaps nology Association). most significant, it makes end use piracy a criminal 19. In the United States, under law a patent application must offense, with a maximum five-year prison term. Another specify any and all of the "prior art" known to the appli- provision of the new law permits copyright owners from cant. Thus it is possible to track knowledge-building across outside Indonesia to retain Indonesian legal representa- people, firms, countries, and regions, and time. Recent tion. Under past law, a complainant had to come to court studies have analyzed patent citations to trace the sources in Indonesia to claim infringement. Foreign companies of original knowledge underlying patented innovations. (and certainly individual artists) often found this was not See Jaffe, Tranjtenberg, and Henderson (1993); Almeida worth the time and effort. Indonesia has adopted the (1996); Almeida, Song, and Grant (2002); Song, Almeida, World Intellectual Property Organization (WIPO) Copy- and Wu (2002). right Treaty and intends to ratify the WIPO Performance 20. Sorensen and Stuart (2000), Rosenkopf and Nerkar (2001), and Phonograms Treaty. Among other things, these and Song, Almeida, and Wu (2002) have used self-citations changes will make it illegal for Internet users to circum- to evaluate the extent of path dependence in innovation vent electronic protection devices on binary files and activities. transmission protocols. 114 East Asia Integrates 30. This increase seems to have come at the expense of fewer wan." NBER Working Paper No. 8528. National Bureau of trademark cases. These data are from the national police Economic Research, Cambridge, Mass. force. IPBN (Indigenous Peoples' Biodiversity Network). 1995. 31. The money the agent received from the resulting fines was "Indigenous People, Biodiversity, and Health." COURTS contributed to a program to purchase computers for schools. Canada IPBN Factsheet, November. 32. Details of the survey results and descriptions of the indus- Jaffe, Adam B., Manuel Trajtenberg, and Rebecca Henderson. tries appear in the technical appendix to this chapter at 1993. "Geographic Localization of Knowledge Spillovers as wwww.worldbank.org/trade. Evidenced by Patent Citations. Quarterly Journal of Econom- 33. The median estimates from the survey were that Indonesia ics 108: 577­98. has 230 software firms, employing some 5,800 workers. By Kim, Linsu. 1997. Imitation to Innovation: The Dynamics of contrast, survey work in Lebanon in 1997 estimated that Korea's Technological Learning. Cambridge, Mass.: Harvard the country was home to some 490 software companies Business School Press. employing about 3,000 workers (Maskus 2000b). Clearly, Maskus, Keith E. 2000a. Intellectual Property Rights in the Global the Indonesian market potential is far larger than that of Economy. Washington, D.C.: Institute for International Eco- Lebanon. nomics. 34. Indeed, it may be that some aspects of Indonesia's intellec- ------. 2000b. "Strengthening Intellectual Property Rights in tual property regime--in particular, the limitations on fair Lebanon." In Bernard Hoekman and Jamel Zarrouk, eds., use and the provisions for patenting software--are overly Catching Up with the Competition: Trade Opportunities and restrictive from the standpoint of the country's develop- Challenges for Arab Countries. Ann Arbor: University of ment needs. In software, the tighter copyright law to be Michigan Press. introduced in 2003 may substantially limit fair use provi- OECD (Organisation for Economic Co-operation and Develop- sions. In addition to the prior restraint that the user of a ment). 1999. Main Science and Technology Indicators, 1999. computer program may make only one archival copy, edu- Geneva. cational users of software and printed materials will be per- Prakash, S. 1999. "Towards a Synergy between Biodiversity and mitted only one copy for their use before they are required Intellectual Property Rights." Journal of World Intellectual to attain licenses for multiple copies. The new law outlaws Property 2 (5). the decompilation (reverse engineering) of computer code. Rosenkopf, Lori, and Atul Nerkar. 2001. "Beyond Local Search: 35. For example, a hectare of land in West Ecuador and the Ama- Boundary-Spanning, Exploration, and Impact in the Optical zon--biodiversity hot spots of the world--is valued at $20.63 Disc Industry." Strategic Management Journal 22: 287­306. and $2.59, respectively (Simpson, Sedjo, and Reid 1996). Simpson, R. D., R. A. Sedjo, and J. W. Reid. 1996."Valuing Biodi- 36. It is now possible to claim a patent in any or all of the follow- versity for Use in Pharmaceutical Research." Journal of Polit- ing steps toward identifying useful biological activity: genetic ical Economy 104 (1): 163­85. sequence, cloning method, expression of protein from Song, Jaeyong, Paul Almeida, and Geraldine Wu. 2002. "Learn- sequence,biological activity of protein,and method of action. ing-by-Hiring: When Is Mobility Useful in Inter-firm Knowledge Transfer?" In Linda Argote, B. McEvily, and R. Reagens, eds., Management Science (Special Issue on Knowl- References edge) 49 (4): 446­63. Sorensen, Jesper B., and Toby E. Stuart. 2000. "Aging, Obsoles- Almeida, Paul. 1996. "Knowledge Sourcing by Foreign Multina- cence, and Organizational Innovation." Administrative Sci- tionals: Patent Citation Analysis in the US Semiconductor ence Quarterly 45: 81­112. Industry." Strategic Management Journal 20: 251­59. Swanson, T., and T. Goeschl. 1999."Ecology, Information, Exter- Almeida, Paul, Jaeyong Song, and Robert M. Grant. 2002. "Are nalities and Policies: The Optimal Management of Biodiver- Firms Superior to Alliances and Markets? An Empirical Test sity for Agriculture." In G. H. Peters and J. von Braun, eds., of Cross-Border Knowledge Building." Organization Science Food Security, Diversification and Resource Management: 13 (2): 147­61. Refocusing the Role of Agriculture. Brookfield, Vt.: Ashgate. Hu, Albert G. Z., and Adam Jaffe. 2001. "Patent Citations and World Bank. 1993. The East Asian Miracle. New York: Oxford International Knowledge Flow: The Cases of Korea and Tai- University Press. 7 Trade and Competitiveness Aspects of Environmental and Labor Standards in East Asia Keith E. Maskus Rapid economic growth is often blamed for the is that foreign direct investment (FDI) and trade widespread deterioration of the natural environ- respond at least in part to country standards. If ment in East Asian economies through deforesta- weak environmental and labor standards provide tion, urbanization, and industrialization. Put dif- cost advantages that motivate domestic and foreign ferently, it is claimed that domestic and firms to locate in areas with such standards, the multinational companies gain a competitive advan- standards may stimulate growth in exports. If weak tage from the willingness of public authorities to or ineffective standards are an important spur to permit environmental resources to be used indis- competitiveness and export growth, and firms at criminately or without charge. Similarly, it is least implicitly demand such weakness through claimed that a permissive government attitude their location decisions, governments may compete toward the protection of workers' rights permits to offer low standards in a "race to the bottom." firms to suppress wages and working conditions, Now that many countries have significantly generating a cost advantage on the labor side. A reduced their tariffs and other border restraints on particular concern is that labor and environmental trade, standards that raise costs have presumably standards may have been weakened in the after- increased their influence on trade. math of the Asian financial crisis of 1997­98, or An alternative hypothesis is that exceptionally because of additional trade liberalization in the late weak environmental and labor standards restrain 1990s, and that this weakening has boosted exports economic competitiveness, exports, and growth.1 since the crisis. Poor environmental stewardship Unchecked or untreated discharges of chemicals and a failure to support at least core labor rights into water supplies can harm agricultural produc- raise issues of human rights, health maintenance, tion and workers' health status and limit productiv- educational attainment, and sustainability. ity. Widespread reliance on child labor limits edu- This chapter reviews theoretical and empirical cational enrollments and the development of labor analysis to help clarify these issues. Three compet- skills, and potentially restricts both the moderniza- ing hypotheses are influential in the literature. One tion of the output mix and growth rates. To the The author is grateful to Kanemi Ban for helpful comments at the seminar at the Tokyo office of the World Bank. 115 116 East Asia Integrates extent that firms require high productivity, low huge, reflecting the complexity of the subject.2 This absenteeism, and clean resources, weak standards section summarizes important institutional ques- can deter investment and restrict exports. In this tions and the available evidence. view, competitiveness flows more readily from Many different kinds of regulations provide higher social protection than from lower. workers with rights in labor markets: Both of these hypotheses suggest that competi- tiveness, trade, and FDI depend on the effects of · Basic rights include prohibition of slavery and environmental and labor protection regulations physical coercion, elimination of discrimina- (and their enforcement) on costs. A third claim tion, and a ban on the exploitative use of child reverses this causation, arguing that the optimal labor. standards for a country depend on factor endow- · Civic rights include freedom of association, col- ments, development levels, technology, and national lective bargaining, and expression of grievances. preferences (Anderson 1996; Antweiler, Copeland, Together, these provisions constitute the so- and Taylor 2001). In this context, regulations are called core labor standards, which embody prin- endogenous to changes in openness to trade and ciples that are supposed to be universally FDI. Openness may either worsen or improve the respected as a matter of human rights.3 natural environment, depending on prevailing con- · Survival rights provide for a living wage or mini- ditions, but to the extent that openness raises mum wage, limited hours of work, information incomes, demand for stronger environmental pro- about working conditions and job hazards, pro- tection may be expected to rise. Similar comments tection from occupational hazards, and com- apply to rules on the protection of workers. These pensation for employment-related accidents. higher demands for stronger environmental protec- · Security rights include protection against arbi- tion and protection of workers have been observed trary firing, rights to severance payments, and in recent decades in Japan, the Republic of Korea, access to health and retirement benefits and sur- Taiwan (China), and Singapore as development has vivors' compensation. proceeded, in part because of export-led growth. These hypotheses cannot be tested conclusively. These various rights are granted by governments One reason is that the relationships between labor to workers through regulations on employer prac- standards or environmental standards, on the one tices.4 Such regulations vary considerably across hand, and international trade, on the other, are com- countries, and even within countries. plex and involve causality in both directions. An important distinction may be made between Another reason is that it is difficult with the available core labor standards and other rights. To a consid- information to capture the subtleties of labor rights erable degree, core labor standards may be inter- or environmental conditions for statistical analysis. preted as policies buttressing the basic freedoms in Even so, by reviewing the balance of evidence it is the workplace. The elimination of coercive forms of possible to draw useful guidelines for policy. The labor and discrimination gives workers and firms next section considers general issues in the area of wider choices, and costly distortions in labor mar- labor standards and trade, and it is followed by a dis- kets are removed (Maskus 1997; Martin and cussion of the situation in East Asian nations. The Maskus 2001). Recognition of the rights of workers next sections repeat this structure for environmental to associate freely and engage in collective bargain- regulation. They are followed by a review of environ- ing allows worker groups to counter the distor- mental protection issues arising in the Doha round tionary impacts of single employers. By contrast, of multilateral trade negotiations. A concluding sec- survival rights and security rights award workers tion offers policy recommendations. benefits that might not otherwise emerge in com- petitive labor markets. But they also raise the costs and reduce the flexibility of firms. Moreover, as the Labor Standards: Background strength of such mandates rises, firms may find and Institutions themselves less able to compete.5 This problem The literature on relationships between labor mar- might be particularly difficult for minimum wage ket regulations and international competition is laws, which can contribute to unemployment in the Trade and Competitiveness Aspects of Environmental and Labor Standards in East Asia 117 event of an economic crisis or political collapse, as since 1997; indeed, there is some evidence of a gen- suggested by a recent study of Indonesia (Suryahadi eral trend toward raising them.6 Nevertheless, a and others 2003). review of legislation finds that Korea is the only In addition to regulating employers' practices, country to have significantly strengthened its regu- governments may directly support workers latory framework for labor protection. China has through public services. Such services consist of done little to improve its formal recognition and social safety net programs, including temporary protection of union rights. Meanwhile, allegations unemployment compensation, and programs to persist that governments in the region are raising improve workers' ability to function in labor mar- roadblocks to union activities, and enforcement of kets, including job training, wage subsidies, labor laws is often thought to be weak.7 employment services, job creation through public Workers in the informal sector are not directly works projects, and education. affected by changes in labor legislation. They have Table 7.1 compares the policies of key East Asian few rights and may work in conditions that are less economies on workers' rights and support pro- sanitary and more dangerous than those in formal grams by indicating their decisions on ratification employment (Maskus 1997). Child labor tends to of the eight "fundamental" conventions of the be concentrated in the informal sector. Employ- International Labour Organisation (ILO) that ment in the informal economy depends on many cover the core labor standards. In the table, the factors (including the effects of labor regulations many ratifications since 1996 are indicated by bold- on costs in the formal economy), but especially on face, and they suggest that worker protection stan- the rate of overall growth in the economy. dards have improved since the Asian financial cri- One impact of the Asian financial crisis was a sis. These data reveal that legislated labor standards shift toward informal employment in many East have not been reduced in the East Asian region Asian economies (Betcherman and Islam 2001), TABLE 7.1 Ratifications of Fundamental ILO Conventions Covering Worker Rights, Selected East Asian Countries Convention (convention number) Worst Free forms Abolition dom of of Equal Nondis- of Minimum child Forced forced remuner- crimin- associ- Collective age labor labor labor ation ation ation bargaining (138) (182) (29) (105) (100) (111) (87) (98) Cambodia Yes No Yes Yes Yes Yes Yes Yes China Yes No No No Yes No No No Hong Kong (China) Yesa n.a. Yesa Yesa No No Yesa Yesa Indonesia Yes Yes Yes Yes Yes Yes Yes Yes Korea, Rep. of Yes Yes No No Yes Yes No No Malaysia Yes Yes Yes No Yes No No Yes Philippines Yes Yes No Yes Yes Yes Yes Yes Singapore No Yes Yes No Yes No No Yes Thailand No Yes Yes Yes Yes No No No Vietnam No Yes No No Yes Yes No No n.a. Not applicable. Note: Words in boldface indicate ratification decisions since 1996. a. Indicates Hong Kong's intention to comply with the conventions indicated, even though Hong Kong is not a member of the International Labour Organisation (ILO). Sources: OECD (2000) and www.ilo.org. 118 East Asia Integrates which suggests that working conditions deterio- many East Asian countries fell dramatically as those rated on average. Although the shift was temporary economies became richer (U.S. Department of in much of the region, it has persisted in Indonesia Labor 1994). Second, richer households gain greater and the Philippines.8 access to credit markets, permitting them to invest more in acquiring skills. This investment can play an important role in decisions by rural households Evidence on Labor Standards, to keep their children in school, but it is also impor- Trade, and Competitiveness tant for increasing adult productivity, which sup- Because the provision of workers' rights affects the ports better working conditions. Third, the demand costs of hiring workers, it may influence interna- for tighter labor protection increases. This demand tional price competitiveness. For example, the abil- reflects both the impact of rising real wages--which ity of firms to require employees to work long hours make workers demand more protection to safe- without a premium for overtime, or to discharge guard their incomes from dismissal or injury--and workers with little notice or severance pay, presum- the nature of labor standards as public goods that ably reduces their labor costs per unit of output, become more affordable as incomes increase. which may be significant in labor-intensive prod- Another important determinant of labor rights ucts. Extensive government support programs for is the endowment and output mix of the economy. training and employment services can reduce aver- In countries with abundant unskilled labor, firms in age labor costs, depending on how such programs labor-intensive sectors may view workers as easily are financed. Unfortunately, it is quite difficult to replaceable and have few incentives to train them. In ascribe changes in trade performance to such economies with greater endowments of skilled labor actions, both because they are difficult to measure and training, workers offer more differentiated-- (particularly at the firm level) and because the rela- and more valuable--skills that make employers tionships are complex. This section considers these more concerned about retaining them and about relationships and reviews some general evidence. avoiding the output losses associated with occupa- tional injuries. Rapid growth emanating from the accumulation of human capital tends to increase the Effects of Openness on Labor Standards and demand for better working conditions. Conditions of Work Yet another determinant of labor rights is the Table 7.2 gives an aggregate index of the strength of nature of competition in both the product and labor standards across several developing countries labor markets. A firm that is the sole or main in the late 1990s (Verite 2002). Among East Asian employer in a local labor market can suppress economies, China, Indonesia, and Malaysia rank wages or working conditions. near the bottom of this scale. Regulations in the A government's willingness to legislate and labor market depend on several influences, includ- enforce better labor standards depends on the politi- ing economic openness, which vary across coun- cal-economic situation. A strong positive correlation tries and over time (Freeman 1994; Fields 1995; exists between democracy and the rights of workers OECD 1996; Maskus 1997). to freely associate (OECD 2000), as corroborated by As the table suggests, a primary determinant of the last column of Table 7.2. In countries without labor standards is the level of income per person or effective political competition, it may be difficult to per family in an economy. The poor may be little establish stronger standards, even where economic able to refuse dangerous work. Clearly, extensive efficiency could be improved by doing so or where poverty is the main factor underlying high rates of much of the public would prefer such standards. child labor participation (Grootaert and Kanbur 1995). As incomes rise, the treatment of labor Effects of Changes in a Country's Trade and improves for several reasons. First, the contribution Investment Regulations on Labor Standards of children to household income becomes less important, and schooling becomes a higher-valued Tariff cuts could have several distinctive impacts on alternative as parents become more capable of workers, depending on the economic circum- investing in their children. The child labor supply in stances of each country: Trade and Competitiveness Aspects of Environmental and Labor Standards in East Asia 119 TABLE 7.2 Overall Ranking of Labor Standards Labor Fundamental Freedom Standards ILO PPP per House Index, conventions capita GDP, Index, Country late 1990sa 1995 1999 1999 (US$) 1998­99b Chile 83.3 3 7 8,720 5.5 Argentina 80.8 6 7 12,258 5.0 South Africa 78.0 1 5 8,994 6.5 Korea, Rep. of 71.3 0 3 15.778 6.0 Brazil 67.8 5 5 7,130 4.5 Philippines 63.8 5 6 3,803 5.5 Thailand 63.0 2 3 6,095 5.5 Colombia 62.0 6 6 5,821 4.5 Peru 59.0 6 6 4,626 3.5 Venezuela 56.0 7 7 5,586 5.5 Mexico 55.3 5 5 8,329 4.5 Turkey 55.0 5 8 6,374 3.5 India 46.5 3 3 2,250 5.5 Malaysia 46.3 2 4 8,215 3.0 Indonesia 44.5 3 7 2,873 3.0 Egypt 44.0 6 7 3,423 2.0 Pakistan 40.0 5 5 1,813 3.5 China 36.3 1 2 3,620 1.5 Correlations with Labor Standards Index 0.71 0.77 Note: PPP = purchasing power parity. a. This index is a weighted average of four broad categories: ratification of the fundamental International Labour Organisation (ILO) conventions (see Table 7.1); a comparison of national labor laws with ILO standard recom- mendations; an evaluation of each government's capacity to implement its laws and policies; and an assessment of the level of compliance with or violations of the core labor standards. b. This index is the average of the figures for political rights and civil liberties from the Freedom House Web site (www.freedomhouse.org). Here they have been scaled so that a figure of 1 means the least political freedom and a figure of 7 means the most political freedom. Sources: Labor Standards Index (VERLS) from Verite (2002); GDP data from World Bank (2001c). · Relative price changes from trade liberalization · Trade liberalization can increase poverty under would tend to push capital and labor into prod- certain circumstances.10 Wages in formal manu- ucts in which the country has a comparative facturing may fall significantly for workers advantage. This situation could either worsen or whose incomes had previously been protected. improve net working conditions in the econ- Workers who are laid off could end up in the omy, depending on whether conditions in the informal economy, pushing down wages there. expanding industries are worse or better than Trade liberalization increases the prices of those in other sectors.9 export goods, which may have a large weight in · In labor-abundant countries, tariff cuts ordinar- household consumption baskets. To the extent ily would raise the real wages of less skilled labor that openness to trade reduces wages and and increase aggregate income. These impacts increases poverty, labor standards and worker should induce workers to demand stronger protection will deteriorate. This impact would labor standards. As a result, over time employers be especially problematic in those regions that should be willing to provide better working con- are not competitive in producing exportable ditions. goods. In East Asian developing economies, 120 East Asia Integrates however, greater openness to trade has increased generally a force for raising labor standards rather incomes, reduced poverty, and improved work- than the other way around. ing conditions.11 · As the product markets of different countries Evidence on the Effects of Labor Standards on Trade become more integrated through trade liberal- Performance and FDI ization, firms may seek out locations with weaker standards. Studies of practices in devel- Have weak labor standards in East Asian develop- oped countries have found no evidence of a ing economies contributed positively or negatively "race to the bottom" in labor standards (Maskus to the export performance of these economies and 1997; OECD 2000). Unfortunately, it appears to their attractiveness for inward FDI? This is a dif- that no studies have been done of competition ficult question to answer. First, economic theory in labor standards among developing countries cannot establish a clear relationship between weak- that produce similar goods for export. ness in labor standards and international trade competitiveness.14 Second, labor standards and working conditions are only two of numerous Labor Standards in Export Processing Zones determinants of relative costs, and it is virtually Some analysts view export processing zones impossible to isolate their impacts using aggregate (EPZs), or free trade zones, as examples of compet- data. A truly informative answer would require itive standard setting.12 In many instances, workers detailed survey work at the firm level and would within EPZs have been denied the rights to union- track the dynamics of labor use over several years. ize and bargain collectively, and safety and health Third, available aggregate measures of labor rights conditions are poor (Moran 2002). This complaint are crude and cannot capture the subtleties of cost has been leveled at some of the EPZs in China, impacts or enforcement efforts, nor can they con- Indonesia, and the Philippines, suggesting that East trol adequately for other influences on trade. Evi- Asian countries may be competing in this way. dence on this subject must be treated with caution. However, no evidence was found that standards in As for the relationship between labor standards EPZs had been weakened in the aftermath of the and international trade performance, Aggarwal's Asian financial crisis or because of additional trade (1995) study of manufacturing sectors in develop- liberalization in the late 1990s. ing countries found that workers in export-ori- By contrast, there is evidence that the operations ented firms received higher wages and benefits than of multinational enterprises, even within EPZs, those in less export-oriented firms. And countries tend to improve working conditions and wages rel- with weak labor rights did not have higher import ative to those in the rest of the economy.13 It is not penetration ratios in the United States than coun- difficult to understand why multinational firms pay tries with stronger labor rights. Rodrik's (1996) higher-than-average wages in developing countries. cross-country econometric analysis found no rela- First, these firms invest much more in advanced tionship between basic measures of labor standards industries, such as electrical equipment, electron- and international trade flows. Similarly, the Organ- ics, industrial machinery, and automobile parts, isation for Economic Co-operation and Develop- than in the less advanced industries such as gar- ment (OECD 1996) found no relationships ments, textiles, footwear, and toys. Jobs in the between labor standards and measures of export advanced industries require higher skills and pay performance. Nor could the OECD authors detect considerably higher wages than those in the less any correlations between measures of revealed advanced industries. Second, these firms tend to comparative advantage and attempts to suppress produce for export, which requires sustained qual- labor union rights, or any association between indi- ity, specific skills, and limited absenteeism, and they vidual countries' core labor standards and the are willing to pay higher wages in return. Third, prices of their exports of textiles and apparel to the multinational firms typically bring superior tech- United States. They concluded that differences in nologies to the factory floor, which raise productiv- core labor standards have no detectable effects on ity and wages relative to those of domestic firms. In patterns of specialization, competitiveness, or this regard, openness to foreign direct investment is exports. Trade and Competitiveness Aspects of Environmental and Labor Standards in East Asia 121 Van Beers (1998), analyzing OECD trade data, At best, however, the data simply do not support found a weak association between stricter labor this notion. Data on the shares of labor-intensive standards and lower exports of labor-intensive products in total manufactured exports show that, goods. His index of labor standards included such except perhaps in China and Indonesia, the output measures as maximum working hours, conditions mix shift has, over time, shifted away from labor- of employment contracts, and minimum wages, intensive goods.16 The export data cannot reveal suggesting that strong security rights can negatively much about how labor standards affected changes affect trade performance in richer nations. It may in trade shares, because numerous other factors be, however, that inflexible labor market mandates could have driven these changes.17 But data on reduce competitiveness rather than fundamental labor conditions in this period provide useful com- labor standards.15 plementary insights (ILO 2002). Despite the large As for the relationship between labor standards increases in unemployment between 1995 and 1999 and FDI, Aggarwal (1995) found no association in Hong Kong (China), Indonesia, Korea, Malaysia, between U.S. foreign direct investment and poor the Philippines, and Thailand, there was virtually labor standards; if anything, FDI from the United no change in reported hours worked per week. States was less concentrated than expected in coun- Moreover, in these countries the number of tries with low standards. Rodrik (1996) found some reported injuries per 100,000 fell sharply during indication that FDI from the United States was 1995­99. Thus, at least on these simple indicators, lower than expected in countries with limited core firms (in the formal sector) were not pressuring labor standards. OECD (2000) reported that larger workers to work harder or more dangerously. inflows of FDI are associated with stronger labor For a more formal test of the hypothesis that rights. This study also reported growing evidence labor standards have affected East Asian exports of that export processing zones with poor working labor-intensive goods, and did so more strongly conditions are less likely to attract sustained, long- after the onset of the financial crisis, an economet- term FDI than those with better conditions. Kucera ric analysis was undertaken.18 The underlying (2001) regressed aggregate flows of FDI into many models are estimated for 1995 and 1999 to see if developing countries on several indexes of labor any differences may have arisen during the crisis in standards. He found no statistical indication that terms of the impact of standards on exports. (If weak labor rights attract FDI, and some of his coef- weak labor rights were a factor raising exports of ficients suggested that the opposite result holds. labor-intensive manufactures, one would expect to The analysis just described suggests that weak find a negative coefficient in both years. But if the labor standards are capable of expanding exports negative coefficient was larger in 1999 than in 1995, most readily in labor-intensive goods such as it would suggest that differential labor rights pro- apparel, textiles, footwear, and miscellaneous manu- vided a stronger impetus to exports in the latter factures such as toys and sporting goods. The Asian year.) This is at best a partial and crude test of the financial crisis may have reduced the relative costs of notion that labor rights were effectively weakened production in labor-intensive exports for several during the crisis and that they helped to boost reasons. First, even if legislated labor standards were exports in the post-crisis period. not diminished during this period, authorities may Three measures of labor standards were used: an have signaled a reduced commitment to enforcing index of the strength of freedom of association (FA) workers' rights within labor-intensive industries. rights in the exporting countries; the index of four Second, the costs of acquiring inputs from subcon- labor standards developed by Verite (2002--VERLS, tractors may have fallen quite sharply during the see Table 7.2); and the number of the eight funda- crisis if wages fell in the subcontractors' firms. mental ILO conventions that were ratified by the Third, exchange rate changes may pass through exporting countries by 1995 or 1999 (see Table 7.1). more quickly to competitive sectors. Under such cir- For two of these three measures of labor stan- cumstances, one would expect effectively weaker dards, the results suggest that bilateral export vol- labor standards to have raised the share of labor- umes in labor-intensive goods actually rise as intensive manufactures in total exports between worker protection is increased--a finding that is 1995 (before the crisis) and 1999 in East Asia. particularly true within East Asia (Table 7.3).19 The 122 East Asia Integrates TABLE 7.3 Impacts of Labor Standards on Labor­Intensive Exports from Developing Countries Model 1 Model 1 Model 2 Model 2 Model 3 Model 3 Variable (1995) (1999) (1995) (1999) (1995) (1999) NAFTA 2.37*** 3.53*** 2.52*** 3.62*** 2.07*** 2.86*** Political freedom ­0.30*** ­0.19*** ­0.82*** ­0.91*** ­0.27*** ­0.11** East Asia 1.05* 3.13*** 2.80*** 4.28*** 1.59*** 0.05 South Asia 8.18*** 5.46*** 16.9*** 6.11*** ­0.09 ­1.56 Latin America 2.37** 1.34 1.55* 6.18*** 0.69 0.72 FA rights 0.60*** 0.66*** (1.37) (1.59) FA rights · East 0.94*** ­0.32 Asia (2.13) (1.32) FA rights · South ­2.42*** ­1.52*** Asia (0.71) (1.15) FA rights · Latin ­1.18** ­1.03** America (0.27) (0.57) VERLS 0.09*** 0.14*** (5.28) (8.22) VERLS · East Asia 0.01 ­0.02 (5.41) (7.81) VERLS · South Asia ­0.21*** ­0.04 (3.31) (7.94) VERLS · Latin ­0.03** ­0.11*** America (4.56) (5.57) ILO-F ­0.16** ­0.52*** (­0.62) (­2.69) ILO-F · East Asia 0.38*** 0.42** (­0.10) (­1.93) ILO-F · South Asia 0.85*** 0.71*** (­0.23) (­2.25) ILO-F · Latin ­0.19** ­0.27 America (­0.92) (­3.27) No. of observations 1,700 1,700 1,700 1,700 1,700 1,700 R­squared 0.74 0.76 0.75 0.77 0.74 0.77 * Significant at 10 percent level. ** Significant at 5 percent level. *** Significantly different from zero at 1 percent level. Notes: FA = freedom of association; ILO-F = fundamental International Labour Organisation conventions. The coefficients come from gravity models of bilateral exports from 17 developing countries to 20 OECD importers listed below. Each equation is estimated as an augmented gravity model with controls for importer and exporter GDP, importer and exporter population, bilateral distance, and industry fixed effects. Standard errors are robust to heteroskedasticity. Figures in parentheses are elasticities calculated at sample means. Export values are included for the set of labor-intensive industries listed below. The regressions include standard gravity variables (logs of importer and exporter GDP levels, logs of importer and exporter population levels, and the log of distance between trading partners), a dummy variable for NAFTA when the exporter is Mexico and the importer is either the United States or Canada, an index of political freedoms from the Freedom House Web site (www.freedomhouse.org), industry fixed effects, regional fixed effects, a measure of labor standards, and interaction terms between labor standards and regional dummies. Data on GDP (in billions of U.S. dollars in purchasing power parity terms) and population (in millions) are from World Bank (2001c). Distance is the number of kilometers (in thousands) between capital cities. The index of political freedoms is the simple average of the "political freedom" and "civil liberties" indicators from Freedom House (www.freedomhouse.org), rescaled so that an increase in the index here signifies an increase in political rights. Labor-intensive industries are defined as ISIC (International Standard Industrial Classification) 321, 322, 323, 324, and 390. Exporters are Argentina, Brazil, Chile, Colombia, Mexico, Peru, Venezuela (Latin America); India, Pakistan (South Asia); China, Indonesia, Republic of Korea, Malaysia, the Philippines, Thailand (East Asia); Egypt, and South Africa. Importers are Australia, Austria, Canada, Denmark, Finland, France, Germany, Greece, Japan, Ireland, Italy, the Netherlands, New Zealand, Norway, Portugal, Spain, Sweden, Turkey, United Kingdom, and United States. Source: The author. Trade and Competitiveness Aspects of Environmental and Labor Standards in East Asia 123 balance of evidence does not support the view that farming communities. The collapse also seemed to weak labor rights promote exports in these goods, speed deforestation in Indonesia and Vietnam, nor does it suggest decisively that during the finan- although a reduction in export demand in Japan cial crisis weak labor rights translated into stronger and Korea offset this problem initially (Marinova labor-intensive exports. 1999; World Bank 1999, 2001b). Finally, there are The policy message that emerges is that East indications that fishing stocks and coral reefs have Asian developing countries need not delay in intro- come under greater stress as a result of rising ducing core labor standards. Doing so will not unemployment, although most such pressures are reduce their competitiveness in labor-intensive long term rather than cyclical (World Bank 1999; manufactured exports and could well increase it. World Resources Institute 2002). However, they should exercise caution in imple- Like for labor standards, there is little indication menting stronger measures that could limit flexibil- that governments actively reduced their legislated ity in formal labor markets.20 or formal environmental regulations during the crisis period. In fact, since 1997 some economies have adopted stricter regulation or greater incen- Environmental Protection: tives for conservation: Background and Institutions It is beyond the scope of this chapter to attempt a · China tightened its industrial point source pol- comprehensive review of environmental regulatory lution control and deforestation regulations in regimes. All the major East Asian economies have the late 1990s (World Bank 2001a), and it has some form of regulatory regime and institutions increased pollution discharge fees to levels at aimed at protecting the environment or promoting least as high as treatment costs (APEC 2001). sustainable resource use. But these regimes vary · Hong Kong is subsidizing projects to develop considerably in terms of resource commitments, green production technologies, develop new enforcement capacity, and regulatory mechanisms waste treatment systems, and promote ISO (Esty and Cornelius 2002). 14000 environmental standards. Vietnam has Did the onset of the financial crisis affect envi- announced a reforestation program (World ronmental use and policy in the region? Reports Bank 2001a). suggest that the crisis did alter various indicators of · Korea's Ministry of Environment has expanded environmental damage, though in different ways. its programs for pollution prevention, including In many cities, the short-term effect of economic certification of enterprises, and it has begun collapse was improved air quality as industrial pro- requiring compliance with certain OECD stan- duction slackened and as vehicles were used less in dards. It also has expanded tax incentives and reaction to lower product demand and, in some procurement policies for recycling (World Bank cases, sharply higher fuel prices (World Bank 2000). Korea adopted a new law in December 1999). By 1999, however, air pollution indexes were 1998 that revises the Waste Management Act. regaining and perhaps exceeding their previous lev- The new law clarifies and increases landowners' els. Water quality is relatively insensitive to short- liabilities for abandoned waste, which has term changes in industrial activity and household become a significant and costly problem to clean consumption, and so water quality measures up. By 1999 Korea had eased up on its regula- changed little. But sanitation conditions deterio- tions in the textile dyeing industry, despite a rated markedly in many locations because growing large drop in output. However, perhaps related amounts of industrial and toxic waste were not to the financial crisis, Korea decided to delay being disposed of properly. Industrial pollution revising some air quality standards in 1997. And worsened somewhat in some countries, perhaps recently the government has shifted toward because of weaker regulatory control and reduced deregulation of onerous mandates and self-cer- compliance by firms. In several countries such as tification and voluntary compliance by busi- Indonesia, some urban dwellers responded to the nesses. financial crisis by returning to the countryside, · Thailand recently implemented a series of plans which placed more stress on natural resources and for enhancing and conserving national environ- 124 East Asia Integrates mental quality (APEC 2001). This approach Evidence on Environmental includes both tighter pollution control, imple- Protection, Trade, and menting the "polluter pays" principle, and pro- Competitiveness grams to promote the development of new tech- nologies. The links between declines in economic activity and environmental degradation are complex and However, in the countries affected strongly by not easily predicted. The same is true for the links the slowdown, the associated fiscal problems have between openness to trade and FDI and use of the reduced public expenditures on environmental environment. These impacts presumably work in protection (World Bank 1999, 2001a).21 This both directions and depend on circumstances. reduced spending has been a significant problem in Indonesia, where the administrative and regulatory How Does Openness Affect Environmental framework does not encourage sustainable use of Protection and Use? the environment. It is likely that the collapse of cen- tral government authority exacerbated Indonesia's Grossman and Krueger (1993) describe the impacts fiscal difficulties. Forest loss has continued, land of changes in an economy's fundamental vari- use management problems remain severe, and ables--endowments, prices, technologies, and poli- inadequate wastewater treatment and the dumping cies--on the aggregate use of the environment.22 of hazardous wastes have not abated. Some com- Drawing on their framework, I use the word emis- mentators view Indonesia's ongoing devolution of sions as a proxy for all forms of environmental use, administrative authority to local governments as including air and water pollution, deforestation, damaging for environmental protection, because habitat destruction, and waste deposits. In the sim- these governments may not have effective capaci- plest terms, the amount of emissions an economy ties or budgets for the job and may be particularly generates depends on three essential factors: the prone to weak and nontransparent enforcement. In size of the economy, the share of output produced the Philippines, fiscal stringency has had a mixed by emission-intensive ("dirty") sectors, and the effect on incentives for environmental resource use degree of emissions intensity in those sectors. (World Bank 1999). Environmental budget cuts Any change in pollution emissions can therefore were reported, but fiscal pressures also resulted in be decomposed into three effects. The scale effect reductions in subsidies to coal production and fuel refers to an increase in emissions associated with a consumption, with associated increases in user larger gross domestic product (GDP), holding con- prices. stant the relative mix of outputs and pollution Remarkably little research has been done on the intensities across sectors--that is, a 10 percent relationships among the environment, the financial increase in all productive factors, everything else crisis, and trade in East Asia. The following research held constant, should raise pollution by 10 per- agenda, which could be pursued by national cent.23 The composition effect refers to a change in authorities, international organizations, or research the share of dirty goods in the GDP, which may institutions, would substantially improve under- come about because of a price change. With a con- standing of the processes at work: stant scale of the economy and no change in emis- sions intensities per industry, a rise in the share of · How the Asia financial crisis affected the use of dirty goods would increase total pollution. The environmentally damaging agricultural inputs technique effect refers to a change in the amount of and whether changes in such use were needed to emissions per unit of output across sectors (an sustain trade flows "emissions-intensity" change)--for example, as · Whether exchange rate changes associated with producers change technologies in response to a the crisis altered demand for imported inputs higher pollution tax.24 This typology can be used to that might have limited toxic discharges analyze the impacts of trade liberalization and FDI · Whether multinational firms had a greater or on the environment. lesser propensity than domestic firms to ignore Trade liberalization reduces impediments to cost-increasing environmental regulations. imports and exports25 and affects use of the environ- Trade and Competitiveness Aspects of Environmental and Labor Standards in East Asia 125 ment through its effects on prices, which then filter Important evidence is available on the effects of through to production and consumption. For exam- trade liberalization on the environment, and what ple, suppose that an East Asian developing economy follows is only a selective and brief review. has abundant labor and therefore has a comparative The most prominent study is by Antweiler, advantage in labor-intensive goods. Freeing up trade Copeland, and Taylor (2001), covering the effects of would raise the price of labor-intensive goods in this trade liberalization on air pollution in 44 coun- economy and shift capital and labor into production tries.29 They find that international trade generates of these goods. If these goods are produced with relatively minor changes in concentrations of air cleaner technologies on average, overall emissions pollution when it alters the composition of output, will fall as the composition of output shifts away but that the associated technique and scale effects from dirtier goods. (Note that emissions could rise if reduce pollution. Overall, they find that within the goods that a country exports are dirtier; compar- their sample greater openness to trade actually ative advantage is an important factor.) The shift in reduces air pollution on average. the composition of output after trade liberalization Beghin and others (2000) consider the relation- raises the aggregate productivity in the economy, ships between trade liberalization and pollution in and thereby raises real GDP. The scale effect of trade Chile. In their computable general equilibrium liberalization is thus to increase pollution. Finally, as model, unilateral liberalization substantially wors- aggregate and per capita incomes rise from freer ens air pollution by providing cheaper and dirtier trade, so does the demand for a cleaner environ- energy sources. But if this trade policy were com- ment.26 Thus the technique effect of trade liberaliza- bined with an appropriate tax on emissions, Chile tion is to improve the environment by reducing would reap significant welfare gains.30 emissions per unit of output.27 Dean (2002) analyzes the impacts of trade liber- The same framework can be used to assess the alization on water pollution in Chinese provinces effects of an increase in foreign direct investment during 1987­95, a period in which there was both on environmental quality. If multinational firms an extensive pollution levy system and a significant exist disproportionately in dirtier industries, their opening to trade. She finds that trade liberalization arrival in a country has an effect like an increase in has aggravated environmental damage, because the capital stock that worsens the environment China has a comparative advantage in pollution- through the composition effect. But if they exist intensive goods, but that greater openness has also disproportionately in cleaner industries, their raised per capita incomes, mitigating the environ- arrival can improve the environment. By expanding mental costs through stronger regulation. She also economic activity, multinational firms would gen- finds that emissions per unit of industrial output in erate a scale effect that increases the effects of firms' China would have been much higher without trade activities on the environment.28 Perhaps most fun- reform, so that China's opening to trade was benefi- damentally, the operations of multinational firms cial for the environment overall. can generate beneficial technique effects. For a vari- Arunanondchai (2001) assesses the impacts that ety of reasons, they are likely to transfer cleaner cuts in tariffs and export taxes on logging and tim- technologies, developed in their home economies, ber products would have on Indonesian and to recipient countries (Moran 2002). If they pay Malaysian exporters. She finds that trade liberaliza- higher-than-average wages without generating off- tion would not necessarily raise log production, setting unemployment, the impact would be higher because it might not raise net producer prices, but per capita incomes and an induced demand for that foreign tariff cuts in plywood and sawn lumber stronger environmental protection. Thus, even if would generate significant gains for Indonesia. She multinational firms do not alter much the compo- also finds that removal of export taxes on logs sition of output, one would expect their operations would reduce world log prices, tending to worsen to improve environmental stewardship. In this con- the joint welfare of exporting nations. text, economies that are more open to FDI, other APEC (1999) assesses the effects of selected non- things being equal, would tend to have stronger tariff restraints on trade in forest products. The environmental standards and cleaner technologies. authors find that multilateral removal of taxes and However, this is an empirical issue. subsidies would generate some gains for timber 126 East Asia Integrates exporters, but they claim that the data were insuffi- Levinson and Taylor (2001) find that in the cient to assess the environmental impacts of such United States during the 1970s and 1980s, the liberalization. largest relative increases in sectoral net imports In summary, the literature suggests it is not pos- were in those industries experiencing the largest sible to predict the implications of trade liberaliza- increases in environmental control costs. Thus tion for use of the environment and economic wel- trade flows do react to such regulation. Unfortu- fare in general, but the balance of econometric and nately, because there are no comparable measures empirical evidence suggests that it can be benefi- in developing countries, it is impossible to deter- cial. Studies of the effects of FDI on environmental mine whether this result holds more widely. damage are surprisingly scarce.31 Moran (2002) has The literature gives little support for the hypoth- provided some anecdotal evidence that multina- esis that FDI responds to international variations in tional firms are associated with improvements in environmental costs. For example, Eskeland and use of the environment, but such improvements Harrison (1997) find almost no evidence that depend on the industry. multinational firms investing in developing coun- tries are attempting to escape higher environmental costs in their home countries. They also find that Evidence on Environmental Standards and foreign-owned plants in developing countries are Competitiveness less polluting than comparable domestic plants. The "pollution haven hypothesis" has attracted Wheeler (2001) discusses why the "race to the bot- much study. What is the evidence that weak envi- tom" idea in environmental standards makes little ronmental protection either generates an export sense in terms of attracting FDI. He finds that advantage or attracts FDI? In principle, one might indexes of air pollution actually improved expect weak regulation to be more strongly identi- markedly in the major cities in China, Brazil, and fied with trade performance and FDI for the envi- Mexico during an era of extensive investment ronment than for labor standards. After all, unlike inflows. Smarzynska and Wei (2001) examine firm- workers, the environment cannot complain or level data in 24 transition economies, controlling shirk when it is treated badly. Therefore, the ability for corruption levels, and find little support for the to exploit it without regulation could reduce costs, pollution haven idea (see also Jaffe and others at least up to the limits from congestion. 1995). Levinson (1996) finds no evidence that dif- ferences in environmental standards across states Past Evidence. This prediction is borne out in a 24- affect the location choices of manufacturing plants. country study of environmental standards and This view is not universal, however. Lucas, trade by Wilson, Otsuki, and Sewadeh (2002).32 Wheeler, and Hettige (1990) claim that increasingly Their results suggest that more stringent environ- strict environmental regulations in OECD coun- mental laws reduce net exports of pollution-inten- tries led to the displacement of pollution-intensive sive goods. Thus weak environmental regulation is industries. List and Co (1999) find that FDI in the associated with higher trade performance. More- United States was negatively related to regulatory over, they find that a hypothetical trade agreement expenditures per firm. Mani, Pargal, and Huq to harmonize environmental protection laws at lev- (1997) discover that in India spending on environ- els higher than those in developing countries would mental damage abatement was higher in more pol- reduce trade by up to 11 percent a year. lution-intensive industries and was a factor in plant Their study is unique in its findings, however. location decisions. Keller and Levinson (2002) Earlier studies discovered no evidence that a coun- point out problems with endogeneity in such mea- try with stricter environmental standards would sures. The issues are inherently empirical. have lower exports of pollution-intensive goods (Tobey 1990; Low and Yeats 1992; Xu 1999). Most Evidence from East Asia. Available measures sug- observers have concluded that, because environ- gest that environmental protection regulations are mental controls typically amount to a small per- not very stringent in East Asia. Table 7.4 lists two centage of total costs, they are relatively unimpor- indexes of environmental regulation in several tant in determining trade patterns. developing countries. The first is an environmental Trade and Competitiveness Aspects of Environmental and Labor Standards in East Asia 127 sustainability index, which is a compilation of sev- widely across these economies.33 Indonesia pro- eral categories involving resource use, pollutant vides the most straightforward support for the idea concentrations, and infrastructure. The correlation that the financial crisis pushed the composition of of environmental sustainability with per capita output toward pollution-intensive sectors--a shift GDP is positive and significant at 0.37, but well toward dirtier industries took place in its total below unity. The environmental sustainability exports and its exports to the European Union and index is correlated positively with the political free- the United States. This shift is consistent with the dom ranking in Table 7.2, but only weakly. The sec- fact that Indonesia's output of dirty goods rose rel- ond environmental policy index measures the ative to that of other goods during the crisis (World stringency of environmental regulations. This Bank 2001b). The export share of dirty industries index too is positively correlated with per capita also rose in Korea from 1997 to 1998 before falling GDP, but it is not correlated with the measure of again in 1999, and there is some indication of simi- political freedom. lar increases in the Philippines and Thailand in One might expect a reduction in costs resulting 1999.34 from weaker, or more weakly enforced, environ- To investigate further the relationship between mental standards to be reflected in increases in the environmental protection and exports of pollu- share of pollution-intensive goods in manufactured tion-intensive goods, I undertook an econometric exports for East Asian economies. But, in fact, analysis based on models whose essential structure trends in shares of pollution-intensive exports vary is the same as those applied earlier to the labor TABLE 7.4 Environmental Policy Indexes Environmental Environmental Sustainability Regulatory Regime PPP per capita Country Index, 2001 Index, 2001a GDP, 1999 (US$) Argentina 62.9 ­0.732 12,258 Brazil 57.4 ­0.077 7,130 Chile 56.6 0.177 8,720 Colombia 54.8 ­0.416 5,821 Peru 54.3 ­0.722 4,626 South Africa 51.2 ­0.029 8,994 Venezuela 50.8 ­1.079 5,586 Malaysia 49.8 ­0.127 8,215 Egypt 46.4 ­0.224 3,423 Turkey 46.3 n.a. 6,374 Mexico 45.3 ­0.602 8,329 Thailand 45.2 ­0.389 6,095 Pakistan 43.4 n.a. 1,813 Indonesia 42.5 ­0.758 2,873 India 40.7 ­0.759 2,250 Korea, Rep. of 40.3 ­0.121 15,778 China 37.5 ­0.348 3,620 Philippines 35.6 ­1.014 3,803 Correlation with GDP per capita 0.37 0.41 Correlation with Freedom House Index, 1998­99 0.17 ­0.03 n.a. Not applicable. Note: PPP = purchasing power parity. a. Most values are negative because the measure is scaled across all countries (including OECD members) to be zero for the average regime. Sources: Sustainability and regulatory regime indexes are from Esty and Cornelius (2002). GDP data are from World Bank (2001c). 128 East Asia Integrates standards.35 The two measures of environmental Overall, it is difficult to reach any confident con- standards used are the environmental sustainability clusions from the results of the numerical analysis. index and the environmental regulatory regime From the changes in export shares by country, it index shown in Table 7.4. Also included are the appears that those countries hardest hit by the index of political freedom, a dummy for joint Asian crisis tended to shift more output and importer-exporter membership in the North exports into pollution-intensive sectors. But the American Free Trade Agreement (NAFTA), and econometric analysis provides little evidence that regional fixed effects. The results are shown in weaker environmental standards are associated Table 7.5. with higher exports of pollution-intensive goods. TABLE 7.5 Impacts of Environmental Standards on Pollution-Intensive Exports from Developing Countries Model 1 Model 1 Model 2 Model 2 Variable (1995) (1999) (1995) (1999) NAFTA 2.38*** 2.33*** 2.33*** 2.82*** Political freedom 0.38*** ­0.82 ­0.35*** ­0.08 East Asia 15.8*** 25.7*** ­2.12*** ­1.25*** South Asia ­237.4*** ­247.6*** 0.18 ­0.09 Latin America 21.2*** 29.8*** ­2.44*** ­2.95*** Environmental 0.46*** 0.61*** sustainability (ES) (21.6) (28.7) ES · East Asia ­0.30*** ­0.51*** (16.1) (19.3) ES · South Asia 5.85*** 6.17*** (51.9) (60.6) ES · Latin America ­0.46*** ­0.64*** (13.2) (16.9) Environmental regime (ER) 16.0*** 14.9*** (8.32) (7.75) ER · East Asia ­16.3*** ­14.1*** (5.01)a (4.44)a ER · South Asia ­13.1*** ­10.9*** (7.57) (7.00) ER · Latin America ­14.2*** ­13.9*** (5.44) (4.87)a Number of observations 2,880 2,880 2,880 2,880 R­squared 0.55 0.58 0.56 0.58 * Significant at 10 percent level. ** Significant at 5 percent level. *** Significantly different from zero at 1 percent level. Note: Each equation is estimated as an augmented gravity model, with controls for importer and exporter GDP, importer and exporter population, bilateral distance, and industry fixed effects. Standard errors are robust to heteroskedasticity. Figures in parentheses are elasticities calculated at sample means. Pollution-intensive goods are ISIC (International Standard Industrial Classification) 332, 341, 351, 352, 353, 356, 371, and 372. Exporters are Argentina, Brazil, Chile, Colombia, Mexico, Peru, Venezuela (Latin America); Bangladesh, India (South Asia); China, Indonesia, Republic of Korea, Malaysia, the Philippines, Thailand, Vietnam (East Asia); Egypt, and South Africa. Importers are Australia, Austria, Canada, Denmark, Finland, France, Germany, Greece, Japan, Ireland, Italy, the Netherlands, New Zealand, Norway, Portugal, Spain, Sweden, Turkey, United Kingdom, and United States. a. Not significantly different from zero at the 10 percent level. Source: The author. Trade and Competitiveness Aspects of Environmental and Labor Standards in East Asia 129 At best, there is no strong or systematic indication can take place only after prior written notifica- that the crisis significantly augmented exports tion by the exporting country authorities to the based on pollution havens in East Asia. competent authorities of the importing country and to countries through which they might transit. Each shipment of hazardous waste must Environmental Protection and the then be accompanied by documentation or be Doha Round deemed illegal. Exports of toxic wastes to certain The Doha Round of multilateral trade negotia- countries are banned outright. Cross-border tions at the World Trade Organization (WTO) movements can take place, however, if the would be the first round to consider linkages exporting nation lacks the capacity to manage or between environmental protection and trade. dispose of the hazardous waste in an environ- Many developing countries have raised concerns mentally sound manner. The Basel Convention about the potential implications of environmental clearly anticipates that importing governments protection for their trade opportunities. This sec- may closely regulate the disposal of hazardous tion reviews the environmental provisions of the wastes because they must agree to the documen- Doha Declaration and the scope of the upcoming tation and may monitor execution of the trans- negotiations, and comments on potential obliga- action. The convention falls short of permitting tions for East Asian economies. individual countries to ban such imports, The Doha Declaration calls on countries to though a proposed amendment provides for countries to ban exports. Whether importers · Clarify relationships between WTO rules and could issue such bans presumably depends on a trade obligations under various multilateral envi- legal interpretation of Article XX of the General ronmental agreements (MEAs).36 Further clarifi- Agreement on Tariffs and Trade (GATT), which cation is needed about whether trade sanctions permits trade restraints in order to protect envi- issued under the Montreal Protocol, the Con- ronmental and human health. However, such vention on International Trade in Endangered regulation must be consistent with national Species (CITES), and other agreements are con- treatment, so that imported waste is not treated sistent with WTO obligations. To date, such worse in the commercial sense than is domestic sanctions have been effectively exempted from waste. In poor countries, waste treatment tends the WTO, and this exemption is likely to become to be weakly regulated, suggesting that import a formal interpretation in the Doha Round. It is limitations could be problematic under WTO conceivable that a broad interpretation could rules. limit export opportunities in certain goods in · Reduce or eliminate trade barriers to environmen- the future, even as it gives East Asian countries tal goods and services. Such barriers raise costs to greater scope to limit imports. However, accord- countries that must import technology if they ing to the Doha Declaration, negotiations on wish to mitigate environmental problems. such linkages must be undertaken without alter- Obligations to cut import tariffs and open ser- ing the WTO rights of countries that are not vice markets may run into political economy members of an associated MEA. Thus there concerns, but they can effectively complement should be little reason for concern about facing domestic environmental policy. trade restrictions by virtue of not joining MEAs. · Clarify rules on the effects of environmental regu- As for toxic waste disposal, developing coun- latory measures on market access. These rules tries need to know whether current or potential seem to be aimed mainly at efforts by developed WTO rules allow them to ban imports of toxic and middle-income countries to restrict trade in wastes and products. A clarification of restraints order to achieve (indirectly) some environmen- under the Basel Convention and WTO obliga- tal goal.37 It remains to be seen whether such tions is needed. Under the Basel Convention, clarification would expand the scope of unilat- which was ratified or acceded to by all major eral action to restrict trade or whether it would East Asian economies by 2000, transboundary impose additional obligations on countries that movements of hazardous wastes or other wastes limit imports. 130 East Asia Integrates · Identify cases where reducing trade restrictions stewardship. If anything, it finds a continuing trend and economic distortions would benefit the envi- toward stronger regulation. However, public expen- ronment, trade, and economic development. ditures devoted to social safety nets for workers and Prominent examples would be subsidies to coal to enforcement of environmental laws suffered in and other forms of carbon-based energy and to some countries affected by the 1997­98 financial water use in agriculture. Reductions in such sub- crisis, and there is anecdotal evidence of declines in sidies bear considerable promise for improving working conditions and environmental protection use of the environment while raising market during the crisis. access to new technologies and agricultural Unfortunately, it is quite difficult to trace products (Anderson 1996). Negotiations on this whether these processes have boosted exports. Eco- point should give East Asian economies an nomic theory and evidence from past studies sug- opportunity to rationalize their trade policy gest that weak labor rights are not correlated with with resource subsidies and other policies. export performance, even in labor-intensive goods, · Clarify provisions of the intellectual property and that they are negatively correlated with an agreement in the WTO that could affect the trade economy's attractiveness to foreign direct invest- and transfer of environmental technologies. There ment. They also suggest that weak environmental are concerns that stronger private property regulations are not a significant determinant of rights that exclude unauthorized use--includ- exports or FDI in most circumstances. Given this ing by governments--of technologies, products, background, it is not surprising that the economet- and services able to reduce environmental prob- ric work performed for this study finds no indica- lems might make environmental management tion that weak labor rights or environmental stan- more costly. dards are positively correlated with export · Consider the scope for environmental labeling performance. There is evidence on the labor side requirements on products. Effective labeling that stronger rights are associated with higher requirements could be beneficial in restraining exports of labor-intensive goods, although that the use of blunter trade restrictions, although finding depends on the measure of labor standards meeting the costs of labeling could be high. used. Environmental protection indexes also seem to be positively but weakly correlated with East Finally, the declaration pledges that developed Asian exports in pollution-intensive goods. In sum- countries will commit to increasing technical and mary, although it is difficult to be confident about financial assistance for meeting international envi- such inferences, I could find no significant evidence ronmental needs. in the data that weak labor rights or environmental This set of negotiating objectives aims to address standards boost East Asian exports. the existing inconsistencies in and shortcomings of If the results are taken seriously, important pol- the trading system rather than permit the introduc- icy messages emerge. East Asian developing coun- tion of extensive new trade control regimes. There- tries have weaker measured labor and environmen- fore, East Asian developing economies should not be tal standards, relative to per capita income, than do overly concerned about the introduction of environ- other developing regions. To the extent that these mental issues into the WTO, because those issues are policies have supported flexibility in labor markets, already included in various ways. The scope of nego- they may be beneficial for export competitiveness. tiations as set out in this section would not raise However, these countries have considerable room much concern about market access, although devel- for improving their protection of fundamental oping country members should perhaps be particu- labor rights and environmental standards. Doing so larly wary about labeling requirements. would not reduce their ability to export labor- intensive goods or pollution-intensive goods; indeed, export growth can be compatible with rais- Policy Implications ing core labor standards and environmental protec- The analysis in this chapter finds little evidence that tion. In short, the developing economies of East East Asian governments have relaxed their stan- Asia can feel comfortable in applying greater dards on working conditions or on environmental weight to social considerations in setting their poli- Trade and Competitiveness Aspects of Environmental and Labor Standards in East Asia 131 cies without significant fear of deterring export may not be strictly comparable, they do not suggest that growth. child labor use significantly increased during the crisis. Edmonds and Pavcnik (2002), who studied household sur- veys in Vietnam, discovered that an increase in the real price of rice in the 1990s, associated largely with a decision to Endnotes phase out the export quota on rice, raised net incomes of 1. Such a situation can arise directly if, for example, absence rural households enough to reduce the use of child labor of collective bargaining by workers permits monopsonistic sharply. By their estimates, a 30 percent increase in the price employers in particular regions to restrain wages and out- of a kilogram of rice led to a decline of nine percentage put (Martin and Maskus 2001). points in the use of child labor. The children, especially sec- 2. For extensive reviews, see Maskus (1997), OECD (1996, ondary school­age girls, were able to go to school instead. 2000), and Elliott (2001). 12. EPZs are regions within an economy that offer firms-- 3. The right to strike is not generally considered a core labor often multinational enterprises (MNEs)--tax advantages, standard. free importation of inputs (or duty drawbacks), infrastruc- 4. Many such benefits may be provided by employers volun- ture development, and other inducements to locate there tarily or through collective bargaining arrangements. Thus and produce goods for export (Madani 1999; Moran 2002). government regulation may simply provide a floor for 13. Romero (1995) reports extensive evidence that wages are these rights. significantly higher in manufacturing jobs within EPZs than 5. An across-the-board set of requirements might be offset by in similar positions outside. Moran (2002) provides ample depreciation of the home currency, with little net impact on anecdotal evidence that MNEs pay wages 20­25 percent international competitiveness (Sykes 1995). The claim that higher than those paid by domestic firms for similar jobs. strong civic and security rights may raise costs is not univer- He also describes a process of continuing improvements in sally accepted; institutional labor economists argue that the 1990s in safety conditions, minimum wages, and bene- such rights increase the attachment of workers to firms and fits packages within several EPZs, including two (Mactan skills, which can lower costs. It is an empirical question. and Baguio City) in the Philippines (also see Jacobson 6. Note, however, that ratification signals only an intention to 1999). Drezner (2001) reviews theories of why multina- comply with a convention, not a country's actual labor mar- tional firms are likely to be a force for raising standards. ket policies. The ILO has little scope for enforcing these 14. See the discussion in the technical appendix to this chapter rights. Conversely, a decision not to ratify does not mean at www.worldbank.org/eaptrade. that the inherent rights in a convention are denied. Govern- 15. Mah (1997), who studied 45 developing countries, found ments may provide such rights while objecting to the lan- that the ratio of exports to gross domestic product (GDP) guage of a particular convention. The United States, for correlated negatively with decisions to ratify fundamental example, has ratified only two of these conventions (105 conventions on freedom-of-association rights and rights to and 182), but its practices largely comply with ILO expecta- nondiscrimination. But the lack of control variables in the tions. A review of labor standards in eight East Asian coun- equations estimated renders these results questionable. tries appears in the technical appendix to this chapter at 16. Figures on the shares of labor-intensive manufactured www.worldbank.org/eaptrade, the Web site for this volume. exports in 1995­99 for eight East Asian economies are in 7. According to information on the ILO Web site, between Appendix Figure 1 of the technical appendix to this chapter 1997 and 2001 unions issued two complaints about govern- at www.worldbank.org/eaptrade. ment practices in China, one in Hong Kong, one in Cam- 17. For example, the growth in volume of exports of apparel bodia, one in Indonesia, two in Korea, one in the Philip- and textiles to key markets may have been constrained by pines, one in Thailand, and none in Malaysia, Singapore, or quotas. It is also possible that the price impacts of depreci- Vietnam. ating exchange rates were relatively greater for labor-inten- 8. In the Philippines, the gross primary school enrollment sive goods, tending to reduce their values if not volumes. rate, which is strongly and negatively correlated with child 18. Details of this analysis appear in the technical appendix to labor, fell from 99.2 percent in 1997­98 to 98.1 percent in this chapter at www.worldbank.org/eaptrade. 1998­99 (Lim 2000). Overall, however, there is no evidence 19. The results using the third measure differ from those using of a significant increase in child labor use during the crisis the first and second ones. Close consideration of the three in the Asia-Pacific region. measures provides insight. The OECD measure of free- 9. In export-oriented, labor-intensive sectors in labor-abun- dom-of-association rights is directly keyed on a central and dant countries that have weak labor standards, conditions fundamental basic labor right, and the Verite Labor Stan- of work would deteriorate. However, if the effect of trade dards Index (VERLS), though it combines four approaches, liberalization is to draw workers from the informal sector, incorporates enforcement and effective recognition of four net conditions could improve. Overall, whether the infor- core labor standards, and thus also focuses on basic rights. mal sector expands or contracts would depend on the cir- Both of these measures are suited to analyze the impacts of cumstances (Maskus 1997). core labor protection on export performance, and these 10. See Dollar and Collier (2001) and McCulloch, Winters, and impacts appear to be positive. On its face, the third mea- Cirera (2001) for an extensive discussion. sure--the number of ILO ratifications--should also be an 11. Dollar and Collier (2001) and Dollar and Kraay (2002) are indicator of core labor rights, but, as explained earlier, it is a two of many sources on the effects of more open trade on questionable measure of a country's actual commitment to East Asia. The child labor force participation rate in the East improved working conditions. Asia­Pacific Region fell from 21.5 percent in 1995 to 19 per- 20. Hasan and Quibria (2002) reach a similar conclusion, cent in 2000, despite the economic crisis (UNICEF 1999; although their focus is on the determinants of poverty ILO 2002). Although for various reasons these estimates reduction rather than trade competitiveness. 132 East Asia Integrates 21. The Korean government reduced its environmental expen- Project over the period 1971­96, they decompose emis- ditures from 2.5 percent of total expenditures in 1997 to 2.3 sions into scale (GDP), composition (capital-labor endow- percent in 1998, although this reduction was not especially ment ratios), and technique (real income) effects, interact- severe in relation to those for other programs (World Bank ing each with a measure of openness to trade. 2000). 30. This insight underscores an important observation about 22. Copeland and Taylor (2001) provide a clear presentation of the nexus between trade policy and environmental policy. this framework. In general, if environmental distortions are internalized 23. It is assumed that net output (total output less resources efficiently, open trade enhances welfare. devoted to pollution abatement) is produced with constant 31. It is possible they exist, but I could not locate any. returns to scale. 32. They regressed net exports in five pollution-intensive 24. One important reason that a government would increase industries on measures of factor endowments and environ- the pollution tax (or generally increase environmental reg- mental laws in 24 countries (6 OECD countries, including ulation) is that voters demand a cleaner environment as Korea, and 18 developing countries) over the period incomes rise, justifiably assuming that environmental 1994­98. health is a normal good. Thus in the applied literature, the 33. Figures on shares of pollution-intensive manufactured term technique effect generally refers to the idea that any- exports are in Figure 7.1 in the technical appendix to this thing that raises per capita income generates an endoge- chapter at www.worldbank.org/eaptrade. nous increase in environmental taxes, thereby reducing the 34. Grossman and Krueger (1993) found no contribution of pollution intensity of production. In this sense, the term is U.S. pollution intensity to U.S. imports from Mexico. related to the so-called Kuznets environmental curve. How- 35. It is noteworthy that those countries experiencing the ever, it also refers to autonomous improvements in envi- greatest stress from the crisis--Indonesia, Korea, the ronmental technologies. Philippines, and Thailand--registered these increases in 25. It could also mean domestic deregulation of services and export shares. This tendency was not evident in Malaysia changes in other policies, such as intellectual property and Taiwan (China). Again, such calculations are only sug- rights, the effects of which are left aside here. gestive, because many factors other than declining environ- 26. Per capita income would be higher despite changes in the mental costs or weak environmental standards could distribution of income. explain such changes in export shares. Three comments are relevant here. First, the impacts of 36. Details of this analysis appear in the technical appendix to openness to trade are less clear when account is taken of this chapter at www.worldbank.org/eaptrade. input flows. Because inputs themselves may be dirty or 37. One important MEA is the Convention on Biodiversity, clean, there is an additional comparative advantage effect to whose commitments on the exploitation of genetic consider. (For example, if trade liberalization increases resources are at odds with rights established under the access to imported coal and petroleum, the net effect may be intellectual property agreement in the WTO. The Conven- expanded output of dirty final goods even if there is a direct tion on Biodiversity does not itself implicate trade policy comparative disadvantage in them. Moreover, tariff cuts directly, but a recently added provision, the Protocol on could expand the consumption of cheaper imported fuels.) Bio-safety, essentially permits countries to exclude imports Second, many analysts suggest that as incomes rise there will on the basis of the precautionary principle. By permitting be an incentive to improve environmental technologies unilateral import restraints that could be more rigorous endogenously as a result of the higher demand for environ- than those in the WTO's Agreement on the Application of mental protection associated with rising incomes. One Sanitary and Phytosanitary Measures, this policy could direct means by which this could come about is the greater negatively affect exports of genetically modified foodstuffs access that more open economies have to foreign technolo- from developing countries. For the East Asian countries, gies. In this context, trade openness provides an endogenous such exports are not significant, so this issue may be of lim- technique effect that could be decisive in improving envi- ited importance unless the principle is extended to other ronmental stewardship. Third, for the standard technique forms of production processes. effect to operate, citizen demand for a cleaner environment 38. One example is the shrimp-turtle case, in which the United must be mediated effectively through an increase in the cost States threatened to limit imports of shrimp from certain of polluting the environment. In some cases, this mediation Southeast Asian countries failing to use turtle-excluder takes place through informal pressures (Wang and Wheeler devices in their fishing fleets. 1999), but, generally, the government must respond to citi- zen preferences and raise environmental charges as incomes grow. Therefore, once again the openness and efficiency of References the political process matter for environmental policy, as they do for labor standards. Barrett and Graddy (2000) find that increases in political freedoms significantly improve envi- The word processed describes informally reproduced works that ronmental quality. may not be commonly available through libraries. 27. It is conceivable that trade openness in rich countries con- tributes to better environmental quality in developing Aggarwal, Mita. 1995. "International Trade, Labor Standards, country exporters by moving workers out of environmen- and Labor Market Conditions: An Evaluation of the Link- tally harmful rural activities. However, this hypothesis has ages." Working Paper 95-06-C. U.S. International Trade not been tested. Commission, Washington, D.C. 28. Not all would agree that FDI expands economic activity as Anderson, Kym. 1996. "The Intrusion of Environmental and much as it eliminates domestic competition. Labor Standards into Trade Policy." In Will Martin and L. 29. Using data on sulfur dioxide concentrations in 293 sites in Alan Winters, eds., The Uruguay Round and the Developing 44 countries from the Global Environment Monitoring Countries. Cambridge: Cambridge University Press. Trade and Competitiveness Aspects of Environmental and Labor Standards in East Asia 133 Antweiler, Werner, Brian R. Copeland, and M. Scott Taylor. 2001. Hasan, Rana, and M. G. Quibria. 2002. "Poverty and Patterns of "Is Free Trade Good for the Environment?" American Eco- Growth." ERD Working Paper No. 18. Asian Development nomic Review 91: 877­908. Bank, Manila. APEC (Asia Pacific Economic Cooperation). 1999. Study of Non- ILO (International Labour Organisation). 2002. Every Child Tariff Measures in the Forest Products Sector. Committee on Counts: New Global Estimates of Child Labor. Geneva. Trade and Investment. Singapore: APEC. Jacobson, Mark. 1999. "The Philippines: The Case of Economic ------. 2001. Survey of Environmental Markets in APEC. Com- Zones." In Alleviating Poverty through Foreign Direct Invest- mittee on Trade and Investment. Singapore: APEC. ment: Country Case Studies. Washington, D.C.: World Bank. Arunanondchai, May. 2001. "Trade Policy and the Welfare of Jaffe, A., S. Peterson, P. Portney, and R. Stavins. 1995. "Environ- Southeast Asian Timber Exporters: Some Implications for mental Regulation and the Competitiveness of U.S. Manu- Forest Resources." Research Report. Economy and Environ- facturing. What Does the Evidence Tell Us?" Journal of Eco- ment Program for Southeast Asia. nomic Literature 33: 132­63. Barrett, Scott, and Kathryn Graddy. 2000. "Freedom, Growth, Keller, Wolfgang, and Arik Levinson. 2002."Environmental Reg- and the Environment." Environment and Development Eco- ulations and FDI to U.S. States." Review of Economics and nomics 5: 433­56. Statistics 84, no. 4. Beghin, John C., Bradley J. Bowland, Sebastien Dessus, David Kucera, David. 2001. "The Effects of Core Workers' Rights on Roland-Holst, and Dominique van der Mensbrugghe. 2000. Labor Costs and Foreign Direct Investment: Evaluating the "Trade Integration, Environmental Degradation, and Public `Conventional Wisdom.'" ILO Discussion Paper Health in Chile: Assessing the Linkages." Environment and DP/130/2001. International Labour Organisation, Geneva. Development Economics 7: 241­67. Levinson, Arik. 1996. "Environmental Regulation and Manufac- Betcherman, Gordon, and Rizwanul Islam, eds. 2001. East Asian turers' Location Choices: Evidence from the Census of Man- Labor Markets and the Economic Crisis. Washington, D.C.: ufactures." Journal of Public Economics 62: 5­29. World Bank. Levinson, Arik, and M. Scott Taylor. 2001. "Trade and the Envi- Copeland, Brian R., and M. Scott Taylor. 2001. "International ronment: Unmasking the Pollution Haven Effect." George- Trade and the Environment: A Framework for Analysis." town University, Washington, D.C. Processed. NBER Working Paper 8540. National Bureau of Economic Lim, Joseph Y. 2000. "The East Asian Crisis and Child Labor in Research, Cambridge, Mass. the Philippines." ILO/IPEC Working Paper 2000. Available at Dean, Judith M. 2002. "Does Trade Liberalization Harm the http://www.ilo.org/public/english/standards/ipec/publ/pol- Environment? A New Test." Canadian Journal of Economics icy/papers/philippines/. 35 (4): 819­42. List, John. A., and Catherine Y. Co. 1999. "The Effects of Envi- Dollar, David, and Paul Collier. 2001. Globalization, Growth, and ronmental Regulations on Foreign Direct Investment." Jour- Poverty: Building an Inclusive World Economy. Oxford: nal of Environmental Economics and Management 40: 1­20. Oxford University Press. Low, Patrick, and Alexander Yeats. 1992. "Do Dirty Industries Dollar, David, and Art Kraay. 2002. "Growth Is Good for the Migrate?" In Patrick Low, ed. "International Trade and the Poor." Journal of Economic Growth 7 (3): 195­225. Environment." World Bank Discussion Paper 159. Washing- Drezner, Daniel W. 2001. "Fears of a `Race to the Bottom' on ton, D.C. Wages and Labor Standards: Why Aren't They Justified?" Lucas, R. E. B., D. Wheeler, and H. Hettige. 1990. "Economic University of Chicago. Processed. Development, Environmental Regulation, and the Interna- Edmonds, Eric, and Nina Pavcnik. 2002. "Does Globalization tional Migration of Toxic Industrial Pollution." Paper pre- Increase Child Labor? Evidence from Vietnam." NBER sented at Symposium on International Trade and the Envi- Working Paper 8760. National Bureau of Economic ronment, Washington, D.C. Research, Cambridge, Mass. Madani, Dorsati. 1999. "A Review of the Role and Impact of Elliott, Kimberly A. 2001."Finding Our Way on Trade and Labor Export Processing Zones." Policy Research Working Paper Standards?" International Economic Policy Briefs no. 01-5. 2238. World Bank, Washington, D.C. Institute for International Economics, Washington, D.C. Mah, Jai S. 1997. "Core Labor Standards and Export Perfor- Eskeland, Gunnar S., and Ann E. Harrison. 1997. "Moving to mance in Developing Countries." World Economy 20: Greener Pastures? Multinationals and the Pollution Haven 773­85. Hypothesis." World Bank, Washington, D.C. Processed. Mani, M., S. Pargal, and M. Huq. 1997. "Does Environmental Esty, Daniel C., and Peter K. Cornelius. 2002. Environmental Per- Regulation Matter? Determinants of the Location of New formance Measurement: The Global Report 2001­2002. Manufacturing Plants in India." Policy Research Working Oxford: Oxford University Press. Paper 1718. World Bank, Washington, D.C. Fields, Gary S. 1995. Trade and Labor Standards: A Review of the Marinova, Nadja. 1999. "Indonesia's Fiery Crises." Journal of Issues. Paris: Organisation for Economic Co-operation and Environment and Development 8: 70­81. Development. Martin, William J., and Keith E. Maskus. 2001. "The Economics Freeman, Richard B. 1994."A Hard-Headed Look at Labor Stan- of Core Labor Standards: Implications for International dards." In W. Sengenberger and D. Campbell, eds., Interna- Trade Policy." Review of International Economics 9: 317­28. tional Labor Standards and Economic Interdependence. Maskus, Keith E. 1997. "Should Core Labor Standards Be Geneva: International Labour Organization. Imposed through Trade Policy?" Policy Research Working Grootaert, Christian, and Ravi Kanbur. 1995. "Child Labor: An Paper 1817. World Bank, Washington, D.C. Economic Perspective." International Labour Review 134: McCulloch, Neil, L. Alan Winters, and Xavier Cirera. 2001. Trade 187­203. Liberalization and Poverty: A Handbook. London: Centre for Grossman, Gene M., and Alan B. Krueger. 1993."Environmental Economic Policy Research. Impacts of a North American Free Trade Agreement." In Moran, Theodore H. 2002. Beyond Sweatshops: Foreign Direct Peter Garber, ed., The U.S.-Mexico Free Trade Agreement. Investment and Globalization in Developing Countries. Wash- Cambridge, Mass.: MIT Press. ington, D.C.: Brookings Institution. 134 East Asia Integrates OECD (Organisation for Economic Co-operation and Develop- Verite. 2002."Report to California Public Employees' Retirement ment). 1996. Trade, Employment, and Labor Standards: A System (CalPERS): Emerging Markets Research Project." Study of Core Workers' Rights and International Trade. Paris. Available at verite@verite.org. ------. 2000. International Trade and Core Labor Standards. Wang, Hua, and David Wheeler. 1999. "Endogenous Enforce- Paris. ment and Effectiveness of China's Pollution Levy System." Rodrik, Dani. 1996."Labor Standards in International Trade: Do Development Research Group. World Bank, Washington, They Matter and What Do We Do about Them?" Overseas D.C. Development Council, Washington, D.C. Wheeler, David S. 2001. "Racing to the Bottom? Foreign Invest- Romero, Ana Teresa. 1995. "Labor Standards and Export Pro- ment and Air Pollution in Developing Countries." Journal of cessing Zones: Situation and Pressures for Change." Develop- Environment and Development 10: 225­45. ment Policy Review 13: 247­76. Wilson, John S., Tsunehiro Otsuki, and Mirvat Sewadeh. 2002. Smarzynska, Beata, and Shang-Jin Wei. 2001. "Pollution Havens "Dirty Exports and Environmental Regulation: Do Stan- and Foreign Direct Investment: Dirty Secret or Popular dards Matter to Trade?" Development Research Group. Myth?" CEPR Discussion Paper 2966. Centre for Economic World Bank, Washington, D.C. Policy Research, London. World Bank. 1999. "Environmental Implications of the Economic Suryahadi, Asep, Wenefrida Widyant, Daniel Perwira, and Crisis and Adjustment in East Asia." East Asia Environment Sudarno Sumarto. 2003. "Minimum Wage Policy and Its and Social Development Unit, Discussion Paper 01. Wash- Impact on Employment in the Urban Formal Sector." Bul- ington, D.C. letin of Indonesian Economic Studies 39 (1): 29­50. ------. 2000. Korea: The Environmental Dimension of the Crisis: Sykes, Allan. 1995. Product Standards in a Globalizing World. A Step Back of a New Way Forward? Washington, D.C. Washington, D.C.: Brookings Institution. ------. 2001a. East Asia Regional Overview Special Focus: Envi- Tobey, J. A. 1990. "The Effects of Domestic Environmental Poli- ronment. Washington D.C. cies on Patterns of World Trade: An Empirical Test." Kyklos ------. 2001b. Indonesia: Environment and Natural Resource 43: 191­209. Management in a Time of Transition. Washington, D.C. UNICEF (United Nations Children's Fund). 1999. The State of ------. 2001c. World Development Indicators 2001. Washington, the World's Children, 1999: Education. Geneva. D.C.: World Bank. U.S. Department of Labor. 1994. By the Sweat and Toil of Chil- World Resources Institute. 2002. Reefs at Risk in Southeast Asia. dren. Vol. 1, The Use of Child Labor in U.S. Manufactured and Available at: www.wri.org/wri/reefsatrisk/reefriskseasia.html. Mined Imports. Washington, D.C. Xu, X. 1999. International Trade and Environmental Regulation: Van Beers, Cees. 1998. "Labor Standards and Trade Flows of A Dynamic Perspective. Huntington, N.Y.: Nova Science Pub- OECD Countries." World Economy 21: 57­73. lishers. Part III Reinforcing Social Stability through broad sharing of benefits 8 Household Welfare Impacts of China's Accession to the WTO Shaohua Chen Martin Ravallion China's recent accession to the World Trade Orga- and factor prices stemming from the accession. The nization (WTO) brings with it sharp reductions in approach respects the richness of detail that is import tariffs, quantitative restrictions, and export available from a modern integrated household sur- subsidies, with implications for the domestic struc- vey, allowing us to go well beyond the highly ture of prices and wages and thus for the welfare of aggregative types of analysis one often finds. Our Chinese households. Even if the trade reforms have analysis measures the expected impacts across the little effect on poverty and income distribution in distribution of initial levels of living, but it also the aggregate, the impacts may vary across house- examines the way in which the impacts vary by hold types and regions, with implications for the other household characteristics, including location design of compensatory policy responses. and demographic characteristics. A reasonably In China, the economic geography of poverty, detailed "map" of the predicted welfare impacts by and its interaction with the geographic diversity in location and socioeconomic characteristics there- the impact of policy reforms, is high on the domes- fore emerges. tic policy agenda. A policy analysis that simply The first section of this chapter briefly reviews averages such differences would miss a great deal of the approaches that typically have been used in what matters to the policy debate. In this chapter, analyzing the welfare impacts of trade reform. It is data from national household surveys for rural and followed by a summary of the approach taken in urban areas are used to measure and explain the this chapter. The penultimate section describes the welfare impacts on households of changes in goods results of our analysis, and our conclusions follow. The authors are grateful to Tamar Manuelyan-Atinc and Will Martin for their encouragement to undertake this task, to Elena Ianchovichina and Will Martin for the estimates of the price impacts of China's trade reform that are crucial to this study, and to Pingping Wang, Yan Fang, Liqun Peng, Honge Gong, and Min Yuan for their help in matching vari- ables from China rural/urban household surveys to the categories of the general equilibrium model. The comments of François Bourguignon, John Cockburn, Neil McCulloch, Sangui Wang, Shujiro Urata and participants in the Fourth Asian Development Forum in Seoul, the seminar at the Tokyo office of the World Bank, and the National Bureau of Sta- tistics in Beijing are gratefully acknowledged. 137 138 East Asia Integrates Measuring the Welfare Impacts of will have very different effects on inequality Trade Reform depending on the level of economic develop- ment--raising inequality in rich countries and The literature contains much debate about the wel- reducing it in poor ones.1 But the opposite out- fare impacts of greater openness to trade. Some come is possible when economic reforms, including authors argue that external trade liberalization trade liberalization, increase the demand for rela- benefits the poor in developing countries, while tively skilled labor, which may well be less equally others assert that the benefits are captured more by distributed in poor countries than in rich ones. the nonpoor. Assessments have focused particularly Regressions for inequality across countries show on the expected impacts on relative wages (notably some evidence of a negative interaction effect between skilled and unskilled labor) and relative between openness to trade and initial gross domes- prices (such as between food staples and luxury tic product (GDP) per capita (Barro 2000; Raval- imports). lion 2001; Milanovic 2002). What does the evidence suggest? One might In principle, these problems can be dealt with by hope to provide a conclusive answer by comparing introducing suitable nonlinearities (including changes over time in measures of inequality or interaction effects) into the regressions based on poverty between countries that are open to external compilations of country aggregates. However, the trade and countries that are not. Thus several stud- concerns go deeper. For example, although aggre- ies have combined levels of measured inequality, or gate inequality or poverty may not change with changes over time in measured inequality or trade reform, there may be both gainers and losers poverty, with data on trade openness and other at all levels of living. Indeed, when surveys have control variables, using aggregate cross-country tracked the same families over time, it is quite com- data sets (for example, see Bourguignon and mon to find that many people have escaped from Morisson 1990; Edwards 1997; Li, Squire, and Zou poverty, while others have fallen into poverty, even 1998; Lundberg and Squire 1999; Barro 2000; Dol- though the overall poverty rate may have moved lar and Kraay 2002; and Milanovic 2002). rather little.2 As another example, geographic dis- There are reasons, however, to be cautious in parities in access to human and physical infrastruc- drawing policy inferences from such studies. First, ture, between and within developing countries, concerns have emerged about data and economet- affect the prospects for participating in the growth ric specification. Differences in survey design and generated by reform, and these disparities tend to processing between countries, and over time within be correlated with incomes.3 A reform may well countries, can distort the measured levels and the entail sizable redistributions between the poor and apparent differences in inequality. Second, the the rich, but in opposite directions in different extent to which cross-country data sets can detect regions within countries. One should not be sur- any underlying effects of greater openness or other prised to find no correlation between growth and covariates is unclear. Another issue is whether the changes in inequality, or that, on average, a policy volume of trade should be treated as exogenous in reform has virtually no impact on inequality. Yet these cross-country regressions. It is clearly not a nonrandom distributional changes could well be policy variable as such, and it may well be highly going on beneath the surface of the average impact correlated with other (latent) attributes of country statistics. A policy analysis that simply averages performance independently of trade policy. Attri- over such differences would miss a great deal of bution of inequality impacts to trade policy what matters to the debate on policy. reforms per se is clearly problematic. This chapter follows a different approach for Furthermore, starting conditions vary widely which the attribution to trade policy changes is among reforming countries, and averaging across unambiguous and the diversity of welfare impacts this diversity can readily hide systematic effects of is not lost. It studies welfare impacts at the house- relevance to policy. For example, countries differ in hold level of changes in commodity and factor their initial levels of economic development. It has prices attributed to a specific trade policy reform-- been argued that greater openness to external trade China's accession to the WTO. Household Welfare Impacts of China's Accession to the WTO 139 Our Approach in This Chapter conomic "map" of impact, showing how it varies with other "non-income" characteristics, such as Past approaches to studying the welfare impacts of location. In this way, we are better able to answer specific trade reforms have tended to be either par- the questions policymakers are asking about who tial equilibrium analyses, in which the welfare gains and who loses from reform. impacts of the direct price changes resulting from In our analysis, we use past estimates of the tariff changes are measured at the household level, direct and indirect impacts of China's WTO acces- or general equilibrium analyses, in which second- sion on prices for both commodities and factors of round responses are captured in a theoretically production, as reported in Ianchovichina and Mar- consistent way, but with considerable aggregation tin (2002), and apply standard methods of first- across household types.4 Although partial equilib- order welfare analysis to measure the gains and rium analysis requires little or no aggregation of the losses at the household level.7 primary household data, it misses potentially The price changes induced by the trade policy important indirect effects on prices and wages. change are simulated from the CGE model used by General equilibrium analysis has the power to cap- Ianchovichina and Martin (2002). This competitive ture these effects by simulating the economy-wide market-clearing model originated in the Global impacts on markets. Standard computable general Trade Analysis Project (GTAP).8 In this model, the equilibrium (CGE) models, however, entail consid- revenue implications of the trade policy change are eration aggregation across household types, with reflected in changes in indirect tax rates. rarely more than six or so "representative house- The CGE model is applied to the data from large holds." Such models are crude tools for welfare dis- national sample surveys of households in urban tributional analysis. and rural areas of China carried out by the National The challenge for applied work is to find an Bureau of Statistics (NBS). The general equilibrium approach that respects the richness of detail avail- analysis generates a set of price and wage changes. able from a modern integrated household survey, These changes embody both the direct price effects while assuring that the price changes attributed to of the trade policy change and the "second-round" reform are internally consistent with economy- indirect effects on the prices of nontraded goods wide equilibrium conditions. Anyone addressing and on factor returns, including effects that make this challenge can in principle build the CGE model themselves felt through the government's budget onto the household survey, such that the number of constraint. Because the price changes are based on households in the model is the number sampled in an explicit model, their attribution to the trade pol- the survey.5 In our analysis, this degree of integra- icy reform is unambiguous, thereby avoiding the tion would require a CGE model of extraordinarily identification problems common to past attempts high dimensions, with 85,000 households (the total to estimate the distributional effects of trade policy sample size of the surveys we use). This route is not reform using cross-country comparisons. a feasible one at present. The welfare impacts are derived from a house- Instead, we follow an intermediate approach in hold model that incorporates own-production which the reform-induced commodity and factor activities. The CGE and household-level analyses price changes simulated from a general equilibrium are not integrated--with 85,000 households in the model are carried to the level of all the sampled survey; doing so would require an extraordinarily households in the survey.6 In measuring the welfare large CGE model. The microsimulations are built impacts, we use standard tools of analysis familiar on economic assumptions that are consistent with from past work on the welfare effects of price the CGE model--notably, that households take changes associated with tax and trade policy prices as given and that those prices clear all mar- reform. Our approach imposes minimal aggrega- kets--but no attempt was made to ensure full con- tion conditions on the survey data, within unavoid- sistency between the microanalysis and the predic- able data limitations. In addition to calculating the tions of the CGE model. Chen and Ravallion overall effects of a trade reform on poverty and (2004) describe our methods of measuring the wel- inequality, we are able to provide a detailed socioe- fare impacts in detail. 140 East Asia Integrates The survey data were taken from the 1999 Rural the net gain (or loss) for each household can be Household Survey (RHS) and the 1999 Urban computed.11 Table 8.1 summarizes the results. Sec- Household Survey (UHS), both carried out by tion 1 of the table gives the mean gains for the peri- China's National Bureau of Statistics. The RHS sam- ods 1995­2001 and 2001­07, split by urban and ple covers 67,900 households, and the UHS covers rural areas. 16,900.9 NBS kindly provided the microdata for Section 2 of Table 8.1 shows the impacts on three provinces (Liaoning, Guangdong, and income inequality, both actual (for the baseline Sichuan), called here the "test provinces." The com- year, 1999) and simulated for the two stages of the puter program to implement the estimation trade reform. The simulated income distribution is method was written for these data, after which NBS obtained, for the first stage, by subtracting the esti- staff ran the program on the entire national data set. mated gains over 1995­2001 from the 1999 Before China's accession to the WTO in 2001, its incomes at household level and, for the second economy had already begun to adapt to the expected stage, by adding the household-specific gains from change. Chinese trade reform can thus be considered 2001­07 to the 1999 incomes. Thus the first simula- as having two stages: a lead-up period, in which tar- tion gives the distributional impact of the price iffs started to fall in anticipation of WTO accession, changes during the first stage of the reform--that and the period from 2001 onward. Ianchovichina is, what the baseline distribution would have and Martin (2002) argue that 1995 is a plausible looked like without the reforms--and the second beginning of the lead-up period, and we use their shows the impact of the post-2001 price changes-- estimates of the price changes induced by WTO that is, how those changes are expected to affect the accession for the periods 1995­2001 and 2001­07. baseline distribution, looking forward. The discussion that follows focuses mainly on the Section 3 of Table 8.1 gives the headcount index latter period, although estimates of the welfare of poverty as measured by various poverty lines. impacts for the lead-up period are also provided. The "official poverty line" gives estimates based on We calculated the welfare impacts to the survey the poverty lines used by the National Bureau of data for 1999. This year was chosen in part for data Statistics, while the "$1/day" and "$2/day" lines are reasons; it was the most recent year for which access those from Chen and Ravallion (2001). to the microdata could be obtained. Also, we believe The overall gain in mean income is about 1.5 that choosing a year near the middle of the lead-up percent, all in the period leading up to WTO acces- period rather than at the beginning or end might sion. There is almost no impact on inequality, diminish possible biases caused by any nonlinearity either in the period leading up to WTO accession or in the welfare impacts of price and wage changes. predicting forward. The aggregate Gini index increases slightly, from 39.3 percent without WTO accession to 39.5 percent post-WTO. Results In 1999 the incidence of poverty would have Summarized here are the results described in Chen been slightly higher had there been no trade policy and Ravallion (2004). Appendix Table 8.1 and changes over the lead-up period to WTO accession, Appendix Table 8.210 show the predicted changes in but that from 2001 to 2007 will increase very relative prices and wages in China during slightly as a result of the price changes expected to 1995­2001 and 2001­07, respectively, as obtained be induced by the remaining tariff changes. from the China GTAP model of Ianchovichina and The impacts of WTO accession on poverty as Martin (2002). The tables also show the mean net measured by a wide range of poverty lines are revenue per capita for urban and rural areas, based revealed in Figures 8.1 and 8.2, which give the on the 1999 rural and urban household surveys. cumulative distributions of income for both the baseline and the two simulated distributions, for the poorest 60 percent in rural areas and the poor- Measured Welfare Impacts of WTO Accession est 40 percent in urban areas. Based on the relative price changes from the China In the aggregate, then, WTO accession has virtu- GTAP model and production/consumption shares ally no predicted impact on poverty and inequality. from the 1999 rural/urban household survey data, Disaggregating the results, predicting forward from Household Welfare Impacts of China's Accession to the WTO 141 TABLE 8.1 Predicted Aggregate Impacts on Welfare of Rural and Urban Households Rural Urban National 1. Mean gains (yuan per capita) 1995­2001 34.47 94.94 55.49 (1.54%)a 2001­07 ­18.07 29.45 ­1.54 (­0.04%)a 2. Inequality impacts (Gini index as %) Baseline (1999) 33.95 29.72 39.31 Simulated: less gains 1995­2001 33.90 29.68 39.27 Simulated: plus gains 2001­07 34.06 29.65 39.53 3. Poverty impacts (headcount index, %) Official poverty line Baseline (1999) 4.38 0.08 2.92 Simulated: less gains 1995­2001 4.56 0.08 3.04 Simulated: plus gains 2001­07 4.57 0.07 3.04 $1/day (1993 PPP) Baseline (1999) 10.51 0.29 7.04 Simulated: less gains 1995­2001 10.88 0.28 7.28 Simulated: plus gains 2001­07 10.81 0.28 7.23 $2/day (1993 PPP) Baseline (1999) 45.18 4.07 31.20 Simulated: less gains 1995­2001 46.10 4.27 31.88 Simulated: plus gains 2001­07 45.83 3.97 31.60 Note: PPP = purchasing power parity. a. Percentage of mean income. Source: The authors. FIGURE 8.1 Poverty Incidence Curves: Rural Percent of people below poverty line 60 50 Pre-WTO 40 Post-WTO (middle) 30 Baseline distribution 20 10 0 400 600 800 1,000 1,200 1,400 1,600 1,800 2,000 Annual per capita income (yuan) Source: The authors. 142 East Asia Integrates FIGURE 8.2 Poverty Incidence Curves: Urban Percent of people below poverty line 40 35 Pre WTO income 30 25 Baseline distribution 20 in 1999 (middle) 15 Post WTO income 10 5 0 500 1,000 1,500 2,000 2,500 3,000 3,500 4,000 4,500 Annual per capita income (yuan) Source: The authors. WTO accession, we focused on three indicators of between proportionate gains and the mean income impact at the household level: the absolute gain or of the province (Figure 8.4). loss in yuan, the percentage gain or loss, and One spatially contiguous region stands out as whether the change is a gain or a loss. Our interest losing the most from trade reform--the northeast in the first two measures is obvious enough. We provinces of Heilongjiang, Inner Mongolia, Jilin, included the third to help determine where there and Liaoning (Appendix Table 5.3). Both the might be high concentrations of losses--whether absolute and proportionate impacts are highest in in specific geographic areas or socioeconomic this region; indeed, more than 90 percent of farm- groups. ers in Heilongjiang and Jilin are expected to experi- About three-fourths of rural households and ence a net loss in income. one-tenth of urban residents are predicted to lose When households are ranked by initial income, real income in the period 2001­07. Farm income is there is a notable difference between urban and predicted to drop by 18 yuan per person, while rural households. Figures 8.6, 8.7, and 8.8 plot the urban per capita income rises by 29 yuan (Appen- results given earlier against percentiles of the dix Table 5.3 and Appendix Table 5.4). The drop in income distribution. For example, the mean impact rural income is caused by the drop in the wholesale in yuan per capita at the median income level is prices of most farm products, plus higher prices for revealed by looking at the 50th percentile in Figure education and health care (Appendix Table 5.2). Yet 8.6. (Figure 8.6 gives the horizontal differences in farmers will benefit from the drop in some con- Figures 8.1 and 8.2 plotted against the point on the sumer prices and from the increase in nonfarm vertical axis.) labor wages. Urban residents will enjoy lower prices The absolute gains tend to be higher for higher- for most farm products and higher wages, but they income households in urban areas and lower for also will be hit by higher fees for education and low-income households in rural areas. Nationally health care. (combining urban and rural areas using the weights Impacts differ widely across regions, as evident described in the technical appendix to this chapter), in Figures 8.3, 8.4, and 8.5, which plot the results by there is a hint of a U-shaped relationship, though provinces ranked by mean income per person.12 still with the highest absolute gains for the rich. The mean absolute gains tend to be highest in the This pattern flips when one looks at the propor- richest provinces in both urban and rural areas tionate gains (Figure 8.7). In urban areas, the pro- (Figure 8.3), although there is no correlation portionate gains tend to fall as income rises, but in Household Welfare Impacts of China's Accession to the WTO 143 FIGURE 8.3 Mean Gains by Provinces: Absolute Gains in Yuan Per Capita Net gain/loss per capita (yuan) 60 40 Urban 20 Total 0 ­20 ­40 Rural ­60 ­80 1 6 11 16 21 26 31 Province ranked by provincial per capita income Source: The authors. FIGURE 8.4 Mean Gains by Provinces: Proportionate Gains in Percent Percent gain/loss per capita 2 1 Urban Total 0 ­1 Rural ­2 ­3 1 6 11 16 21 26 31 Provinces ranked by provincial per capita income Source: The authors. rural areas and nationally they rise with income. In of gains and losses as a function of household char- the aggregate, a higher proportion of gainers is evi- acteristics. We estimated the model for the three test dent as one moves up the income ladder, which is provinces--Liaoning, Guangdong, and Sichuan-- driven by the rise in the number of gainers as for which complete microdata were available. The income increases within rural areas (Figure 8.8). key characteristics considered included the age of household head, education and demographic char- acteristics, and land (interpreted as a fixed factor of Incidence of Gains and Losses production because it is allocated largely by admin- Which types of households gain and which lose? istrative means in rural China). We considered rural The technical appendix to this chapter describes in and urban areas separately; there are some differ- formal terms the model for explaining the incidence ences in the explanatory variables between them. 144 East Asia Integrates FIGURE 8.5 Mean Gains by Provinces: Percentage of Gainers Percent of gainers 100 90 80 Urban 70 60 50 Total 40 30 20 Rural 10 0 1 6 11 16 21 26 31 Provinces ranked by provincial per capita income Source: The authors. FIGURE 8.6 Mean Gains in Yuan, by Income Percentile Net gain or loss per capita (yuan) 40 30 Urban 20 10 0 National ­10 ­20 ­30 Rural ­40 ­50 ­60 0 10 20 30 40 50 60 70 80 90 100 Percent of population ranked by per capita income (assuming urban price is 15% higher) Source: The authors. The results are given in Appendix Tables 8.6 and late covariates of potential relevance in thinking 8.7 for rural areas and Appendix Tables 8.8 and 8.9 about compensatory policy responses. for urban areas. Because these results are averages Looking first at rural areas, in all three provinces across the impacts of these characteristics on the the predicted gain from trade reform tends to be consumption and production choices that deter- larger for larger households. There is also a U- mine the welfare impact of given price and wage shaped relationship with the age of the household changes, they are difficult to interpret. We view head, such that the gains reach a minimum at about them as mainly of descriptive interest, to help iso- 50 years of age (47 in Liaoning, 52 in Guangdong, Household Welfare Impacts of China's Accession to the WTO 145 FIGURE 8.7 Mean Percentage Gain, by Income Percentile Percent net gain or loss per capita 2.0 1.0 Urban 0.0 National ­1.0 Rural ­2.0 ­3.0 ­4.0 ­5.0 ­6.0 0 10 20 30 40 50 60 70 80 90 100 Percent of population ranked by per capita income (assuming urban price is 15% higher) Source: The authors. FIGURE 8.8 Percentage of Gainers, by Income Percentile Percent of gainers 100 90 Urban 80 70 60 50 Total 40 Rural 30 20 10 0 10 20 30 40 50 60 70 80 90 100 Percent of population ranked by per capita income (assuming urban price is 15% higher) Source: The authors. and 55 in Sichuan). The gains are smaller for agri- graphic effect is that younger households (those cultural households. They are larger for households with a higher proportion of children under six) with more employees, more workers in township tend to be gainers in Liaoning. and village enterprises, more migrant workers, and For agricultural households, predicted losses are less cultivated land (although the last finding is sig- significantly higher than average in six counties in nificant only in Liaoning). The only strong demo- Liaoning (losses of from 3 to 5.6 percent versus the 146 East Asia Integrates provincial average of 1.3 percent), seven counties in This incidence analysis suggests that encourag- Guangdong (from 2.5 to 5.3 percent versus the ing labor flexibility is central to protecting house- provincial average of 0.8 percent), and six counties holds adversely affected by reforms. But workers in Sichuan (from 2.8 to 5.7 percent versus the need skills to be flexible, and they must be allowed provincial average of 0.7 percent)--see Appendix to move occupationally and geographically, sug- Table 8.10. gesting that labor market reforms must be carried In urban areas, the gains tend to be larger for out in parallel with trade reforms. In China, even smaller households (except in Guangdong). As in though its degree of labor mobility is considerable rural areas, there is a U-shaped pattern (except in (as evidenced by the large migrant population), Liaoning), with the smallest gains for households restrictions and impediments remain. With whose heads were 66 years of age in Guangdong and increased labor flexibility, more rural households 51 years in Sichuan. By contrast with rural areas, can potentially benefit from WTO accession. How- where we find no relationship between education ever, to assess properly the impact of changes in levels and welfare gains, in urban areas the gains tend internal labor mobility policy on household wel- to be larger for less well-educated households.13 fare, we would require a richer data set and addi- There are signs of some sectoral effects, although tional behavioral assumptions, which suggests a only significantly so in Liaoning, with higher gains future research agenda. for those in government jobs. Signs of larger gains Naturally, this approach has other limitations. A are evident among those whose employer is the gov- case in point: some dynamic gains from trade liber- ernment. Retirees tend to gain less than others. alization may not be captured by the model used to generate the relative price impacts. For example, trade may well facilitate learning about new tech- Conclusions nologies and innovation that brings longer-term In the aggregate, the finding is that China's acces- gains in productivity. These effects may be better sion to the WTO has only a small impact on mean revealed by studying time-series evidence, com- household income, inequality, and incidence of bined with cross-country comparisons. poverty. There is, however, a sizable, and at least The geographic differences identified in welfare partly explicable, variance in impacts across house- impacts arise entirely from differences in house- holds of different characteristics. Rural families tend hold consumption and production behavior. In to lose; urban households tend to gain. Impacts are reality, trade reform is likely to have differential larger in some provinces than others, and largest in impacts on local prices as a result of transport or the geographically contiguous northeast region of other impediments to internal trade. Our approach Heilongjiang, Inner Mongolia, Jilin, and Liaoning. does not incorporate such differences, and doing so In this region rural households depend more on would pose data and analytical problems. This feed grain production (for which falling prices are direction might, however, be a fruitful one for expected from WTO accession) than those else- future work in settings in which the necessary data where in China. Such impact"mapping"can be used on prices and wage levels by geographic area are to guide targeted safety net programs. available. Within rural or urban areas of a given province, Another limitation of the approach is that, as the gains from WTO accession vary with observ- explained in the technical appendix, we had to able household characteristics. The most vulnera- make linear approximations in the neighborhood ble households tend to be rural, to be dependent on of an initial optimum for each household. In other agriculture, to have relatively few workers, and to applications using this method, doing so may be have weak economic links to the outside economy deceptive if the price or wage changes are large, or if through migration. There are also some strong geo- a household is initially out of equilibrium, perhaps graphic concentrations of adverse impacts. For because of rationing (including involuntary unem- example, agricultural households in certain coun- ployment). In principle, there are ways to deal with ties of Guangdong, Liaoning, and Sichuan incur these problems by estimating complete demand predicted welfare losses of 3­5 percent of their and supply systems, allowing for rationing. This incomes. avenue may prove to be a fruitful one for future Household Welfare Impacts of China's Accession to the WTO 147 research, but it should be noted that such methods legal registration system (Hukou). As in other countries, the generate their own problems, such as those arising RHS gives data on the remittances of migrant workers, but does not provide information about the migrant workers from incomplete data on price and wage levels at themselves, who (unlike in other countries) are not sampled the household level. in the urban survey either. It is therefore difficult to measure While acknowledging these limitations, we impacts through labor mobility and rural-urban transfers in the present study. Comparisons between the RHS and UHS believe that the type of approach offered here can also pose problems. For example, income in the RHS still illuminate the likely short-term distributional includes income in-kind (such as from own-farm produc- impacts of economy-wide reforms, with minimum tion and other household enterprises), but income in the UHS ignores some in-kind components, notably subsidies aggregation. By avoiding unnecessary aggregation received from the government. For further discussion in the of the primary household-level data, the relatively context of the RHS, see Chen and Ravallion (1996). simple tools described here can also offer insights 10. All the appendix tables cited in this section appear in the into the kinds of policy responses that might be technical appendix to this chapter at www.world bank.org/eaptrade. See also Chen and Ravallion (2004). needed to compensate losers from reform. 11. Using equation 3, as explained in the technical appendix. 12. For the province rankings, see Appendix Table 8.5 at www.worldbank.org/eaptrade. Endnotes 13. This finding may be biased by the fact that, as explained in the technical appendix at www.worldbank.org/eaptrade, we 1. Wood (1994) makes a qualified argument along these lines. had to use education levels to identify skilled labor (noting 2. Jalan and Ravallion (1998) report evidence of such churn- that wages of unskilled nonfarm workers are predicted to ing using panel data for rural China. Baulch and Hoddinott increase relative to those of skilled labor, as shown in (2000) review evidence for a number of countries. Appendix Tables 8.6­8.8). 3. In the context of China's lagging poor areas, see Jalan and Ravallion (2002). 4. For an overview of alternative approaches to assessing the References welfare impacts of trade policies and examples, see McCul- loch, Winters, and Cirera (2001). Barro, Robert. 2000."Inequality and Growth in a Panel of Coun- 5. The only example of this full integration known to us is tries." Journal of Economic Growth 5: 5­32. that by Cockburn (2002), who built a classic, trade-focused Baulch, Bob, and John Hoddinott. 2000. "Economic Mobility CGE model onto the Nepal Living Standards Survey cover- and Poverty Dynamics in Developing Countries." Journal of ing about 3,000 households. Development Studies 36 (6): 1­24. 6. In an antecedent to our approach, Bourguignon, Robilliard, Bourguignon, François, and C. Morisson. 1990. "Income Distri- and Robinson (2003) also take price changes generated by a bution, Development and Foreign Trade." European Eco- CGE model to survey data (for Indonesia). Methodologi- nomic Review 34: 1113­32. cally, the main difference is that we derive first-order welfare Bourguignon, François, Anne-Sophie Robilliard, and Sherman impacts analytically from a standard competitive farm- Robinson. 2003. "Representative versus Real Households in household model, and Bourguignon and his colleagues gen- the Macro-Economic Modeling of Inequality." Working erate income impacts at the household level from a micro- Paper 2003-05. DELTA, Paris. econometric model of income determination. Chen, Shaohua, and Martin Ravallion. 1996."Data in Transition: 7. Details of the methodology and data treatment are given in Assessing Rural Living Standards in Southern China." China the technical appendix to this chapter at www.world Economic Review 7: 23­56. bank.org/eaptrade. The appendix includes a discussion of ------. 2001. "How Did the World's Poor Fare in the 1990s?" the main assumptions underlying the CGE framework. See Review of Income and Wealth 47 (3): 283­300. also Chen and Ravallion (2004). ------. 2004. "Welfare Impacts of China's Accession to the 8. Papers describing the standard GTAP with applications can WTO." World Bank Economic Review, forthcoming. be found in Hertel (1997). A full discussion of the assump- Cockburn, John. 2002. "Trade Liberalization and Poverty in tions of the general equilibrium model and the results of its Nepal: A Computable General Equilibrium Micro Simula- application to China's accession to the WTO can be found tion Analysis." University of Laval, Quebec. Processed. in Ianchovichina and Martin (2002). Dollar, David, and Art Kraay. 2002. "Growth Is Good for the 9. The full sample of the UHS in 1999 was about 40,000 Poor." Journal of Economic Growth 7 (3): 195­225. households. Since 2002 the central office has kept the data Edwards, S. 1997. "Trade Policy, Growth, and Income Distribu- on all 40,000 households. tion." American Economic Review, Papers and Proceedings 87 Over the past 15 years, the NBS has made a great effort to (2): 205­10. improve both the RHS and UHS, focusing on sample cover- Hertel, T. W., ed. 1997. Global Trade Analysis: Modeling and age, questionnaire design, survey methodology, and data Applications. Cambridge: Cambridge University Press. processing. The number of variables in the surveys has Ianchovichina, Elena, and William Martin. 2002. "Economic increased dramatically, with additional details on income, Impacts of China's Accession to the WTO." Paper presented expenditure, savings, housing, and productivity, among at Seminar on WTO Accession, Policy Reform and Poverty other things. However, some problems remain in the 1999 Reduction in China, World Bank, Beijing, June 28­29. RHS and UHS. For its sample frame, the RHS relies on its Jalan, Jyotsna, and Martin Ravallion. 1998."Transient Poverty in sampled counties from 1985. The UHS excludes rural Post-reform Rural China." Journal of Comparative Economics migrants, because the base of the UHS sample frame is the 26: 338­57. 148 East Asia Integrates Li, Hongyi, Lyn Squire, and Heng-fu Zou. 1998. "Explaining Milanovic, Branko. 2002. "Can We Discern the Effect of Global- International and Intertemporal Variations in Income ization on Income Distribution?" Policy Research Working Inequality." Economic Journal 108: 26­43. Paper 2876. World Bank, Washington, D.C. Lundberg, Mattias, and Lyn Squire. 1999. "Growth and Inequal- Ravallion, Martin. 2001."Growth, Inequality and Poverty: Look- ity: Extracting the Lessons for Policymakers." World Bank, ing Beyond Averages." World Development 29 (11): 1803­15. Washington, D.C. Processed. Wood, Adrian. 1994. North-South Trade, Employment and McCulloch, Neil, L. Alan Winters, and Xavier Cirera. 2001. Trade Inequality. Changing Fortunes in a Skill-Driven World. Liberalization and Poverty: A Handbook. London: Centre for Oxford: Clarendon Press. Economic Policy Research. 9 Trade in Sectors Important to the Poor: Rice in Cambodia and Vietnam and Cashmere in Mongolia Jehan Arulpragasam Francesco Goletti Tamar Manuelyan Atinc Vera Songwe Trade liberalization has profoundly changed the of the organizational structure of a particular com- determinants of national economic growth and modity or sector is crucial for designing a set of social development. World trade is now growing measures that enables the poor to take fuller advan- much faster than the world gross domestic product tage of greater access to markets. (GDP), so that growth opportunities are typically The studies underpinning this chapter illustrate greater for exports than they are for domestic sales. the kind of analysis--using a supply chain But the extent to which poor people, especially poor approach--that is needed to inform policymakers producers, benefit from expanded trade opportuni- and the directions that need to be taken to ensure ties often depends on complementary measures that trade opportunities reach the poor. The illus- designed to tackle behind-the-border constraints to trative case studies focus on two sectors--rice in efficient production and exports. These measures Cambodia and Vietnam and cashmere in Mongo- include ones to foster the development of competi- lia. Why rice and cashmere? For one important rea- tive markets and public action to provide informa- son: in the countries considered, the two commodi- tion, reduce transaction costs that are often high as a ties provide a livelihood for the majority of poor result of corruption, deliver public services, and people, and therefore any gains in productivity or address market and collective action failures. Poor in the share of producers in the value chain will producers are not only economically disadvantaged lead to improvements in poor people's welfare. The but also often politically powerless, and when their value chain is the steps of the production system interests are pitted against those of more powerful (from raw producer to consumer) as well as the actors, they frequently lose. Understanding the linkages among the networks of producers, institutional and political economy underpinnings exporters, importers, and retailers along the way. The authors gratefully acknowledge helpful comments from Florian Alburo and participants in seminars held at the Insti- tute for Southeast Asian Studies of the National University of Singapore and at the World Bank office in Ulaanbaatar. 149 150 East Asia Integrates In Cambodia and Vietnam, as in much of Asia, resistance to full and effective implementation of most of the poor earn a living by growing rice, and reforms (Braguinsky and Yavlinsky 2000). the rice culture permeates the farming traditions of Second, at the same time, new entrants do not the region. In Mongolia, livestock herding, a tradi- always readily appear to fill in market opportunities tion that is centuries old, provides a livelihood for created by the removal of administrative con- most of the rural population. These are traditional straints to entry. In addition to noncompetitive economic activities of the poor, determined by his- behavior by incumbent economic actors, often tory and country context, and higher productivity abetted by state agencies, low capacity, limited and returns to producers in these subsectors would access to working capital, and low population den- have a vital impact on the livelihoods of the poor. sity are factors that commonly constrain the devel- Why value chain analysis? The main remaining opment of the private sector in many transition constraints to improvements in the welfare of poor economies (Sachs, Zinnes, and Eilat 2000; Small- producers engaging in rice and cashmere produc- bone and Welter 2003). tion are related to sectoral policy. Domestic trade Third, a common market failure in early transi- policies, by and large, no longer represent impor- tion economies is imperfect information--about tant impediments, which means that exporters the changes in structural policies that affect the sec- receive the undistorted prices signals from interna- tor and about relevant prices. tional markets. At the other end of the export Fourth, information services are just one of the chain, producers have incentives to respond to many ancillary services and institutions on which a price signals, because they now have full property fully developed market economy relies for effi- rights over their assets or the returns to those ciency. Indeed, value chains in transition assets, thanks to the increased security of land economies are often hindered by a lack of the rele- tenure in Vietnam and the privatization of live- vant institutions, technology, and infrastructure stock in Mongolia. But between the two ends of the that may have developed in step with market devel- supply chain is a host of institutional and policy opment, in response to market demands, in other constraints that tend to interfere in the transmis- more fully developed market economies sion of the price signals and reduce the ability of (Havrylyshyn and van Rooden 2003). Institutional poor producers to benefit from expanding oppor- failure or missing markets in areas such as trans- tunities. port services, insurance, credit, and research and For transition economies such as Vietnam, extension are just some examples of lacunae in Cambodia, and Mongolia, institutional constraints linked markets that challenge the development of are particularly prevalent.1 Indeed, transition efficient value chains. economies making the complex move from a com- Fifth, the freeing up of prices is expected to lead mand economy regime to a more market-oriented to more competitive and efficient behavior. But one are subject to particular challenges and con- behavioral changes are not instantaneous and are straints that value chain analysis is well placed to themselves closely linked with institutional change identify and assess. and the development of markets. For example, First, the removal of the administratively dic- there may be a role for collective action or public tated value chains found under a command econ- action to raise the capacity of nascent private sector omy (e.g., where a certain set of economic agents is entrepreneurs to figure out how collective behavior commanded to sell to another set of economic might improve their efficiency, or how quality agents, perhaps with the prohibition of entry of enhancements might improve their returns (e.g., other agents) is expected to (indeed, intended to) through the establishment of grading systems). result in a structural transformation. State-owned Moreover, in many transition economies, although enterprises that may have had a monopoly or a the move to efficient markets calls for the develop- monopsony under such an arrangement will ment of market-related institutions and behavior, it become subject to competition. Low levels of effi- also calls for dismantling the institutions of petty ciency that may have been protected might prompt corruption and nuisance bribes that can add signif- closures or a general restructuring of the value icantly to transaction costs (North 1991; World chain, raising serious political economy issues and Bank 1996; Boeva 2002). Trade in Sectors Important to the Poor: Rice in Cambodia and Vietnam and Cashmere in Mongolia 151 Value chain analysis (see Box 9.1) is a useful tool The analysis in the rest of the chapter describes for understanding these constraints and proposing the considerable impediments to both upgrading a set of measures to relieve them. Valuable in the the value of rice and cashmere production and context of any economy, it is particularly useful in increasing the share of poor producers in the value diagnosing the constraints produced by market chains. In all three countries, powerful constituents failures, institutional and capacity failures, as well appropriate rents through explicit preferences as policy and actual policy implementation failures (credit for state-owned enterprises in Vietnam, in transition economies. Value chain analysis evalu- export tax in Mongolia), industrial structure (near ates the behavior of each major participant (pro- monopsony for a few millers in Cambodia), and ducers, processors, traders, exporters) in the system the toleration of corrupt practices (illegal fees). that links production to final consumption and Meanwhile, poor producers remain trapped in low- provides a detailed account of the constraints par- productivity states in the absence of improvements ticipants face at different stages of the chain. in the delivery of public services such as the poor The value chain analysis approach considers road network and inadequate research and exten- international trade relations as part of a series of sion. Our analysis shows that the poor can benefit networks of producers, exporters, importers, and from the expanded opportunities presented by retailers whereby knowledge and relationships are global integration, provided that these institutional developed to gain access to markets and suppliers. constraints are addressed. Giving voice to poor pro- In the shorter term, the relations among different ducers' interests by placing these issues on the pol- participants in a sector, in turn, determine the icy agenda is crucial for fostering reforms that means by which benefits are distributed within the unleash the productivity potential of poor people chain and influence the way different actors try to and increase their bargaining power. improve their positions within the chain. From the The first section of this chapter analyzes the constraints identified, it is possible to derive policy value chain in rice in Cambodia and Vietnam and recommendations to raise the returns of poorer makes policy recommendations to address the households within the sector. These recommenda- constraints that face different groups of partici- tions can then be validated and assigned priority pants. It then validates these recommendations by means of quantitative methods. Over the longer using quantitative methods and weighs their effec- term, measures to eliminate constraints need not tiveness in reducing poverty. The next section ana- be conflictual in nature, as would be the case in a lyzes the cashmere value chain in Mongolia and zero-sum game. Policy, institutional, and gover- makes policy recommendations. A concluding sec- nance changes can indeed eliminate constraints, tion follows. improve efficiency and competitiveness within the chain, and permit incremental value to be "added" Rice and the Poor in Cambodia to the chain (e.g., by improving the quality of the and Vietnam final product or better placing it in the global mar- ketplace). Rice is a central feature of life in Cambodia and Developing a value chain is tantamount to Vietnam. More than 80 percent of the population transforming the production function. The added of both countries is rural, and most of the rural value can result in important gains from trade for population produce rice (Table 9.1). Rice covers all participants: a positive-sum game. In fact, the much of the cultivated area (64 percent in Vietnam focus on the links among various actors in the and 84 percent in Cambodia) and makes major value chain can be a powerful means of building contributions to agricultural GDP (35 percent in consensus around reform. The common objective Vietnam and 22 percent in Cambodia) and to agri- of all participants--to enhance the value of their cultural exports. It is the main staple food, repre- product in the global marketplace--could form an senting 60­70 percent of the calories in the diet of important constituency for, if not motivator of, the average rural household and absorbing about reform at various levels of the economy (e.g., pro- 30 percent of household expenditures. Food secu- ducers, traders, processors, exporters, etc.) with rity in both countries is closely identified with rice, knock-on benefits for other sectors of the economy. as supported by recent history.2 152 East Asia Integrates BOX 9.1 Value Added from Value Chains Kaplinsky defines the value chain as "the full for the retailer while increasing producers' share range of activities . . . required to bring a prod- in the value chain by cutting out the middlemen. uct or service from conception through the Strategies devised to improve the overall value intermediary phases of production, delivery to in the chain could focus on upgrading upstream final consumers, and final disposal after use" actors, which may take the form of either devel- (Kaplinsky 1999: 121). The value chain method- oping new, higher-value market niches or ology derives from two strains of literature: the expanding the range of activities--for example, a business literature on strategy and organization manufacturer might expand into distribution or (Porter 1990) and the literature on global com- research and development. The growing role of modity chains promoted by Gereffi and devel- supermarket procurement systems in developing oped in numerous studies in the late 1990s countries is a key development in this regard (UNCTAD 2000). (Reardon and others 2003; Weatherspoon and Value chain analysis focuses on the interac- Reardon 2003). For example, exporters able to tion of actors along each step of the production place packaged carrots in UK supermarkets can system (from the producer of the raw material to earn a premium 7­15 times that for exporting consumer) as well as the linkages within each set ordinary bulk carrots (Dolan and others 1998). of actors. Such an approach thus considers inter- Yet at the same time, such systems are leading to national trade relations as part of a series of net- consolidation and a shift toward nontraditional works of producers, exporters, importers, and wholesalers, as well as the introduction of strin- retailers, whereby knowledge and relationships gent private standards, posing compliance chal- are developed to gain access to markets and lenges for small farmers. suppliers. By mapping the range of activities in The role of governance structures is important the chain, value chain analysis allows for a in how such upgrading by suppliers occurs, as is decomposition of earnings and rewards among the support of government and other institutions different parties in the chain. As a diagnostic (UNCTAD 2000). Indeed, there may be a regula- methodology, it allows identification of where tory or supportive role for the public sector in this rents are being captured, as well as the nature regard. In addition, the overall value of the chain and extent of barriers to entry along the chain. often benefits from improvements in the quality of Value chain interventions can both affect the raw material at the early stages of the value chain, relative gains of the economic agents that par- pointing to the benefits of cooperative solutions ticipate in the chain and enhance the total value among, for example, smallholder producers, that accrues to the chain, enabling all partici- traders, and processors. Presumably, there would pants to gain. Value chain analysis can con- be a common interest in ensuring that small- tribute to an understanding of the governance holder suppliers reduce the risks of plant disease structures within a value chain and to an under- and enhance the quality of their supply. standing of where relative gains are likely to As market integration moves to sourcing of accrue. Also important, it can lead to strategies high-quality products for niche markets, action that improve the relative gains of certain actors is also warranted to ensure that the average (such as poor smallholder producers at the bot- smallholder does not get marginalized (Dolan tom of the chain) and to strategies in which all and others 1998; Reardon and others 2003). actors can benefit from the gains to trade. Also important to improving the overall value in Strategies to improve relative positions may the chain are measures taken to reduce market- involve cooperative solutions through vertical or ing margins and transaction costs arising from horizontal integration within the chain. For infrastructure and transport constraints. There is example, farmer associations may help to substantial evidence, for example, that lowering improve the bargaining power of producers vis- transport margins through better rural roads à-vis traders, or direct contractual relations benefits not only farmers but also other eco- between smallholders and retailers might nomic agents along the value chain (Hayami enhance the predictability and quality of supply and Kawagoe 1993). Trade in Sectors Important to the Poor: Rice in Cambodia and Vietnam and Cashmere in Mongolia 153 TABLE 9.1 Rice in Cambodia and Vietnam Indicator Unit Vietnam Cambodia Agriculture as share of GDPa percent 24 39 Agricultural labor as share of total laborb percent 69 85 Rice as share of total cultivated areac percent 64 84 Rice as share of GDP in agricultured percent 35 22 Share of rural households growing ricee percent 80 78 Share of rural households selling ricef percent 47 na Rice exportsg thousand metric tons 3,600 310 Rice exportsh US$ million 644 55 Rice exports as share of agricultural exportsi percent 27 51 Agricultural exports as share of total exportsi percent 16 12 Total production of ricej thousand metric tons 20,150 2,600 Total rice production per capitaj kilograms per capita per year 252 217 Share of rice in total calorie consumptionk percent 64 70 Budget share of rice in rural household food budgetk percent 36 29 Budget share of food in total rural household expendituresl percent 51 69 a. Data for Vietnam refer to 1999 and for Cambodia to 2001. b. Data for both Vietnam and Cambodia refer to 1998. c. Data for Vietnam refer to 1999 and for Cambodia to 2000. d. Data for Vietnam refer to 2000 and for Cambodia to 2001. e. Data for Vietnam refer to 1998 and for Cambodia to 2002. f. Data for Vietnam refer to 1998. g. Data for Vietnam refer to 2001 and for Cambodia to 2001 and include rice equivalent of 400,000 metric tons of paddy. h. Assume average price for Vietnam of US$179 per metric ton and for Cambodian paddy of US$90 per metric ton and rice of US$350 per metric ton. i. Data for Vietnam refer to 2000 and for Cambodia to 2001. j. Data for Vietnam and Cambodia refer to 2001. k Data for Vietnam refer to 1998 and for Cambodia to 1998. l. Data for Vietnam refer to 1998 and for Cambodia to 1999. Sources: Royal Government of Cambodia (1999, 2001); Socialist Republic of Vietnam, Statistical Yearbook (various issues); Socialist Republic of Vietnam (1998); World Bank Study Team, 2002. Poor households contribute 15­25 percent of with few modern inputs. These varieties usually total rice production, cultivating very small hold- fetch higher prices than HYV rice, because they are ings. The average poor household has a rice land- higher-quality and local consumers prefer them. holding of less than a hectare in Cambodia and less The scope for increasing rice yields in the Mekong than half a hectare in Vietnam (Table 9.2). In Viet- and Red River Deltas is limited, but the scope in the nam, most farmers outside of the productive more remote and mountainous areas of Vietnam, Mekong and Red River Deltas grow rice on small and in all provinces of Cambodia, is considerable. plots for subsistence reasons. Income per hectare of rice ranges between US$100 and $250.3 For the poor, rice provides a Rice Exports smaller share of total income than for the average In both countries, liberalization policies have rice farmer (Table 9.3). pushed up rice supplies considerably over the past Most of the poor in both countries grow tradi- decade, at average rates of 4­5 percent a year, mostly tional rice varieties, which, unlike modern high- as the result of increases in yield and cropping yielding varieties (HYVs), can thrive on poor land intensity (ANZDEC 2000; OPCV 2002). The pro- 154 East Asia Integrates TABLE 9.2 Land Size and Rice Profit, Cambodia and Vietnam Indicator Cambodia Vietnam Average landholding size (hectares) 1.71 1.08 Average size of cultivated rice land (hectares/household) 0.99 0.87 Average household size 5.3 4.9 Size of cultivated rice land by the poor (hectares/household) < 1 < 0.5 Rice production gross margin on labor (US$/hectare) 49 96 Rice production gross margin, including labor (US$/hectare) 147 216 Sources: Cambodia: Goletti, Bhatta, and Srey (2002) and ACI (2002a); Vietnam: ACI (2002b). TABLE 9.3 Rice and Income of the Poor, Cambodia and Vietnam (percent) Indicator Cambodia Vietnam Rice as share of total farmers' income 50.4 42.5 Rice as share of income of the rural poor 40.4 47.5 Sources: Cambodia: Goletti, Bhatta, and Srey (2002); Vietnam: Minot and Goletti (1998). duction increases have allowed both countries to unlikely to grow much faster than the population achieve self-sufficiency and to export rice. Vietnam (at 2 percent a year). The world rice market is thin has become a major player in international markets. and growing only slowly, but it presents good Indeed, with exports of 3.5­4.5 million metric tons, opportunities for export gains by countries such as it is now the world's second largest exporter after Vietnam and Cambodia that have a comparative Thailand (see Appendix Table 9.3).4 Cambodia has advantage in rice, especially if global trade liberal- begun to export rice (milled) to international mar- ization continues in the direction indicated by kets (316,000 metric tons currently) and paddy Uruguay Round trade negotiations. Vietnam and (threshed, unmilled rice) to Vietnam. Cambodia are in competition with only a few other The values of rice exports are significant for main suppliers, and the customers are many, with both countries. For Vietnam, they range from $600 diverse preferences. million to $1 billion a year and represent 30­50 Currently, Cambodia and Vietnam have a 2 per- percent of agricultural trade. For Cambodia, rice cent and a 16 percent share of world rice exports, plus paddy exports provide about $40 million a respectively. If they want to maintain or increase year and represent about 40 percent of agricultural their shares, they will need to maintain or increase exports, surpassing rubber and forest products. their productivity relative to that of their competi- Cambodia and Vietnam have a comparative and tors. Their success will also depend on establishing competitive advantage in rice relative to other links with value chains that go beyond their regional and international rice producers.5 Accord- national boundaries. ing to the computed domestic resource costs for dif- Some opportunities for export gains may lie in ferent rice varieties and production patterns, both diversifying into higher-quality rice. Rice comes in Vietnam (with domestic resource cost ratios for rice many varieties, grades, standards, and qualities. Thai- ranging from 0.42 to 0.66) and Cambodia (with the land,as a highly successful rice exporter,has been able same ratios ranging from 0.87 to 0.77) have a com- to accommodate the diverse requirements of foreign parative advantage in rice and would gain signifi- markets--high quality, low quality, fragrant, nonfra- cantly from using more efficient farming practices.6 grant, long grain, short grain, parboiled, normally Exports are likely to be the main driver of fur- milled, and so forth. If a producer like Vietnam con- ther production expansion in rice in the medium tinues to focus on delivering on contracts for low- term, given that the domestic demand for rice is quality rice, it will lose the opportunity to export to Trade in Sectors Important to the Poor: Rice in Cambodia and Vietnam and Cashmere in Mongolia 155 other markets, to diversify its clientele, and possibly enterprises, and exporters. Other participants to increase its overall market share. include transporters, seed companies, agrochemical In neither Vietnam nor Cambodia does trade companies, agricultural equipment companies, irri- policy per se offer significant further promise for gation companies, shipping companies, port author- expanding rice exports. Most of the traditional ities, banks, inspection agencies, commerce and tax instruments of trade policy have already been departments, agricultural departments, farm organi- used.7 Some difficulties in import and export still zations, miller organizations, research organizations, exist as a result of irregularities, mostly corruption extension organization, policymakers, and consumer and weak institutions (inspection, customs, ship- organizations (Figure 9.1). ping, accounting, banking), but barriers to inter- The market and power structures evident in the provincial trade have been lifted in both countries. rice markets in Vietnam and Cambodia highlight Further progress in trade liberalization is certainly the particular challenges facing transition possible, but the most effective way to promote rice economies that are struggling to overcome the lega- exports in both countries seems to be sector policy cies of their previous command economy struc- rather than trade policy. tures in which the state and state-owned enter- Even if rice exports expand successfully, both prises played a particularly important role, and that Vietnam and Cambodia will face challenges in are seeing private sector activity (and even owner- ensuring that the expansion benefits the poor. ship) still in the process of being fully institutional- What can value chain analysis reveal about the spe- ized. Value chain analysis is especially useful in cific constraints and challenges for Vietnam and identifying the links in the value chain where mar- Cambodia in expanding rice exports as an element ket failures or idiosyncrasies are hindering achieve- of a strategy for poverty reduction? ment of competitive behavior, diminished margins, or enhanced value. In fact, this kind of analysis reveals that produc- Value Chain Analysis: Constraints on Production ers are not the dominant actors in the chain in and Trade, and Policy Options either Cambodia or Vietnam; nor are there power- Numerous actors take part in the value chain that ful retailers or multinationals. In Vietnam, a good links rice producers to final consumers: farmers, col- deal of power rests with a state-owned enterprise lectors, millers, wholesalers, retailers, state-owned engaged in rice trade, particularly through govern- FIGURE 9.1 Profit Shares for Rice Value Chain, Cambodia and Vietnam Percent 120 100 VINAFOOD/retailer 80 Exporter/wholesaler 60 Transporter 40 Miller Collector/husker 20 Farmer 0 Cambodia Vietnam ­20 Source: ACI (2002a, 2002b). 156 East Asia Integrates ment-to-government contracts involving large Rice processing, too, faces some important con- amounts of low-quality rice. Financial sector mar- straints. The few millers that do operate have kets have yet to develop that can serve the credit monopsony power, allowing them to capture needs of Vietnamese farmers, and ancillary service higher margins than if the milling link were more delivery functions, such as in the delivery of appro- competitive (Figure 9.1). Their reliance on obsolete priate irrigation and seed varieties, have yet not milling equipment results in high levels of broken been developed. In Cambodia, power lies with a rice, reducing the value of the crop. Poor paddy few large millers oriented toward high-quality rice quality, in the form of mixed varieties of seeds and exports, a new Rice Millers' Federation, and traders inadequate postharvest handling, also increases the who ship paddy to Vietnam. Meanwhile, retailers in percentage of broken rice and lowers the price Cambodia suffer high transaction costs related to earned by most farmers. Lack of working capital poor infrastructure and unofficial fees and other and the high cost of credit limit millers' ability to transaction costs, including poor infrastructure. buy paddy from farmers and update their machin- The sections that follow briefly analyze the char- ery. In the end, such conditions encourage the acteristics, constraints, and behavior of each of the unofficial export of paddy to Vietnam and Thai- main groups of participants, and draw policy rec- land and prevent the country from capturing the ommendations for improving efficiencies and dis- value added from rice milling. Because they are tribution in each country.8 unable to produce consistent amounts of standard- ized varieties of milled rice and because they lack of Cambodia. Cambodian rice farmers face serious information about foreign market conditions, rice constraints on increasing productivity and output millers have limited access to foreign markets. quality. These include a lack of breeding-seed stock, Milling is the key bottleneck in the rice value many farmers' lack of title to the land they farm chain in Cambodia. Millers and farmers have (which discourages them from investing in important unexploited opportunities for collabora- improvements), and a lack of access to commercial tion to encourage more productive cultivation of credit, in part because credit application proce- paddy as well as better sorting and postharvest han- dures are cumbersome and not tailored to small dling. In Angkor Kasekam, farmer associations loans. Water use systems, where irrigation is avail- have linked up with millers and processors, provid- able, are highly politicized, with pervasive free rid- ing a model that could well be replicated elsewhere. ing, and many farmers struggle to subsist on rice More important still is increased competition in production from plots that are just not suited to milling to help reduce milling margins in Cambo- rice production. dia that are substantially higher than those in Viet- Increased public investment is needed to main- nam and to increase prices to poor producers. tain the breeder stock of seed, particularly for the Efforts to build capacity (both private and public) traditional rice varieties grown by the poor. Such in marketing information services would be useful investment should be complemented with an to improve knowledge among participants in the appropriate institutional framework to ensure that rice marketing chain. research centers focus on basic research and vari- As for the distribution of both retail products etal development, while private companies focus and exports, poor roads dampen production incen- on the multiplication and sale of certified seed. tives and reduce market access. Unofficial check- The government also needs to redouble its efforts points and port fees raise the costs of rice for Cam- to implement the land law and raise public aware- bodian consumers and lower the competitiveness ness of land rights. And improvements in the func- of all Cambodian rice products, including those tioning of water user groups would go a long way bound for export. Nearly half of transport fees are toward raising productivity and returns. Those unofficial costs. In addition, exporters are con- farmers who cannot profitably produce rice strained by their inability to obtain consistent because of poor natural resource endowments amounts of a standardized quality of milled rice. should be encouraged to diversify into other prod- Increased investment to improve roads, rail- ucts, with the aid of agricultural extension and ways, and ports and to strengthen the market infor- advisory services. mation system would reduce transaction costs and Trade in Sectors Important to the Poor: Rice in Cambodia and Vietnam and Cashmere in Mongolia 157 raise the profits for actors along the chain. So too milling sector. And, as in Cambodia, there is much would efforts to put into place transparent rules for to be gained by encouraging collaboration between export clearance and to build capacity in legal insti- farmers and millers, with millers providing seed tutions to facilitate export transactions. Reducing and other inputs and enhancing farmers' incentives transaction costs and marketing margins can go a to supply higher-quality, less breakable paddy that long way toward improving returns. would realize higher prices. Rice distribution in Vietnam is hampered by Vietnam. Vietnamese farmers seeking credit face inadequate infrastructure, particularly in areas out- complicated borrowing procedures and unsuitable side the Mekong River Delta, and a multilayered dis- repayment schedules. They have difficulty obtain- tribution system makes it difficult for millers and ing high-quality seed in the absence of standard- exporters to obtain high-quality inputs from pro- ized seed specifications. And they are charged the ducers. State-owned enterprises benefit from prefer- same irrigation fees regardless of how much water ences and are able to borrow the entire value of a they use and despite the poor quality of service. contracted shipment of rice, while private firms can As in Cambodia, Vietnam's research institutes borrow only up to 70 percent. Rice exports from and universities should focus on seed research and Vietnam are vulnerable because they are of rela- varietal development, while commercial seed com- tively poor quality and go to only a limited number panies focus on seed multiplication and sale of cer- of markets. Most exports go through a small num- tified seed. Irrigation fees should be set on the basis ber of distributors, and most sales are made through of metered water use rather than area irrigated (at government-to-government contracts. least at the level of farmer groups), and they should Improving physical infrastructure, particularly in be set at levels high enough to cover capital costs remote and mountainous regions, would lower and operations and maintenance. Also as in Cam- transaction costs. Rationalizing state-owned enter- bodia, research is needed on alternative land use prises, permitting their equitization, and leveling the options and possibilities for diversification, espe- playing field by removing their preferential access to cially for farmers who do not have suitable land or credit arrangements would foster greater private sec- other endowments for productive rice farming. tor participation and efficiencies in this link of the Rice processing faces constraints as well. Viet- chain. The diversification of export outlets might be nam's milling sector is in transition, from a encouraged through simple steps such as the devel- bimodal structure--with a large number of small opment of overseas trade promotion activities and and medium-size mills catering to the domestic strategy. Increased efficiency, lower transaction costs, market and a few large mills catering to the export and better placement of a final, higher-quality prod- market--to a more modern industry with fewer, uct would raise prices, lower costs, and benefit the larger mills. Technology in the domestic milling various actors along the value chain. sector is often outdated, resulting in high levels of broken rice, and, as in Cambodia, millers are fun- Overview of Recommendations damentally constrained by a lack of working capi- tal. Small millers are also constrained by their lim- To the extent that the poor are able to benefit from ited storage space, as well as, for some, shortages of increased production, they will also benefit from electricity. In the large milling sector, the preferen- trade. Most poor households in Vietnam and Cam- tial credit arrangements for state-owned mills ham- bodia are farm households, and because 80 percent per the ability of the private sector to compete of farmers grow rice, the poor are likely to benefit effectively. The awarding of government-to-gov- from any improvement in rice productivity or ernment contracts to state-owned mills also pre- value added. They will probably benefit most from vents expansion of the export base to private mills. improvements in productivity and the technology Interventions are needed to improve postharvest associated with traditional varieties. Therefore, storage facilities and build the financial manage- short-term strategies to reduce the number of peo- ment capacity of millers. Leveling the playing field ple in poverty and the severity of poverty should between state-owned and private mills would also focus on how to increase returns from these vari- foster enhanced private sector involvement in the eties. Steps should be taken to improve production 158 East Asia Integrates technologies for traditional varieties, moderate exports, and agricultural incomes. The results postharvest losses,9 introduce increased competi- reported in Table 9.5 show that a combination of tion and more modern technology into milling, different approaches is needed to improve the strengthen infrastructure, and lower transaction income generated by the value chain. Improve- costs (including illegal fees). Particularly significant ments in productivity, postharvest technology, and will be efforts to establish farm organizations. New marketing and institutions can go a long way institutions and markets that increase information toward increasing exports and agricultural income. about prices and permit the agents in a value chain Clearly, policy design would require a more to form strategic alliances can play an important explicit idea of the costs involved and the time frame role in increasing returns to players such as poor required to implement different options. Even so, smallholder farmers. some useful insights can be obtained from a qualita- The role of the public sector in direct invest- tive analysis such as that reported in Table 9.6. ments to increase productivity should be limited to providing irrigation, where irrigation is economi- Effects on Farmers and the Poor. What are the cally viable. Elsewhere, the public sector should marketing margins and profits for the different par- play an enabling role, both through legislation and ticipants in the value chain? Tables 9.7 and 9.8 through provision of services--especially report these at each level of the chain for rice in improved infrastructure and public policies to Cambodia and Vietnam.11 Farmers and millers have reduce the transaction costs of providing credit. the largest marketing margins and profit margins12 For convenience, the recommendations that in both countries, but the distribution of profits dif- emerged from the value chain analysis are grouped fers; Cambodian millers capture a much higher as strategic options related to improvements in share of profits than their Vietnamese counterparts. productivity, postharvest technology, and infra- In absolute terms, Cambodia's rice value chain structure and chain management (Table 9.4).10 realizes a higher profit than Vietnam's (Figure 9.2), This section uses quantitative methods, applied to in part because Cambodia exports higher-quality household data, to evaluate these options and paddy than does Vietnam. Meanwhile, Cambodia's explore their effects on different participants farmers realize lower unit profits than their Viet- within the value chain. namese counterparts. On average, their rice land is only half as productive as that in Vietnam, but the Aggregate Effects. A multimarket model was built average Cambodian rice farmer has more than to explore the probable effects of the various policy twice as much land as his Vietnamese counterpart. options on aggregates, including production, These features suggest that efforts to reduce poverty TABLE 9.4 Strategic Options for Rice, Cambodia and Vietnam Option Adoption of option Productivity Adopting green revolution package Dry season, Cambodia Improving traditional varieties Wet season, Cambodia Upland areas in Vietnam Postharvest technology Reducing postharvest losses Cambodia, Vietnam Improving milling Cambodia, Vietnam Marketing and chain Improving road and port infrastructure Cambodia, Vietnam management Reducing transaction costs Cambodia, Vietnam Establishing niche market (specialty rice) Cambodia, Vietnam Establishing level playing field Private sector and state- owned enterprises, Vietnam Source: ACI (2002a, 2002b). Trade in Sectors Important to the Poor: Rice in Cambodia and Vietnam and Cashmere in Mongolia 159 TABLE 9.5 Effects of Various Policy Options on Rice Production, Income, and Exports: Cambodia and Vietnam Percent change in Rice Agricultural Rice Cambodia Option production income exports Improved productivity Green revolution technologies 4 1.5 31 Traditional varieties 15 7 228 Improved postharvest handling Milling technology 3.1 1.3 24 Postharvest losses 4.6 2 37 Marketing and chain management Marketing costs farm to port 1.6 5 25 Percent change in Rice Total Rice Vietnam Option production income exports Improved productivity Yield improvement 18.2 1.2 39 Improved postharvest handling Milling technology 0.5 0.37 12 Postharvest losses 1.31 0.1 1 Source: Authors' simulation of multimarket model; see ACI (2002a). TABLE 9.6 Qualitative Analysis of Strategic Options, Cambodia Benefit in Benefit/ Cambodia Option agricultural cost income Cost ratio Time frame Improved productivity Green revolution technologies 2 4 0.5 3­10 years Traditional varieties 5 3 1.7 3­10 years Improved postharvest Milling technology 1 1 1.0 3­5 years handling Postharvest losses 3 2 1.5 3­5 years Marketing and chain Marketing costs farm 4 5 0.8 5­10 years management to port Note: Scoring method: 1 = lowest, 2 = low, 3 = average, 4 = high, 5 = highest. Source: ACI (2002a). by raising productivity might be more effective in lions of small rice farmers could be dramatic Cambodia than in Vietnam. Because rice produc- increases in rural nonfarm activities. tion represents between 30 and 50 percent of the Assessing the overall direct impact on the poor of incomes of the poor in the two countries, even small the proposed improvements in the rice subsector improvements in productivity can lead to impor- would require a more in-depth model based on tant direct income gains for the poor. In addition, household data (e.g., those available from living the combined effects of higher productivity for mil- standards surveys). Such a model could trace the 160 East Asia Integrates TABLE 9.7 Marketing Costs and Margins for Rice in Cambodia, 2002 Trans- Whole- Inputs Farmer Collector Miller porter saler Retailer Total Transport cost 15 26 10 Operating cost 34 6 Input cost 355 453 469 526 561 605 Total costs 355 468 503 553 577 605 Price received 355 453 469 526 561 605 625 Value of byproduct 65 Total revenue 355 453 469 592 561 605 625 Profit (%) 355 98 (21.7) 1 (0.3) 88 (14.9) 8 (1.5) 27 (4.5) 20 (3.2) 244 Percent of total profit 40 1 36 3 11 8 100 Marketing margins (%) 56.8 15.7 2.6 9.1 5.6 6.9 3.2 27.5 Markup over farm gate price (%) 0 3.6 31 24 33 38 Notes: Costs are shown in riels per kilogram of paddy rice; yield is calculated at 2 metric tons per hectare; trans- port is from Battambang to Phnom Penh; and milling recovery is 0.64. Sources: Derived from data collected by World Bank Study Team, July 2002; ACI (2002a). TABLE 9.8 Marketing Costs and Margins for Export Rice in Vietnam, 2002 Collector/ Large Trans- VINA- Inputs Farmer husker miller porter Exporter FOOD Total Transport cost 32 4 2 Operating cost 37 47 14 13 Input cost 1,000 1,600 1,717 1,716 1,742 1,727 Total costs 1,000 1,669 1,769 1,732 1,755 1,727 Price received 1,000 1,600 1,717 1,716 1,742 1,727 1,736 Value of byproduct 144.10 Total revenue 1,000 1,600 1,717 1,860 1,742 1,727 1,736 Profit (%) 1,000 600 (37.5) 48 (2.8) 90 (4.9) 9 (0.5) ­27 (­1..6) 9 (0.5) 729 Percent of total profit 82 6 12 1 ­4 1 100 Marketing margins (%) 57.5 34.5 6.7 ­0.06 1.5 ­0.8 0.5 7.8 Markup over farm gate price (%) 0 7.3 16 9 8 9 Notes: Costs are shown in dong per kilogram paddy equivalent. Milling recovery is: brown rice, 0.77; white rice from brown rice, 0.85; white rice from paddy, 0.66. Export price is: (assuming 15 percent broken rice) US$172 per metric ton = 2,632 dong per kilogram. Sources: Derived from data collected by World Bank Study Team, July 2002; ACI (2002b). microeconomic behavior of poor households. and consider the effect of changes on each category Higher consumer prices for rice will undoubtedly in detail. To the extent that the farm gate prices of hurt households that are net consumers of rice, paddy increase more than the retail prices for rice, including the urban poor. Also, it is clear that net not only net sellers of rice but also some of the net sellers of rice in rural areas would benefit from buyers will benefit from increased exports, and a higher prices or increased trade volumes. An in- reduction in overall poverty might result (Minot depth study might look usefully at the distribution and Goletti 1998). of poor households in rural areas, seek to under- Rice farmers in both countries have limited stand who is going to be a net buyer or net seller, capacity to capture a larger share of the profit or Trade in Sectors Important to the Poor: Rice in Cambodia and Vietnam and Cashmere in Mongolia 161 FIGURE 9.2 Profit Structure of Rice Value Chain, Cambodia and Vietnam (US$/ton) Profits of each actor (US$ per 1,000 kilograms) 70 60 50 40 30 20 10 0 ­10 Farmer Collector/ Miller Transporter Exporter/ VINAFOOD/ Total husker wholesaler retailer Cambodia Vietnam Source: ACI (2002a, 2002b). marketing margins produced along the value chain. face relatively much higher costs for transport, pro- Comparisons of four countries at increasing levels cessing, marketing, credit search, and information. of development suggest that the marketing margins Policies that promote infrastructure, informa- of farmers are smaller in the less developed tion systems, and farm organizations will help to economies than in the more developed economies transfer some of the benefits to farmers. Better (Table 9.9). As national income per capita infrastructure will lower the cost of transport from increases, farmers get a progressively bigger size of farm gate to port and the marketing costs of traders. the total price (higher margins) and higher profits. For farmers, such reductions will likely result in The reasons farmers get less in Cambodia and Viet- higher prices and a greater share of the value added. nam than in the other countries shown are related Better information systems will help farmers to partly to infrastructure and partly to political econ- adopt better technology and orient marketing omy. In the United States, where the farm lobby has agents toward the higher-quality rice that is garnered substantial subsidies, leading to farm gate demanded by international markets. The result will prices higher than FOB prices, farmers are better be higher production and higher prices for the raw positioned to influence decisions than are farmers material, and both suggest that farmers will capture in Cambodia, who have very small and dispersed a larger share of the value added. The promotion of holdings, are less effective in influencing policy, and farm organizations will raise the bargaining power of farmers vis-à-vis millers, traders, and exporters. The development and strengthening of farm orga- TABLE 9.9 Marketing Margins of Farmers nizations, and their alliance with milling firms and in Cambodia, Vietnam, Thailand, and United States other agencies in the value chain, can lead to sub- stantial improvements in efficiencies and quality. Country Farm price/FOB price (%) Cambodia 72 Cashmere and the Poor in Vietnam 80 Mongolia Thailand 90 USA 111 Mongolia produces one-fourth of the world's sup- ply of raw cashmere and is the world's second Sources: Cambodia and Vietnam: fieldwork of largest producer after neighboring China. Over the World Bank Study Team, July 2002; ACI (2002a, last decade annual production has more than dou- 2002b). Thailand: personal communication, Tom bled. All cashmere production is traded, and almost Slayton, October 2002. United States: USDA-ERS (2002). all is exported. Cashmere exports of $70 million 162 East Asia Integrates accounted for 13 percent of Mongolia's official 44 percent of livestock herding households earning exports in 2001 and generally have been even less than $100 a year, another 44 percent earning higher (Table 9.10).13 between $100 and $500 a year, and the remaining 12 The cashmere subsector provides more than a percent earning more than $500 a year.14 third of the country's 2.4 million people with When COMECON (Council for Mutual Eco- incomes. It is the single largest employer--providing nomic Assistance) collapsed in 1991, Mongolia pri- jobs for more than 16 percent of the work force and vatized collective farms and livestock and removed accounting for more than 6 percent of GDP in price controls, fueling a boom in the livestock sec- 1993­2001 (see World Bank 2003)--and it is a prin- tor and ushering in substantial increases in the live- cipal source of livelihood for Mongolia's poor. Aver- stock population. The number of cashmere goats age household income from cashmere for herding doubled, to 10.3 million by the end of 2000, and the households is estimated at $220 for 2001. The varia- per capita incomes of herders rose from about $30 tion around this average, however, is substantial, with to $55 between 1991 and 1996. The characteristics TABLE 9.10 Cashmere in Mongolia Indicator Unit 1998 1999 2000 2001 Agriculture as share of GDP percent 39.9 40.3 34.2 28.4 Livestock sector as share of GDP percent 36.1 36.8 30.8 24.8 Agricultural labor as share of total labor percent 49.7 49.5 48.6 48.3 Goat herd as share of total livestock percent 33.6 32.9 34.0 36.8 Cashmere production as share of GDP percent 3.5 6.7 11.7 5.7 Cashmere exportsa metric tons 1,812.1 3,499.6 2,770.6 2,369.1 Cashmere exportsb US$ million 38.7 70.3 90.1 67.8 Cashmere exports as share of total exports percent 8.4 15.5 16.8 13.0 Cashmere exports as share of agricultural exportsc percent 41.6 53.7 53.8 50.0 Agricultural exports as share of total exports percent 17.3 24.7 26.5 24.5 Total production of cashmere metric tons 3,126.7 3,194.8 3,084.5 2,930.0 Total cashmere production per capita kilograms per capita per year 1.3 1.3 1.3 1.2 Cashmere revenue per capita US$ 15.2 26.5 49.1 25.0 Share of cashmere revenue per capita in GDP per capita percent 3.8 6.9 12.2 6.0 Total cashmere production/household kilograms with livestock per capita per year 11.4 11.8 11.5 11.4 Cashmere revenue/household with livestock US$ 124.4 223.4 423.4 229.0 Share of cashmere revenue/household with livestock in GDP per capita percent 31.1 55.9 105.9 57.3 Livestock per capita head 13.6 14.1 12.6 10.7 Livestock/household with livestock head 119.6 124.4 112.5 101.6 Goats per capita head 4.58 4.64 4.26 3.98 Goats/household with livestock head 40.2 40.9 38.2 37.4 a. Data for semiprocessed and finished cashmere products converted to raw cashmere. b. Official exports; value of smuggled cashmere is not included. c. Here the finished cashmere products (tops and garments) are excluded from total cashmere exports and agricultural exports. Sources: National Statistical Office of Mongolia, 2001, and World Bank staff estimates. Trade in Sectors Important to the Poor: Rice in Cambodia and Vietnam and Cashmere in Mongolia 163 of the herder population changed as well. The lian cashmere. These households are considered number of herding households more than doubled rich and are generally not at risk. during the 1990s, to 185,500 (one-third of house- Both the second and third groups of herders are holds nationwide) in 2001, reflecting an influx of nomadic and can afford to move to three or more nontraditional herders. different camps a year. They have access to winter Mongolia now has three groups of herders, dis- shelters by tradition or kinship, and they have tinguished by herd size, income levels, and grazing longer experience in selecting and occupying techniques. The first group is made up of house- rangeland in distant areas (World Bank 2001). holds that were in poverty after the collapse of COMECON and moved into cashmere production Potential Gains for the Poor from Trade in for lack of alternative income opportunities (World Cashmere Bank 2001).15 Unlike traditional herders, most of these families own small herds of up to 100 goats Because a high share of its production is exported, that they graze on common areas, and more than the cashmere subsector in Mongolia is much more 30 percent of them--with fewer than 30 goats on directly influenced by trends in the international average--live in urban areas. The average annual market, including quality standards, than is the rice income of these families from cashmere sales in sector in Cambodia and Vietnam. Looking ahead, 2000 was less than $55 per capita. Their intensive there are many reasons for optimism about the and stationary herd management is a major source international market. Demand for luxury goods is of serious land degradation. As their livestock growing worldwide and is well balanced globally. numbers have increased, land degradation has Demand for finished cashmere articles is highly become more acute, imposing costs on both the income-elastic, and trade in cashmere and cash- poor and the state. Overgrazing increases the inci- mere products is likely to continue growing.16 dence of respiratory diseases from dust, and it low- Quality is an important consideration in assess- ers productivity because herders have to travel fur- ing Mongolia's prospects for expanding export ther to tend their stock. It also exacerbates the earnings and incomes. Mongolia has historically impact of harsh winters on animals and on the supplied some of the world's best cashmere, but households that depend on them. These house- quality has declined significantly over the last 15 holds have been able to operate at the margin years as the country's production has increased. thanks only to the implicit subsidies--free public Quality cashmere commands a 30­40 percent price land and water--and explicit subsidies--free premium in international markets, and quality dis- restocking of goats after the devastation of livestock counts cost Mongolian herders about $18 million in the successive harsh winters of 2000 and 2001. in 2001--23 percent of the income of the average Taking into account the cost of inputs and land household with livestock. degradation, the social returns from the livestock Higher export earnings from cashmere would practices of these households may be negative. help to boost the incomes of a large number of The second group of herders, who own herds of Mongolians. Although information about the rela- from 101 to 500 head, account for the bulk of cash- tive position of herder households in the income mere production. They own 38 percent of the distribution is not available, it is clear that many of national herd and produce more than 65 percent of these households represent the poorest segments of Mongolia's cashmere. Most live in rural areas, pre- Mongolian society. However, the sustainability of viously worked in state-owned livestock coopera- cashmere production as a livelihood for households tives, have herding experience, and bought livestock in the first group just described is open to question. using vouchers when collective enterprises were privatized. They depend almost entirely on live- Value Chain Analysis: Constraints on Production stock herding for income. Not all of these herders and Trade, and Policy Options are poor, but all are vulnerable to bad weather and changes in demand. Many different actors--herders, traders, processors, The third group makes up 2 percent of all herder retailers, and exporters--take part in the value chain families and produces about 3 percent of Mongo- linking producers to final consumers. Others are vet- 164 East Asia Integrates erinary service providers, local and central govern- of cashmere without undue environmental dam- ment officials, herder organizations, graders, and age. The combination of easy entry into cashmere external partners, including donors and nongovern- production and failing institutions for key inputs mental organizations. Most of the power within the has led to deterioration of the water supply, col- value chain resides with the state-owned processing lapse of the fodder industry, and an absence of pas- firm and retailer, Gobi; herders are the largest group ture management. These factors have therefore put of actors but perhaps the least powerful. pressure on environmentally fragile land and This section looks at the policy and institutional increased both the private and social costs of pro- constraints faced by each of the main actors and duction. suggests possible policy responses. Like the situa- Mongolia's existing legal and administrative tion for rice in Cambodia and Vietnam, many of mechanisms for controlling grazing are unlikely to the constraints in the sector represent shortcom- be able to resolve the problems of land degradation. ings of sectoral policies and institutions rather than The primary input for cashmere--grazing and pas- of trade policy. Once again, value chain analysis tureland--continues to be a public good, and pri- proves particularly useful in pointing out the par- vatization is not a realistic option.17 Strengthening ticular challenges to developing efficient and high- social controls would require introducing provi- value products, with returns that benefit the poor, sions into the land laws to allow groups of herders, in the context of transition economies. organized as legal entities, to exclude others from The difficulties faced by transition economies are well-defined areas of pastureland. Other options particularly evident in Mongolia where, with tradi- would include land user fees, whose feasibility tional practices under stress and state institutions would have to be assessed, or modifications of the being dismantled, efficient market development is head tax on livestock to ensure that it reflects the facing the challenge of developing new institutions social costs associated with land degradation.18 and overcoming the legacy of passing ones. For Improvements in land use laws or a revision of the example, in overcoming the history of collectiviza- livestock tax would provide incentives for herders tion in Mongolia, economic agents must spend to rationalize their stocks, pool resources, and some time allaying suspicions and work to develop employ more efficient grazing techniques. Impos- new and efficient institutions through collective ing fees on land use would also spur the develop- action with a view toward benefiting from ment of a sustainable fodder market. Rehabilitating economies of scale. The development of new herder water bores is crucial to increased livestock produc- cooperatives and responsive private sector veteri- tivity and supply and could generate rural employ- nary services to replace defunct communal and ment. Bores and wells are constitutionally the public service institutions has been a slow process. responsibility of local government administrators, Two requirements--understanding the need for but the Gobi Initiative, funded by the U.S. Agency quality control and grading systems and developing for International Development (USAID), has an effective demand for such systems through shown that herders are willing to pay for water appropriate price signals--have yet to be met. The facilities if exclusive use is guaranteed. Support breakdown of traditional pastureland and risk man- from government and donor agencies is needed in agement practices is resulting in overgrazing and scaling up this successful public-private partner- severe losses of herds after a particularly harsh win- ship nationwide. ter (dzud), and an inability to maintain wells and Mongolia is vulnerable to severe winters.19 Bet- bores. Likewise, missing markets, weak institutions, ter herding practices can help to mitigate the and an absence of ancillary markets in services-- impact of severe weather, but herders will remain such as a lack of physical marketplaces, traders, and highly vulnerable without insurance. Only about insurance markets--are challenging the develop- 0.3 percent of Mongolia's livestock was insured in ment of value chains in Mongolia. Indeed, this situ- 2001, when Mongolia had a particularly harsh win- ation plagues many transition economies. ter.20 It is estimated that the government spent that year about 2 billion tugriks (about 1.1 percent of Herders. Herders face several serious impediments GDP) on providing subsidized fodder and other in improving the quality and expanding the supply disaster relief services. Trade in Sectors Important to the Poor: Rice in Cambodia and Vietnam and Cashmere in Mongolia 165 A well-developed insurance market for livestock Tsaiz market in Ulaanbataar--is 600­1,000 kilome- could reduce these costs substantially for the gov- ters from most regional production centers. Almost ernment and help to stabilize herders' incomes. The all herders must sell their cashmere either to traders development of insurance options faces important at the farm gate or at informal provincial market- challenges, including herders' nomadic lifestyle, the places, at discounts of 10­45 percent from capital difficulties of gathering adequate information on city prices. Generally, they sell as individuals, often herd management techniques, and a lack of suffi- on a barter basis, with little knowledge of market cient skills and information to determine appropri- demand and little ability to influence market out- ate insurance premiums. But recent pilot projects comes. They incur additional expenses for trans- aimed at providing index-based livestock insurance porting goods to the market day locations with no provide a good start to the introduction of market- certainty of concluding sales. based techniques for addressing risks.21 The persist- Steps to stimulate growth in Mongolia's cash- ence of an implicit government guarantee of live- mere industry, and in poor herders' incomes in par- stock assets will discourage the development of an ticular, will require strategies to improve marketing insurance market and delay the needed exit of mar- channels, so that herders are able to get their prod- ginal producers. Donors and the government need ucts to processors at least cost and so that they are to desist from restocking herds free of charge and at remunerated fairly for superior quality. Develop- levels too high for long-term sustainability. ment of regional market centers, such as those in Finally, restoring quality is critical to increasing the Gobi Initiative funded by USAID, and improve- the value of Mongolian cashmere and the incomes ments in infrastructure could significantly increase of herders. Reasons for the drop in quality over the the incomes of rural herders. Other promising ini- last decade include herders' lack of information on tiatives are the development of herder cooperatives grades and prices, and on veterinary and other herd and attempts by government to improve herders' improvement topics, together with a lack of grad- access to markets by creating wholesale networks. ing and testing facilities. When herders lack market Mongolia has a good law on cooperatives, but assis- information, they are vulnerable to exploitation by tance is needed to improve governance and finan- processors and traders, and without price differen- cial management in cooperatives. tials between high-quality cashmere and other Vertical integration between herders and proces- grades, herders lack incentives to invest in improv- sors could in principle lead to a sharing of trans- ing their herds. port costs between agents, improve herders' income The way to restore the quality of Mongolian security, and reduce the uncertainty of supply for cashmere is through herd selection, superior genet- processors. But developing such strategic alliances ics, and price incentives that reward quality would be challenging; with their nomadic lifestyle, improvement. Improvements in the legal and regu- herders would find it difficult to enter into formal latory framework could encourage the establish- marketing agreements. Moreover, such formal rela- ment of innovative joint venture agreements and tionships require a lot of trust and strong social partnerships in research and development. New cohesion, which has been eroding in Mongolia. technological developments in rearing cashmere Credible mechanisms to enforce contractual obli- goats could improve the supply and quality of gations would have to be in place and so would a Mongolian cashmere without damaging the envi- system that rewards herders for cooperation and ronment. The provision of services for testing and organization. grading should be encouraged, and the Mongolian herders' association should use internationally Processors. Since Mongolia's transition from a accepted grading standards. Grading will allow for command economy to a more market-oriented product differentiation and will allow herders and economy, the number of cashmere processing and processors to determine the best mix of quality and dehairing companies has risen to more than 30, and price for production and sale. herders can trade directly with international buy- ers, middlemen, or trading agents. Although the Traders and Marketing and Distribution Chan- state abandoned cashmere producers in the wake of nels. Mongolia's one major cashmere market--the the transition, it retained a heavy influence in cash- 166 East Asia Integrates mere processing through the ownership of Gobi, ering the supply of raw cashmere to the domestic the largest processing company, and other joint processing industry and exacerbating the problems venture partnerships in Mongol-Amical and Mon- associated with low-capacity utilization. Smuggling Forte.22 Processing capacity exceeds domestic sup- has grown rampant, given the length and porosity of ply even in the best of years, and in 2000 the four the border with China, and Mongolia's final goods large factories that account for about two-thirds of exports have remained quite limited (Figure 9.3). processing capacity were running at only one-fifth Proceeds from the export tax in 2001 amounted of their capacity. to less than 0.01 percent of total government rev- Policies to reform this segment of the value enues. The decline in recorded raw cashmere chain need to level the playing field among proces- exports cost Mongolia about $100 million in offi- sors; Gobi benefits from continued government cial export earnings between 1996 and 2001. In support in raw cashmere procurement. The indus- addition, by pitting the interests of processors try's excess processing capacity also should be against those of herders, the export tax may have rationalized, which adds to production costs. damaged the prospects for forming strategic alliances within the domestic supply chain. Cashmere Trade Policies. In 1994 the government The case for retaining the export tax is uncon- banned exports of raw and washed cashmere on the vincing. The tax has shown itself to be unenforce- grounds of the need to "protect domestic industry able, but if it were to have its intended effects, the from external competition and increase domestic result would be a transfer of resources in the value value added." In 1997 the export ban was replaced chain away from the poorest members to the more by an export tax on raw cashmere of 4,000 tugriks powerful ones. Abolishing the tax is likely to have per kilogram, equivalent to 13 percent in 2001 beneficial consequences both for herder incomes prices. and for the overall development of the cashmere Although these policies were meant to protect subsector. domestic processing firms by providing them with larger volumes of cashmere at cheaper prices, and to Overview of Recommendations increase the value added by the cashmere subsector to the economy, these goals have not been met. As a matter of priority, measures are needed to stop Rather, these policies have stimulated activities that the degradation of grazing land and ensure more circumvent the ban and the export tax, in fact low- secure supplies of water and fodder. FIGURE 9.3 Share of Cashmere Products in Mongolia's Total Cashmere Exports, 1993­2001 Percent 100 90 80 70 60 50 40 30 20 10 0 1993 1994 1995 1996 1997 1998 1999 2000 2001 Cashmere, greasy/raw Cashmere, dehaired Cashmere tops Cashmere garments Source: Government of Mongolia. Trade in Sectors Important to the Poor: Rice in Cambodia and Vietnam and Cashmere in Mongolia 167 For herders, the scope is significant for expand- another, and others would increase private costs to ing incomes and reducing poverty by upgrading reflect more accurately the economic costs of input the quality of cashmere produced. Simple simula- provision and of externalities, notably environmen- tions suggest that closing half the quality gap tal degradation from overgrazing. between Mongolian and Chinese cashmere would Without subsidies, the marginal segments of the have increased average annual incomes for herder herder population might no longer find cashmere households from $220 to $263, or by 20 percent, in production a sustainable activity. This possibility 2001. This increase is a feasible target over the next raises difficult questions about the role of livestock 6­10 years. herding in Mongolia, which has acted as a safety net Reducing poverty will require a combination of for many households adversely affected by the tran- measures to increase both the value of output-- sition from a planned to a market economy. For through upgrading quality--and herders' share of such families, a transition out of herding would that value. A host of measures would help to increase have to be phased over time and accompanied by herders'share of profits in the value chain: increasing measures to stimulate alternative employment-- their security of access to pastureland; enabling the for example, in storage, maintenance and repair of development of input and insurance markets; pro- wells, and fodder production. viding livestock support services and reliable infor- mation about quality standards and prices that are Conclusions differentiated by quality; developing markets in regional centers, reducing transport costs, and An analysis of value chains provides a good under- increasing the bargaining power of herders; and fos- standing of the behavior and constraints of differ- tering cooperative solutions to help households ent actors involved in the addition of value as a attain economies of scale or strategic alliances to product moves from producers to consumers, and increase the security and quality of supply. at the same time allows the analyst to identify the Government has a necessary role in improving the major constraints in the sector as a whole. Such policies and institutions to support the cashmere analysis reveals that the poor can benefit from the value chain. For the most part, it must play an expanded opportunities presented by global inte- enabling role, but some instances of market failure gration, provided that these institutional con- require nonmarket solutions. These solutions include straints are addressed. Giving voice to poor pro- pastureland management, veterinary services, stan- ducers' interests by placing these issues on the dard setting, and dealing with information asymme- policy agenda is crucial for fostering reforms that try in insurance markets. In addition, there is a case unleash the productivity potential of poor people for revising Mongolia's trade policies for cashmere: and increase their bargaining power. although the government should pursue relaxation of Growth in trade can be achieved through the ban on breeder stock imports from China, in the improvements in productivity, combined with context of discussions on China's accession to the improvements in postproduction technologies and World Trade Organization the tax on exports of raw marketing and institutions. This finding is true for cashmere appears counterproductive. rice in Vietnam and Cambodia, where a small pro- Detailed household survey data are not yet avail- portion of total production is actually traded and able for Mongolia, and without them it is difficult to an even smaller proportion is exported. It is also estimate the net impact of the recommended true for cashmere in Mongolia, where all output is changes on the welfare of various participants in the traded and almost all is exported in some form. cashmere supply chain.23 It seems clear, however, Both rice and cashmere are commodities central that upgrading cashmere quality is likely to increase to the livelihood of a large proportion of poor value, and that reducing transaction costs (by devel- households in the countries studied, but between oping regional market centers) and reducing volatil- the countries there are differences in the prospects ity (through insurance) would yield unambiguous for income gains from improvements in the supply welfare gains. Some of the recommendations, par- chain. These different prospects stem largely from ticularly elimination of the cashmere export tax, differences in the nature of the commodities-- would transfer gains from one set of participants to which affect the prospects for growth in the inter- 168 East Asia Integrates national market, for gains from upgrading quality, 2. For a brief history of rice cultivation in the two countries, and for gains in the relative share of producers, who see the technical appendix to this chapter at www.world bank.org/eaptrade. are the poorest participants in the value chain. 3. All dollar amounts are current U.S. dollars. Although cashmere is a luxury good for which 4. All appendix tables cited in this chapter can be found at demand is highly income-elastic, most rice is a www.worldbank.org/eaptrade. 5. As shown in the technical appendix to this chapter at basic good where growth in the overall market is www.worldbank.org/eaptrade. sluggish and tied to population growth rather than 6. Domestic resource cost (DRC) is an indicator of compara- income trends. Moreover, few poor households tive advantage. It measures the ratio of value added from domestic, nontraded activities to the foreign exchange engaged in rice production are likely to be able to earned or saved from domestic production. A DRC of less raise their incomes from producing high-quality than 1.00 suggests comparative advantage and efficiency in rice; the niche market for high-quality rice is small, production. and Cambodia and Vietnam have limited potential 7. Quotas have been lifted, export taxes are zero, import tariffs are low in Cambodia (7 percent) and, even though tariffs for supplying it. By contrast, for cashmere produc- are relatively high in Vietnam (at 30 percent), they are not ers the prospects for income gains from quality binding. improvements are sizable and would have wide- 8. The background papers for this chapter contain a detailed analysis of the constraints for each actor along the chain spread impacts among all herder households. (see ACI 2000a, 2000b). Based on these papers, a more Within the value chains for both rice and cash- extensive analysis of the specific constraints and policy rec- mere, power rests not with producers but with pro- ommendations related to each link of the value chain is detailed in Appendix Tables 5­12 at www.world cessing firms, especially state-owned Gobi in Mon- bank.org/eaptrade. golia, millers in Cambodia, and the state-owned 9. These losses, which occur at different stages after harvest, trading company in Vietnam. The share that pro- and are the result of moisture, pests, rodents, outdated ducers capture of the final value of the products milling technology, and pilferage. Losses at the village level can be substantial, from 5 to 20 percent of production. differs among countries--and so, therefore, do the 10. Details of these options are described in the technical prospects for raising their incomes through gains in appendix to this chapter at www.worldbank.org/eaptrade efficiency and strategic alliances in the form of 11. This breakdown may not be an appropriate metric for cross-country comparisons of the efficiencies of agents either greater horizontal or vertical integration. In (such as farmers), because farmers in some countries may Vietnam, where farm gate prices are 80 percent of take on more functions (e.g., postharvest storage) than they the border price for rice and farmers capture as do in other countries. In Vietnam and Cambodia, however, the comparison would appear to be appropriate. much as 82 percent of the profits from the value 12. In Cambodia, these two groups together get about 80 per- chain, there is less scope for gains from either cent of the profits, whereas in Vietnam they get about 90 improving efficiency or restructuring the value percent of the profit (see Figure 9.1). chain. In Cambodia, where farm gate prices are 13. These figures considerably understate the importance of the industry to the Mongolian economy. It is estimated that somewhat lower, at 72 percent of the border price, about 45 percent of the raw cashmere produced in 2000-- and the farmers' share of total profits is substan- worth about $50 million--was smuggled to China. This tially less, at 40 percent, there is considerable scope estimate is derived from deducting official exports and sales to domestic processors from total known production. for productivity and efficiency gains along the 14. 2001 Household Income and Expenditure Survey, Mongo- value chain and for improvements in farmers' bar- lian National Statistics Office. gaining position. In Mongolia, it is difficult to 15. The collapse of the Soviet Union ushered in a difficult period of transition for Mongolia. The Mongolian econ- decompose the value chain, but it is clear that omy imploded, and real GDP declined by 20 percent herders capture only a very small portion of profits between 1989 and 1993. The one bright spot was the and that they have significant potential for raising expansion in agriculture and herding, fueled by wholesale their incomes. Finally, all three countries are highly privatization of the livestock sector in 1991. The privatiza- tion of livestock using the voucher system was largely equi- vulnerable to external shocks, and coping with risk table and successfully put assets in the hands of poor is a major concern for households engaged in both households. Every Mongolian citizen received vouchers rice and cashmere production. worth 10,000 tugriks (roughly $10) for the purchase of state assets, particularly livestock and small businesses, but only members of state-owned cooperatives (about 40 per- Endnotes cent of the population) could use vouchers to purchase livestock. The voucher distribution system was not eco- 1. Early on, Fischer and Gelb (1991) flagged the role of insti- nomically optimal, because it created incentives for pro- tutional reforms in the transition. duction but encouraged suboptimal herd sizes. Over time, Trade in Sectors Important to the Poor: Rice in Cambodia and Vietnam and Cashmere in Mongolia 169 concentration was aimed at yielding an economically sus- Goletti, F., A. Bhatta, and C. Srey. 2002. "Farmer Survey: Tabula- tainable herd size, but this process was interrupted in 1999 tion of Results." Discussion Paper No. 5. Agricultural Sector by external shocks. Development Program, ADB TA 3695-CAM. Phnom Penh, 16. Trends in the global cashmere market are outlined in the October. technical appendix to this chapter at www.world Havrylyshyn, Oleh, and Ron van Rooden. 2003. "Institutions bank.org/eaptrade. Matter in Transition, But So Do Policies." Comparative Eco- 17. Private ownership is prohibited under Mongolia's constitu- nomic Studies 45 (1): 2­24. tion and its 2002 land laws. Traditionally, private herds have Hayami, Yujiro, and Toshihiko Kawagoe. 1993. The Agrarian used public lands, with herders having rights to certain areas Origins of Commerce and Industry. New York: St. Martin's at certain times of the year. Mobility and knowledge are Press. essential if the pastures are not to suffer from overgrazing. Kaplinsky, R. 1999. "Globalisation and Unequalization: What 18. Mongolia already has a specific tax on livestock, but the Can Be Learned from Value Chain Analysis." Journal of level of the tax would have to be reassessed to determine Development Studies 30 (2): 117­46. whether it is sufficient to capture the social costs associated Minot, N., and F. Goletti. 1998. "Export Liberalization and with land degradation. Local administrators already have Household Welfare: The Case of Rice in Vietnam." American detailed records of stock ownership by herder that enable Journal of Agricultural Economics 80 (4): 738­49. the tax to be administered efficiently. North, Douglass C. 1991."Institutions." Journal of Economic Per- 19. The winters of 2000 and 2001 were particularly harsh, and spective 5 (1): 97­112. drove down the stock of goats by more than 700,000 ani- OPCV (Overseas Projects Corporation of Victoria). 2002. "Agri- mals. Rising prices cushioned the impact of declining pro- cultural Sector Performance Review." Final Report submit- duction and export volumes, but more than 19,000 ted to Asian Development Bank. TA 3695-CAM. Manila. herders--about 8 percent of the households with live- Porter, M. E. 1990. The Competitive Advantage of Nations. New stock--lost all of their animals. York: Free Press. 20. They were insured by Mongol Daatgal, a state-owned Reardon, Thomas, C. Peter Timmer, Christopher B. Barrett, and insurance company. Julio Berdegue. 2003. "The Rise of Supermarkets in Africa, 21. For example, the livestock insurance program under the Asia and Latin America." American Journal of Agricultural Sustainable Livelihoods Project supported by the World Economics (December). Bank is a public-private initiative aimed at developing a Royal Government of Cambodia. 1999. Cambodia Socioeco- risk index that private insurance companies would use in nomic Survey 1999. National Institute of Statistics, Ministry offering insurance to livestock owners to cover risks arising of Planning. from drought or other weather-related events. Details of ------. 2001. Statistical Yearbook 2001. National Institute of this scheme are in the technical appendix to this chapter at Statistics, Ministry of Planning. www.worldbank.org/eaptrade. Sachs, Jeffrey, Clifford Zinnes, and Yair Eilat. 2000. "The Gains 22. Gobi remains in state hands, despite having been slated for from Privatization in Transition Economies: Is `Change of privatization for some time. Ownership' Enough?" CAER Project II, Harvard Institute for 23. A new household survey, including a special module on International Development. cashmere production, is currently under way and will yield Smallbone, David, and Friederike Welter. 2003. "Institutional useful information to enable a more quantitative estima- Development and Entrepreneurship in Transition tion of both net impacts and distributional changes. Economies." Paper presented at ICSB 48th World Confer- ence--Advancing Entrepreneurship and Small Business, Belfast, June 15­18. References Socialist Republic of Vietnam. 1998. Vietnam Living Standard Survey 1998. Government Statistics Office. ACI (Agrifood Consulting International). 2002a. "Rice Value UNCTAD (UN Conference on Trade and Development). 2000. Chain Study: Cambodia." Report Prepared for the World "Strategies for Diversification and Adding Value to Food Bank. Phnom Penh, September. Exports: A Value Chain Perspective." UNCTAD/DITC/ ---------. 2002b. "Rice Value Chain Study: Vietnam." Report COM/TM/1. Geneva. Prepared for the World Bank. Ha Noi, September. USDA-ERS (United States Department of Agriculture, Eco- ANZDEC. 2000. "Agricultural Sector Program, Inception nomic Research Service). 2002."Rice Outlook." Report RCS- Report." ADB TA 3223-VIE. March. 0802. Washington, D.C. Boeva, Bistra. 2002. "Governance in the Transition Economies: Weatherspoon, Dave D., and Thomas Reardon. 2003. "The Rise Current Issues in Bulgaria." Center for International Private of Supermarkets in Africa: Implications for Agrifood Sys- Enterprise, Washington, D.C. tems and the Rural Poor." Development Policy Review 21 (5): Braguinsky, Serguey, and Grigory Yavlinsky. 2000. Incentives and 333­55. Institutions: The Transition to a Market Economy in Russia. World Bank. World Development Report 1996: From Plan to Mar- Princeton, N.J.: Princeton University Press. ket. New York: Oxford University Press, 1996. Dolan, C., and others. 1998. "Horticulture Commodity Chains: ------. 2001. "Mongolia Participatory Living Standards Assess- The Impact of the UK Market on the African Fresh Vegetable ment." Washington, D.C. Industry." IDS Working Paper 96. Institute for Development ------. 2003. "From Goats to Coats: Institutional Reforms in Studies, University of Sussex. the Mongolian Cashmere Sector." Washington, D.C. Fischer, Stanley, and Alan Gelb. 1991. "The Process of Socialist Economic Transformation." Journal of Economic Perspectives 5 (4): 91­106. 10 Trade and Labor Market Vulnerability in Indonesia, Republic of Korea, and Thailand François Bourguignon Chor-ching Goh The links between openness to international trade Opponents of integration, by contrast, see grow- and economic growth have been the subject of ing inequality between and within nations where many research papers. Although somewhat contro- proponents see average gains in welfare. They speak versial, evidence suggests that openness to interna- about the uneven distribution of gains from open- tional trade tends to raise national income (Frankel ness to trade and the increased uncertainty and vul- and Romer 1999; Irvin and Tervio 2002).1 More nerability associated with being more exposed to controversial are debates about the distribution of the vagaries of international markets and more sus- the benefits from higher economic growth--both ceptible to shocks that can be transmitted rapidly across countries and within countries. Proponents across countries and regions as evidenced during of integration posit that for a small, labor-abun- the 1997­98 East Asian financial crisis. dant economy, opening up to trade leads to rising Recent empirical work analyzing the determi- wages and greater stability. They point to the sub- nants of distributional changes over time shows stantial increases in average real wages in open that indeed some evidence (from Latin America as developing countries over the last several decades well as some Asian countries) suggests that greater as evidence that international trade does indeed integration may be generating higher returns to increase demand for the abundant factor--labor in skills, which, when coupled with unequal access to most developing countries--much like trade the- education, is leading to growing wage inequality in ory predicts. They also argue that trade benefits the some liberalizing economies. Other papers also poor at least as much as the average household, and lend support to the hypothesis that macroeco- that opening a small economy permits access to nomic shocks have become more frequent in more stable international markets and thus greater today's more integrated international markets, even opportunities for hedging and insurance. though their severity and duration may be less We would like to thank the following people for their contributions to this chapter: Tamar M. Atinc, Mun S. Ho, Caro- line Hoxby, Lawrence Katz, William Maloney, and Martin Rama for valuable suggestions and ideas when we embarked on this work; Vivi Alatas (Indonesia) and Dae Il Kim (Korea) for excellent analysis; and Anant Chiarawongse and Binh Nguyen for data assistance. We are also grateful for comments from participants in the seminars held at the World Bank's Tokyo Resident Mission; the Institute of Southeast Asia Studies in Singapore; the Thailand Development Research Institute; the Fourth Asian Development Forum in Seoul; and the World Bank's Jakarta office. All errors remain ours. 171 172 East Asia Integrates because governments are better equipped to affected more by inflation than by openness, and respond to them than in the past. But little empiri- that in Argentina, Bolivia, Brazil, Chile, and Peru cal work has been done to test whether a greater wages became more stable in the 1990s. Although propensity to experience shocks at the macro level job turnover rates increased and informal sector translates into more vulnerability at the level of employment rose, unemployment was stable over individual workers. the period 1970­90, and there was no evidence of a It is indeed possible that many of the concerns higher probability of unemployment or of longer expressed by the opponents of integration reflect unemployment spells. Fajnzylber and Maloney anxiety about the future prospects for employment (2000) found no evidence in Chile, Colombia, and and wages even if the evidence points to an increase Mexico that trade liberalization has increased in average returns to labor. For East Asian countries workers' insecurity. that are closely integrated into world markets, and This chapter is a first attempt to investigate the becoming more so, an examination of whether this topic in an East Asian context. The analysis was concern indeed rests on solid ground should have applied to three East Asian countries: Indonesia, crucial implications in three policy areas: trade, Republic of Korea, and Thailand. We sought to labor markets, and social safety nets. determine whether workers' earnings and working Within the large literature on wage inequality and hours have fluctuated more as countries have liber- wage differentials in relation to trade liberalization, alized their trade regimes. Because no data were only a handful of studies--mostly on Latin Ameri- available for testing that hypothesis directly, a can economies, perhaps because macroeconomic twofold approach was followed. First, we asked volatility appears to be structurally higher there-- whether output and average wages become more examine trade, volatility, and insecurity in the labor volatile as the economies become more open. Sec- market, most of them taking changes in employ- ond, we looked for signs that workers in sectors and ment as the indicator of vulnerability. The literature industries highly exposed to world markets are suggests that trade liberalization has only a small more vulnerable to falling into poverty than those impact on aggregate employment. Papageorgiou, in sectors less exposed. Choksi, and Michaely (1990) found that, by and We begin here by describing the evolution of large, trade liberalization did not significantly raise growth and volatility of output and wages as the unemployment in the 19 countries they examined. economies became more open. Then, in the first Revenga (1994) discovered that Mexico's trade stage of the analysis, we examine the dynamics of reform of 1985­88 reduced employment modestly, earnings and employment across sectors and but did not reduce wages. Cox and Edwards (1996) industries with different degrees of exposure to found that Chile's trade liberalization of the 1970s trade and thus external shocks over time. Entering affected workers' duration of unemployment, but the second stage of the analysis, we explore more that its effect was small relative to those of other vari- formally the relationship between trade intensity ables, and declined over time. Currie and Harrison and workers' vulnerability to falling into poverty. (1997) discovered that during trade liberalization Because our findings for all three countries yielded between 1984 and 1990 in Morocco, changes in similar conclusions, for the sake of brevity we pre- import tariffs and quota coverage had no impact on sent only selected examples in this chapter.2 aggregate employment. Using rotating panel house- hold surveys, Arango and Maloney (2002) saw some Overview of Trade Liberalization evidence of a higher incidence of involuntary sepa- and Labor Market Volatility ration, mostly among skilled workers, in sectors that are opening to trade in Mexico and Argentina, but This chapter examines the relationship between the impact is transitory. trade liberalization and vulnerability by comparing One concern has been that trade liberalization two time periods in each country under study: one may make the demand for labor more elastic. Issues in which the economy was more closed and one in of worker insecurity and economic openness in which it was more open. The periods covered in the Latin America are summarized in De Ferranti and three countries differ, depending on data availabil- others (2000), who found that wage volatility is ity. For Indonesia and Thailand, we used labor force Trade and Labor Market Vulnerability in Indonesia, Republic of Korea, and Thailand 173 surveys, and for Korea, we used both the establish- whether the usual structural measures such as tariff ment surveys (known as occupation wage surveys) rates, nontariff barriers, and tariff revenue conform and labor force surveys. The labor force data cover to this view. 1991­2000 for Thailand, 1986­2001 for Indonesia, and 1976 and 1981­2000 for Korea. Measures of Trade Liberalization We also compared workers' experiences accord- ing to the sectors and industry groups in which All the economies examined in this study were they work, because these sectors and groups have already open in the early 1970s (see, for example, different degrees of exposure to world markets. For Sachs and Warner 1995), but they have become each country, sector definitions are governed by the more open since then. Table 10.1 shows the average source data. Agriculture, fishery, and forestry were import tariffs in manufacturing during the 1980s excluded from the tradable sector because of data and 1990s and their rates of reduction from one constraints.3 decade to the next. In manufacturing, the average Korea has three broad sectors--manufacturing; tariff dropped by 26 percent in Indonesia, by 42 construction and social services; and "others," percent in Korea, and by 8 percent in Thailand. In which include public utilities, wholesale and retail the category of duty called the most-favored-nation trades, transport, communications, and financial status rate, the reductions were even bigger. services.4 In Indonesia, industry classification is As for the import tariffs faced by specific indus- restricted to a few categories, and so it has only two try groups, Figure 10.1 shows the average import sectors: manufacturing and nontradables. Non- tariff rates in 11 industry groups within manufac- tradables include all nonmanufacturing and non- turing. In all three countries, tariffs decreased primary industries.5 Within the manufacturing sec- noticeably in all these industries, with a few excep- tors of Thailand and Korea, we define groups of tions, from the 1980s to the 1990s. industries that have differing degrees of exposure to Core nontariff barrier measures followed the trade, determined by the share of trading volume same evolution (Table 10.2). They fell between (exports and imports) in their output.6 1989­94 and 1995­98 in all three countries, more The main question explored in this section is or less at the same rate as tariffs. Changes in indi- whether changes in trade regimes in the three vidual measures such as licensing, variable levies, economies under study were accompanied by and minimum pricing also generally suggest that changes in the volatility of various aggregate output these economies have become more open in practi- and labor market indicators. But, first, were the cally all dimensions. three economies indeed more open in the 1990s That the preceding measures have been effective than in the 1980s, as commonly believed? To is confirmed by the evolution of import tax rev- answer this question, we begin by determining enue. In all three countries in the 1980s and the TABLE 10.1 Average Import Tariffs (Applied Duty Rates) and Percentage Change in Tariffs of Indonesia, Republic of Korea, and Thailand between 1980s and 1990s Manufacturing Indonesia Korea Thailand Average tariff (applied duty rate) during 1980s 24.23 18.74 39.04 Average tariff (applied duty rate) during 1990s 17.97 10.96 36.06 Percentage reduction (applied duty rate) 25.84 41.50 7.64 Percentage reduction (most-favored-nation status rate) 45.65 46.31 22.46 Source: Simple averages from United Nations Conference on Trade and Development (UNCTAD). 174 East Asia Integrates FIGURE 10.1 Decade Averages of Import Tariffs for Manufacturing Industries: Indonesia, Republic of Korea, and Thailand, 1980s and 1990s 90 90 Indonesia Korea 80 80 70 70 60 1980s 60 1980s 1990s 1990s 50 50 40 40 30 30 20 20 10 10 0 0 Paper Paper tobacco apparel plastics metals metals metals metals furniture tobacco apparel plastics furniture Chemicals Petroleum Chemicals Petroleum and and and equipment equipment and manufactures and and and and manufactures Primary and Primary and Fabricated Fabricated Textiles Wood Rubber Other Textiles Wood Rubber Other beverage, beverage, Machinery Machinery Food, Food, 90 Thailand 80 70 1980s 60 1990s 50 40 30 20 10 0 Paper tobacco apparel plastics metals metals furniture Chemicals Petroleum and and and equipment and manufactures Primary and Fabricated Textiles Wood Rubber Other beverage, Machinery Food, Source: United Nations Conference on Trade and Development (UNCTAD). 1990s, these revenues fell as a proportion of total enue to about 8 percent. In Thailand, import rev- trading value as well as of total government revenue enue also fell substantially in terms of both total (Table 10.3), suggesting that these economies revenue and total import value. This downward became effectively less protected. Korea's import trend is much less pronounced for Indonesia, but tax revenue fell from about 16 percent of total rev- that country's trade liberalization also proceeded Trade and Labor Market Vulnerability in Indonesia, Republic of Korea, and Thailand 175 TABLE 10.2 Core Nontariff Barrier Measures: Indonesia, Republic of Korea, and Thailand, 1989­99 Core nontariff barrier measures Country 1989­94 1995­99 Indonesia 53.6 31.3 Korea 50.5 25.0 Thailand 36.5 17.5 Note: Table shows the percentage of items with various types of nontariff barrier measures among all Standard International Trade Classification (SITC) or harmonized system (HS) two-digit products. Core nontariff barrier measures are licensing, prohibition, quotas, and administered pricing. Nonauto licensing includes various forms of administrative approvals. Source: Michalopoulos (1999). TABLE 10.3 Decade Average Shares of Import Tax Revenue in Total Government Revenue and Total Trading: Indonesia, Republic of Korea, and Thailand, 1980­2000 Import duties as Import duties as percentage of total percentage of Country Decade import value government revenue Indonesia 1980s 4.6 4.7 1990s 3.7 4.5 Korea 1980s 8.3 16.3 1990s 5.1 8.3 Thailand 1980s 12.2 21.1 1990s 7.6 16.4 Source: Trade Analysis and Information System of United Nations Conference on Trade and Development (UNCTAD). on the export side, with export duties as a propor- If there are short-term adverse shocks, they are tion of total tax revenue falling by more than half. likely to be transmitted into fluctuations of earn- ings and employment. Unemployment rates have been stable and uniformly low in all these Volatility of Some Aggregate Indicators economies as they have become more open, and for Did this increased openness make these economies Korea and Thailand, where sectoral unemployment more vulnerable to international shocks, with rates are available, there are no significant differ- greater fluctuations in the growth of output and ences among industries of varied intensity in trade. wages? The evolution of the observed volatility in Less unemployment volatility could have been the growth of the gross domestic product (GDP), achieved at the expense of more variability in manufacturing value added, and wages gives a first wages. However, in general, the year-to-year fluctu- rough answer. Table 10.4 shows the volatility of ation of wages was much less in the more open GDP growth rates by decade, as measured by the 1990s than it was in the previous decade. Moreover, coefficient of variation.7 According to the table, the we found no evidence that volatility of wages was 1990s--the decade of greatest openness (up to and greater in more trade-intensive industries. In excluding the financial crisis of 1997­98)--was not Korea, for example, wages were significantly less only less volatile than the 1980s, but in fact the least volatile in the later, supposedly more open period volatile of the last three decades. The same conclu- than in the earlier period, whereas industries with sion holds when considering only the manufactur- high trade exposure experienced less volatility in ing sector, which appears to be generally more the later period and roughly the same volatility as volatile than the overall economy. other sectors in the first period (Table 10.5). 176 East Asia Integrates TABLE 10.4 Volatility of Annual Growth Rates in GDP and Value Added in Manufacturing, Measured by Coefficient of Variation: Indonesia, Republic of Korea, and Thailand, 1970­96 Coefficient of variation of annual growth rates in GDP Manufacturing (value added) Indonesia 1970­1979 0.16 0.34 1980­1989 0.40 0.55 1990­1996 0.10 0.10 Korea 1970­1979 0.26 0.31 1980­1989 0.52 0.64 1990­1996 0.22 0.30 Thailand 1970­1979 0.39 0.35 1980­1989 0.43 0.75 1990­1996 0.19 0.26 Sources: Bank of Thailand, Bank of Korea, and Bank of Indonesia. TABLE 10.5 Level and Fluctuation of Wage Growth Rates (Males) by Trade Exposure Groups: Republic of Korea, 1976­1997 Low Medium High exposure exposure exposure Annual growth rate 1976­1987 0.022 0.020 0.018 1988­1997 0.064 0.066 0.076 Standard deviation of growth rates 1976­1987 0.077 0.071 0.068 1988­1997 0.038 0.054 0.038 Note: Low trade exposure industries: food and beverage, paper, petroleum, and rubber and plastics. Medium trade exposure industries: wood, chemical, and iron and steel. High trade exposure industries: textile and apparel, fabricated metal, machinery, and others not elsewhere classified. Source: Wage Structure Survey (formerly the Occupational Wage Survey). But the preceding aggregate evidence must be levels or average wage rates may actually hide an interpreted with caution. First, because volatility is increased turnover rate among employees and evaluated for only a small number of years in each higher variance in their earnings over time. subperiod, the results may be strongly influenced Any study of workers' vulnerability to shocks by single observations. In particular, the economic requires panel data that follow individual workers shocks of the 1980s may be responsible for the over time. Unfortunately, the only panel data avail- higher volatility observed during that period, able were for Korea, and for too short a period to be which therefore may not be fully comparable to the really useful.8 For the first stage of our approach, we 1990s, when the crisis that occurred late in that used synthetic panels, which were created by follow- decade is ignored. Second, although volatility is ing cohorts of randomly selected individual workers practically constant at the aggregate or even at the over time in successive cross-sectional surveys. industry level, it may have increased at the individ- Cohort cells in this synthetic panel were defined by ual level. The stability of aggregate employment workers' year of birth, gender, and educational Trade and Labor Market Vulnerability in Indonesia, Republic of Korea, and Thailand 177 attainment. Because the resulting samples are small degrees of exposure to trade. In a flexible labor for some countries, it was sometimes necessary to market where workers are mobile, a shock that use a more aggregate definition of cohorts.9 originates in the tradable sector will be transmitted For the second stage of the analysis, we recov- to other sectors. For Korea, where the time series ered information on the dynamics of individual are longer, there is no noticeable change in trend earnings from the observation of the time patterns when the economy experiences greater openness in of the mean and the variance of earnings in cohort trade. There is also no sign of greater variability in cells. We then used that information to simulate the growth rates. likely vulnerability of workers to poverty under a It is quite possible that the steady and parallel set of simplifying assumptions. trends in a cohort's mean wage hide a variability in individual earnings that increases over time or is very different across sectors. A simple way to check Trends in Earnings and this possibility is to examine whether the variance Employment in Synthetic Panels of individual earnings changes very much over In Korea and Indonesia, where longer time-series time or differs markedly across sectors. If so, such a data are available, average real wages have been ris- change or difference could correspond to some ing steadily since the early 1980s, except for a mod- increase in the variance of transitory earnings, est drop in Korea and a sharper decline in Indone- thereby reflecting increased instability at the indi- sia associated with the 1997­98 financial crisis. In vidual level.10 Thailand, average real wages stayed quite constant The variance of (log) wages within cohort cells between 1991 and 2001. in the three countries does not show any rising Figure 10.2 shows that for Korea (1976­2000) trend, and patterns are similar across the tradable this evolution is the same for a specific cohort and and nontradable sectors as well as across the indus- that trends do not differ between broad sectors or try groups with varying degrees of trade exposure. across industries of varying trade intensity. The An example is given in Figure 10.5, which shows for same consistency is observed for other cohorts and Korea the evolution of the variance of (log) earn- for Thailand (1991­2001) in Figure 10.3 and ings in a male cohort (born in 1950, attained a sec- Indonesia (1986­1999) in Figure 10.4. ondary education) between 1976 and 2000. This similarity suggests that labor markets in the This result in Figure 10.5 appears to contradict three countries are quite integrated, with no evi- the finding in the pioneering work by Deaton and dence of segmentation between tradable and non- Paxson (1994). In an analogous cohort analysis in tradable sectors or among industries with different the United States, the United Kingdom, and Taiwan FIGURE 10.2 Average Log of Real Wages of Males by Broad Sectors (Manufacturing, Services and Construction, and Others) and by Trade Exposure (Low, Medium, and High): Republic of Korea, 1976­2000 Males, 38 years old in 1990, primary education 14.5 14.5 14.0 14.0 13.5 13.5 13.0 13.0 12.5 12.5 1980 1985 1990 1995 2000 1980 1985 1990 1995 2000 Manufacturing Services and Construction Low Medium High Exposure Others Source: Wage Structure Survey (formerly the Occupational Wage Survey). 178 East Asia Integrates FIGURE 10.3 Average Log of Real Wages of Males by Broad Sectors (Manufacturing and Services) and by Trade Exposure (Low, Medium, and High): Thailand, 1991­2000 Males, 35­39 years old in 1991, primary education Average log of real wages Average log of real wages 10.0 10.0 9.5 9.5 9.0 9.0 8.5 8.5 8.0 8.0 7.5 7.5 7.0 7.0 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 Manufacturing Services Low Medium High Source: Thailand Labor Force Surveys. FIGURE 10.4 Average Log of Real Wages of Males by Broad Sectors (Manufacturing and Services): Indonesia, 1986­1999 Males, 35­39 years old in 1986, primary or less education Average log of real wages 7.8 6.8 5.8 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 Manufacturing Services Source: Indonesia Labor Force Surveys (SAKERNAS). (China), these authors found that the variance of the variance of (log) earnings was extremely stable the log of household consumption, income, and in Taiwan between ages 20 and 50 (and then dou- earnings tended to increase with the age of the bled between ages 50 and 65)--which is in full con- household head. formity with our results. Second, our results for The discrepancy between the present result for Korea refer to active wage workers, whereas Deaton Korea and their results for Taiwan, two very similar and Paxson's data for Taiwan include changes in economies, might be more apparent than real, participation and in wage work status. Finally, it is however. First, Deaton and Paxson also found that also possible that a slightly increasing trend in Trade and Labor Market Vulnerability in Indonesia, Republic of Korea, and Thailand 179 FIGURE 10.5 Standard Deviation of Log of Real Wages of a Male Cohort in Low, Medium, and High Trade Exposure Industries: Republic of Korea, 1976­2000 Males, 40 years old in 1990, secondary education Standard deviation .8 .6 .4 .2 0 1980 1985 1990 1995 2000 Low Exposure Medium Exposure High Exposure Source: Wage Structure Survey (formerly, Occupational Wage Survey). inequality might be observed in the Korean data if and this increase might be taken as hiding a slight cohorts with different educational levels were taken increase in disguised unemployment, possibly together, as in Deaton and Paxson. imputable to the rising openness of the economy. Changes in labor market conditions possibly However, the change in the inactivity rate seems associated with more openness might have affected too modest to be of real concern. employment or working hours rather than earning The evolution of average working hours within rates. But Figures 10.6 and 10.7 suggest this is not the cohorts leads to the same conclusion: the absence case. In general, the distribution of employment sta- of an openness effect. Because of data limitations, it tus and of sector of employment within cohort cell was not possible to compare the 1980s and the shows no noticeable change in trend in the 1990s in 1990s in Thailand. For a cohort of educated men, comparison with the 1980s, except, of course, during no significant trend appears to be present in the the 1997­98 financial crisis in East Asia. 1990s until the 1997­98 financial shock (Figure For example, for a cohort of Korean males born 10.7). The evolution was similar in other cohorts. from 1950 to 1954 who attained a primary educa- In most of them, it is true that working hours tend tion or less, the rising trend in the share of the non- to be longer in trade-exposed manufacturing than tradable sector showed no discontinuity until 1998, in trade-protected services. But this variation seems and the same was true of the declining trend in to be true in many countries, and it is more sugges- agriculture and manufacturing (Figure 10.6). These tive of differences in the conditions of production trends may therefore reflect the long-run process of than of differences in trade exposure. Within man- change in the economic structure rather than the ufacturing, by contrast, work hours do not differ effect of a change in the trade regime. Likewise, the according to the trade intensity of industries. unemployment rate remained remarkably constant In summary, looking at the evolution of labor until the 1997­98 crisis. The inactivity rate market indicators for synthetic cohorts, we found increased somewhat a few years before the crisis, no evidence of a systematic difference associated 180 East Asia Integrates FIGURE 10.6 Distribution of Employment Status and Employment Sectors of a Male Cohort: Republic of Korea, 1985­2000 Males, born in 1950­54, primary education or less Percent 60 50 40 30 20 10 0 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 Unemployed Inactive Agriculture Manufacturing Nontradable Source: Surveys of the Economically Active Population, Korea. FIGURE 10.7 Average Monthly Hours Worked of a Male Cohort in Manufacturing and Services: Thailand, 1991­2000 Males, born in 1951­55, secondary education or more Average hours worked per week 60 55 50 45 40 35 30 1991 1992 1993 1994 1996 1997 1998 1999 2000 2001 Manufacturing Services Source: Thailand Labor Force Surveys. with the degree of exposure to trade in any of the to trade exposure. And there was no evidence that three countries studied. No systematic difference trends in earnings or employment changed, or that in workers' earnings or employment was found working hours became more unstable as between the tradable and nontradable sectors, or economies liberalized their trade policies during among manufacturing industry groups according the 1990s. Trade and Labor Market Vulnerability in Indonesia, Republic of Korea, and Thailand 181 Indirect Estimation of Changes in A Simple Model Workers' Vulnerability to Poverty from Cross-Sectional Data The earnings, wit, at time t of individual i belonging to cohort group j may be represented by A steady aggregate evolution of earnings and j j j j employment levels for the whole population or for (10.1) lnwit = Xitt + it cohorts of individuals may hide changes in the vari- ability of conditions for individual workers. The where Xit is a set of characteristics that are not used share of employment in manufacturing or services in the definition of cohort group j--that is, they are may change steadily over time, but, meanwhile, an not from age, gender, and education. In addition, it increasing proportion of people may be switching is assumed that the unobserved residual term jit among sectors or switching between employment follows a first-order autoregressive process and inactivity within a cohort. Likewise, average AR(1)--that is earnings may go up at a steady pace while individual variability in earnings is increasing.11 (10.2) it = jit j j -1+ it . j Any attempt to determine whether there are changes in the variability of individual conditions This is the simplest time dependency assump- in a labor market, and therefore in personal vulner- tion that can be made.13 ability to poverty, would ideally require panel data In the absence of panel data, the dynamic equa- that would allow a sample of people to be followed tion cannot be estimated directly on individual- over time. Unfortunately, such data are seldom level data. However, some indirect estimation is available in developing countries, and Thailand and possible by considering successive observations of Indonesia are no exception.12 individuals in the same cohort, even though those To estimate the individual variability of earnings individuals are not the same from one period to the and the vulnerability of workers to having their next. Indeed, (10.2) implies that earnings fall below a poverty threshold, we used an indirect method based on comparing successive (10.3) 2jt = j 2jt 2 -1 + 2jt cross sections. The idea is as follows: if it can be assumed that all individuals within a cohort face a where 2jt is the variance of the residual term of the stochastic earning process that has common charac- earning equation at time t, and 2j is the variance of teristics, then these characteristics can be recovered the innovation term during the same period in the at the aggregate level without observing actual earn- dynamic equation (10.2). Both variances are evalu- ing paths. Observing the evolution of the mean and ated for the whole cohort j of individuals observed the variance of earnings within a cohort is sufficient at time t. The sequence of 2jt is a time series. Esti- for estimating the common characteristics of indi- mating a standard autoregressive model on this vidual earning processes. On this basis, simple esti- series yields estimates of j and 2jt. Of course, a mates of the probability of a worker observed in year crucial identifying assumption here is that the t falling into poverty in year t + 1 can be obtained. regression coefficient j is constant over time. Now returning to the issue of trade exposure, If the model is well specified and the time series the problem is one of knowing whether the insecu- is long enough, then the estimated j and 2jt ^ ^ rity and uncertainty evaluated with the preceding should have the expected signs and orders of mag- technique changed between less liberal and more nitude. In particular, it should be the case that 0 < liberal time periods, as well as between industries of ^j < 1 and 2jt > 0. However, if well-behaved esti- ^ varying exposure to trade. This comparative esti- mates of j and 2jt (for every t) are not forthcom- ^ ^ mation of vulnerability to poverty was made for ing, then it may be necessary to use alternative esti- each of the three countries. For all countries, a mates of j lying in the confidence interval poverty threshold of 60 percent of the national obtained in the original estimation.14 However, this median wage was used. Findings for the three step seldom proved necessary. countries were very similar, so only a typical selec- With estimates of j and 2jt, it is now possible to tion is presented here. simulate the dynamics of individual earnings, 182 East Asia Integrates according to equations 10.1 and 10.2. One begins the comparison of the evolution of the mean and by drawing a value ^ itj + 1 in the normal distribution the variance of earnings within the typical cohort with mean 0 and variance 2it + 1. Substituting the with vulnerability as proxied by the preceding tech- draw, ^ itj + 1, into equation 10.2, and combining it nique should be informative. with estimated from equation 10.3 and predicted ^ j The definition and estimation of vulnerability ^ it from ordinary least-squares in equation 10.1, j given here do not take into account employment yields an estimate ^ itj + 1. Substituting this value in mobility. Vulnerability is estimated for those indi- (10.1) at time t + 1 then gives the (log) earnings for viduals in a cohort who are employed in each an individual in cohort j at time t + 1, conditional period, and thus ignores the part of vulnerability on the earnings at time t and assuming exogenous that is associated with losing one's job. As revealed changes in characteristics. earlier, net flows into unemployment did not seem Using the preceding reasoning, one can see that to have increased because of the change in trade the probability for individual i, observed at time t, regime in the three countries under analysis. But of receiving e earnings below a survival threshold w gross flows may have become bigger. Not enough at time t + 1, conditional on characteristics and information is available to verify this point. earnings in period t, is given by Also, in comparing the vulnerability of workers in sectors with different degrees of exposure to (10.4) trade, it is tempting to apply the preceding tech- nique in equation 10.4 to cohorts of individuals v^it j +1= Pr(wit j +1< w | Xit ,Xit j j j +1,wit+1 ) employed in the same sector. But then the same pro- j j j j j j viso applies. The resulting proxy for vulnerability is = log(w)- Xit ^ +1 t+1 -^ log(wit )- Xit^t [ ] valid insofar as intersectoral movements did not 2jt+1 change substantially during the period under study. Again, the analysis just described suggests that they where (.) denotes the cumulative density of the did not, in net terms, but this finding is not neces- standard normal. But this expression requires sarily inconsistent with increased vulnerability to determining the characteristics of individual i at poverty through forced mobility. Under these con- time t + 1. Indeed, X jit is observed, but X jit+ 1is not. ditions, our findings must be interpreted cautiously. Without a special reason to do otherwise, the sim- plest approach is to assume the same characteristics Findings in time t and t + 1 except, of course, for age. Such an assumption should be satisfactory if the character- Reasonable estimates of the parameters of the istics in X are truly exogenous. model were obtained for all three countries. In par- According to equation 10.4, vulnerability--that ticular, estimated persistence coefficients, , in is, the probability of being below the earning equation 10.2 range from zero to one, and generally poverty threshold in year t + 1, conditional on significantly so, suggesting that earning shocks in earnings in year t--depends on the following the current period are actually transmitted to con- parameters: initial earnings, individual characteris- secutive periods, but that the effect of the shocks tics, changes in the returns to these characteristics, will eventually fade out.15 For example, Table 10.6 the persistence of earning shocks from one period presents the estimates of 2 for Indonesia (note that to the next, and, finally, the variance of these cohorts are pooled to estimate 2). The estimated 2 shocks. As noted earlier, differences in trade ranges from 0.06 to 0.45. All of these 2 are statisti- regimes did not seem to be associated with differ- cally significantly greater than zero--the smallest t- ences in overall earning inequality. This finding statistic being 3.5--and less than one. An exception suggests either that some of the previous parame- is female workers with secondary education work- ters, or their distribution within cohorts, remained ing in the nontradable sector for whom 2 is not sig- more or less constant over time, or that compensat- nificantly different from zero because of the limited ing variations have taken place. In what follows, no number of observations within these cohorts. Fig- attempt is made to isolate the effect of each set of ure 10.8 presents the estimated for every birth parameters on the evolution of vulnerability. Yet cohort of workers in Korea. The estimates of j Trade and Labor Market Vulnerability in Indonesia, Republic of Korea, and Thailand 183 TABLE 10.6 Estimated 2 Correlation of Variance of Residual Earnings by Education, Gender, and Sector: Indonesia Manufacturing Nontradable Education Male Female Male Female Primary or less 0.3548 0.3746 0.4583 0.4098 Secondary 0.3547 0.3549 0.1841 0.0617 Tertiary 0.4124 -- 0.2165 -- Notes: For female workers, tertiary education falls within secondary education because there are not enough observations for those with tertiary education to treat them separately. Source: Alatas (2002). FIGURE 10.8 Estimated by Birth Year Cohort: Republic of Korea, 1929­69 1 0.9 0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0 1929 1931 1933 1935 1937 1939 1941 1943 1945 1947 1949 1951 1953 1955 1957 1959 1961 1963 1965 1967 1969 Men Women Source: Wage Structure Survey (formerly the Occupational Wage Survey). range from 0.3 to 0.9 in most cohorts and the aver- the extent of vulnerability was extremely limited-- age is 0.74 for men and 0.59 for women.16 with an overall average of 3 percent--reflecting the In Korea, a worker's risk of falling into poverty very low inequality of individual earnings in Korea. differed rather little among the three sectors, and Overall growth in Korea and in the other two these differences narrowed over time. Differences countries between the early 1980s and 1998 was among manufacturing industries with different such that the strongly decreasing trend in absolute degrees of trade exposure narrowed, too, as vulner- poverty hides any variation that might be the result ability decreased. Remarkably, the drop in vulnera- of more trade exposure. bility was the most pronounced for the manufac- In all three countries, not surprisingly, vulnera- turing sector, which has the highest exposure to bility tended to differ widely according to gender trade (Figure 10.9). In the Korean manufacturing and educational attainment. As reflected in Figure sector, vulnerability to relative poverty decreased 10.10 for Thailand and Table 10.7 for Indonesia, the until 1998 and increased again somewhat after- expected earnings of women and people with little ward. Vulnerability may have increased slightly in education were much below national averages. The services and construction a little before the vulnerability of less educated workers to poverty 1997­98 financial crisis, but it remained stable in may be considerable, reaching 20 percent for male the other sectors. It must be stressed, however, that workers in Indonesia. 184 East Asia Integrates FIGURE 10.9 Vulnerability by Sectors and Trade Exposure Groups: Republic of Korea, 1976­2000 Conditional likelihood of earning less than poverty threshold 0.07 0.07 0.06 0.06 0.05 0.05 0.04 0.04 0.03 0.03 0.02 0.02 0.01 0.01 0 0 1979 1981 1983 1985 1987 1989 1991 1993 1995 1997 1999 1979 1981 1983 1985 1987 1989 1991 1993 1995 1997 1999 Manufacturing Low exposure Service and construction Medium exposure Others High exposure Note: The figure shows the evolution of Korean workers' vulnerability as defined in year t by their conditional likelihood of earning less than the poverty threshold in year t + 1 by sector of employment and, within manufacturing, by indus- tries with different degrees of exposure to trade. The evolution is shown for the 21-year period 1979­2000 and for the average across all cohorts in each year of observation. Source: Wage Structure Survey (formerly the Occupational Wage Survey). TABLE 10.7 Vulnerability by Education, Gender, and Sector: Indonesia, 2001 (Conditional likelihood of earnings falling beneath poverty level) Male Female Education Manufacturing Nontradable Manufacturing Nontradable Primary or less 0.1971 0.2147 0.3098 0.4076 Secondary 0.0890 0.1047 0.1552 0.1777 Tertiary 0.0050 0.0261 -- -- Note: For female workers, tertiary education falls within secondary education because there are not enough observations for those with tertiary education to treat them separately. Source: Alatas (2002). In summary, we found no firm evidence to sup- case, the differences in earnings security and vul- port a correlation between workers' vulnerability to nerability associated at any point in time with edu- sinking into poverty and periods of greater eco- cation or gender seem much larger than what could nomic openness, or between their vulnerability and be imputed to trade openness. To reduce the overall their sectors of employment. This conclusion is, vulnerability of workers to poverty, it might be however, conditional on the limitations of the indi- more efficient to tackle these disparities first. rect method used in this analysis to evaluate vul- nerability. The analysis could have been strength- Conclusions ened had we been able to take into account gross flows in and out of employment or across sectors of Determining whether trade liberalization is associ- employment. Yet evidence on net flows does not ated with greater earnings volatility or increased suggest significant differences across industries vulnerability of workers requires comprehensive with different degrees of trade exposure. In any panel data on individual employment status and Trade and Labor Market Vulnerability in Indonesia, Republic of Korea, and Thailand 185 FIGURE 10.10 Vulnerability by Gender and Educational Attainment: Thailand, 1991­2000 Males by educational attainment Females by educational attainment Conditional likelihood of earning less than poverty threshold 0.25 0.25 0.2 0.2 0.15 0.15 0.1 0.1 0.05 0.05 0 0 1992 1993 1994 1995 1996 1997 1998 1999 2000 1992 1993 1994 1995 1996 1997 1998 1999 2000 Primary Primary Secondary Secondary Tertiary Tertiary Source: Thailand Labor Force Surveys. earnings. In the absence of such data, only indirect Now it is time to make a point that will reinforce evidence or indirect proxies for individual earning the general findings in this chapter and at the same volatility and vulnerability may be used. This chap- time suggest another, more subtle way through ter has reviewed such evidence for three East and which openness to trade may actually affect earn- Southeast Asian countries where trade barriers ings volatility. The apparent stability of net employ- were lowered significantly during the 1990s. ment flows and the evolution of earnings and their Indirect evidence revealed no significant change variance in the decade or so preceding the 1997­98 in year-to-year fluctuations in workers' earnings and East Asian financial crisis are meaningful precisely employment after trade was further liberalized in the because the crisis corresponded to quite significant 1990s or in sectors that were more exposed to for- changes in practically all existing trends. Thus the eign competition. The mean and variance of earn- crisis shows that there may be some variability in all ings, and net flows in employment, proved to be very the indicators used in this chapter. The remarkable similar in tradable and nontradable sectors. We did result is therefore that no systematic variation not find any systematic differences in employment showed up at the time the three economies were or earnings volatility. Nor did workers' vulnerability, becoming more open, or in the comparison across as approximated using an original methodology sectors with distinct exposures to trade. This find- developed in this chapter, seem to differ significantly ing seems indeed to confirm that opening up to across manufacturing industries with different trade in the late 1980s and early 1990s did not have degrees of exposure to trade. Our results, in fact, strong negative effects on poverty and vulnerabil- show that vulnerability to relative poverty--defined ity, whereas the 1997­98 crisis did have such effects. as 60 percent of the national median wage--seems In turn, one may wonder whether it was openness to have declined over the last 10­20 years in the three that made that crisis much more serious than the countries and that differences across sectors seem to various shocks that hit the three economies in the have narrowed considerably. 1980s. If this were true, then it could be asserted These conclusions may be criticized because that openness actually contributed to more indi- they rely on indirect evidence. In particular, it may vidual vulnerability in East and Southeast Asia be that the stability of net employment flows and through increased macroeconomic volatility, rather the steady evolution of the mean and the variance than through the interplay of modified microeco- of earnings behind the proxy used for vulnerability nomic mechanisms. The question is open. actually hide an increased instability for individual Although it is not unreasonable to suggest that workers in the labor market. Again, to reach a defi- short-term idiosyncratic shocks in the global mar- nite conclusion would require panel data. ket may affect the more exposed sectors dispropor- 186 East Asia Integrates tionately, the fact that we did not find such evi- activities in the income of rural self-employed. In Korea, we dence suggests that a possible reason is that, in the worked primarily with the establishment surveys, because the labor force surveys do not contain wage information absence of systemic shocks like the one in 1997­98, and the establishment surveys do not include the primary an open economy is subject to many uncorrelated sector. shocks that tend to cancel one another out thanks 4. For Korea, we were able to break up services into the strictly nontradable construction, social, and communal services, to a well-functioning labor market--certainly a and nontradable industries that may depend on trade characteristic of the three economies under study. activity and are categorized as "others." Another possibility is that liberalization in small 5. Services, retail and wholesale, transport, and communica- tions. economies may facilitate access to the interna- 6. We also examined the ranking by trade per worker, which tional market, which provides hedging and self- gives very similar groupings. In Korea, the industries in the insuring opportunities. In addition, these Asian group of highest trade exposure are textile and apparel, fab- economies were already relatively open by the ricated metal, machinery, and others not elsewhere classi- fied; those in the group of medium exposure are wood, 1970s, and have been exposed to world markets chemical, and iron and steel; and those in the group of low and their fluctuations since then. Thus the tariff exposure are food and beverage, paper, petroleum, and reductions of the 1980s and 1990s may have had rubber and plastic. In Thailand, the highest trade exposure group consists of the chemical, primary metal, machinery, only negligible effects on domestic industries. As is and other industries not elsewhere classified; the medium well known, a country's tariff regime is a complex exposure group consists of the food and beverage, tobacco, combination of step function, multiple tiers, and textile, apparel, footwear, wood product, furniture, and fab- ricated metals industries; and the low exposure group is various types of percentage-quantity barriers. made up of the paper product, petroleum, and rubber and Under these conditions, it is extremely difficult to plastics industries. quantify the magnitude of additional exposure The Indonesian data do not permit a comparable break- that actually follows from an observed reduction down. Before 1988, SAKERNAS, the Indonesian labor force surveys from which our data are taken, provided only a in the average tariff rates. rough five-way classification of the sectors in which work- Although we find no obvious link between trade ers were employed: agriculture, manufacturing, trade, ser- and vulnerability, the analysis in this chapter con- vices, and other industries. Between 1989 and 2000 the sec- toral classification was expanded from 5 categories to 18, firms that some workers are more vulnerable than but within manufacturing it provided for only four indus- others. Women workers are more vulnerable to try groups: food, beverages, and tobacco; apparel and tex- falling into poverty, as are workers who have less tiles; wood products; and others. 7. Standard deviation of growth divided by the decade-aver- schooling. This result reflects the dominant evi- age growth. dence in the literature that gender and skill have a 8. Daewoo Economic Research Institute carried out the Korea strong discriminatory power in determining a per- Household Panel Survey from 1994 to 1998. The survey col- son's earnings. An important conclusion is that lected information on income, assets, expenditures, the labor market, and other household and individual charac- remedies to vulnerability to poverty must be teristics. There were no replacements of households, but the sought more in education than in trade openness. data covered split-off households stemming from marriage A more open trade regime is unlikely to justify by or other reasons. The survey was conducted through strati- fied random sampling by street blocks, and it covered all itself the creation of effective safety nets. Such Korean prefectures except Jeju-do. In 1994 there were about insurance systems may be justified independently, 3,500 households and in 1998 about 2,200 households. either because of the presence of a substantial pro- 9. The synthetic panels could be assembled for only a limited time period (except for Korea where establishment surveys portion of unskilled workers with limited capacity are available for as early as 1976, and labor force surveys for to face adverse shocks or possibly because of an as early as 1985). Because of this limitation, it was not pos- increased likelihood of a major macroeconomic or sible to compare systematically the 1980s and the 1990s-- systemic crisis. that is, moderate versus pronounced openness to trade-- for all three countries. The comparison performed on sectors with high and low exposure to trade did not present this difficulty. Endnotes 10. It is also known that the inequality of earnings tends to 1. For a critical view, see Rodriguez and Rodrik (1999). increase with age because the stochastic process behind 2. Detailed and comprehensive findings for Indonesia, Korea, individual earnings is close to a random walk, a substantial and Thailand are available from the authors upon request. proportion of shocks being persistent. 3. The sample of wage workers in the primary sector in the 11. Such an evolution is even consistent with a constant vari- labor force surveys of Thailand and Indonesia is small, and ance of relative earnings. It is sufficient that individuals it is not possible to isolate precisely agriculture from other within a cohort switch rank more and more frequently. Trade and Labor Market Vulnerability in Indonesia, Republic of Korea, and Thailand 187 12. In Korea, there is the household panel survey conducted by Arango, Carlos, and William Maloney. 2002. "Unemployment Daewoo Economic Research Institute, but the five-year Dynamics in Latin America: Estimates of Continuous Time period covered was too short for our analysis. Markov Models for Mexico and Argentina." World Bank, 13. A more general specification would also include a persist- Washington, D.C. Processed. ent component in the innovation term xjt. Some implica- Cox, Edwards A., and Sebastian Edwards. 1996. "Trade Liberal- tions of this specification are considered in Deaton and ization and Unemployment: Policy Issues and Evidence Paxson (1994). from Chile." Cuademos de Economia 33 (99): 227­50. 14. Let us rewrite equation 10.3 as 2jt = aj jt 2 ­ 1+ bj + ujt, Currie, Janet, and Ann E. Harrison. 1997. "Sharing the Costs: where 2jt = Bj + ujt. Let the ordinary least-squares (OLS) The Impact of Trade Reform on Capital and Labor in estimates of the two coefficients aj and bj be A and B. Two Morocco." Journal of Labor Economics 15 (13) (part 2, July): bad cases are possible. Scenario 1: suppose that A is positive S44­S71. but some Bj + ut are negative. We can then try to find a j Deaton, Angus, and Christina Paxson. 1994. "Intertemporal value 2 < A in the confidence interval of A, such that all Bj Choice and Inequality." Journal of Political Economy 102: + ujt are positive. Scenario 2: suppose that OLS gives a neg- 437­67. ative A or a value that exceeds one. In the first case, it is best De Ferranti, David, Guillermo E. Perry, Indermit S. Gill, and Luis to start with small positive values of 2 in the confidence Servén Region. 2000. Securing Our Future in a Global Econ- interval of A and proceed as in scenario 1. In the second omy. Washington, D.C.: World Bank. case, we would start with values below but close to 2 = 1 Fajnzylber, Pablo, and W. F. Maloney. 2000."Labor Demand and and proceed as in Scenario 1. Practically, A was never found Trade Liberalization in Latin America." Latin America and significantly smaller than zero or greater than one with Caribbean Sector Management Unit for Poverty Reduction standard tests--in theory, nonstationarity tests of the and Economic Management. World Bank, Washington, Dickey-Fuller type should be applied to test A = 1. D.C. 15. This finding is in apparent contradiction with the finding Frankel, Jeffrey, and David Romer. 1999. "Does Trade Cause by Deaton and Paxson (1994), who applied a similar Growth?" American Economic Review 89 (3): 1­16. methodology to Taiwan (China). They found that shocks Irvin, D., and M. Tervio. 2002. "Does Trade Raise Income? Evi- on household consumption expenditures--and implicitly dence from the Twentieth Century." Journal of International income--tend to be persistent. Thus there is an increasing Economics 58: 1­18. trend in the variance of log expenditures. The reason for Michalopoulos, Constantine. 1999. Trade Policy and Market the difference with the findings reported in this chapter Access Issues for Developing Countries. Policy Research Work- might be that we focus on individual earnings and wage ing Paper 2214. World Bank, Washington, D.C., October. earners rather than household consumption per capita. We Papageorgiou, Demetrios, Armeane M. Choksi, and Michael therefore ignore shocks linked to the demographic compo- Michaely. 1990. Liberalizing Foreign Trade in Developing sition of the household and labor supply. Countries: The Lessons of Experience. Washington, D.C.: 16. Note that is highest among the youngest cohorts for both World Bank. men and women, because earning changes persistently take Revenga, Ana. 1994. "Employment and Wage Effects of Trade place early in a worker's career. Liberalization: The Case of Mexican Manufacturing." Policy Research Working Paper 1524. World Bank, Washington, D.C. References Rodriguez, Francisco, and Dani Rodrik. 1999. "Trade Policy and Economic Growth: A Skeptic's Guide to the Cross-National The word processed describes informally reproduced works that Evidence." NBER Working Paper No. 7081. National Bureau may not be commonly available through libraries. of Economic Research, Cambridge, Mass. Sachs, Jeffrey, and Andrew Warner. 1995. "Economic Reform Alatas, Vivi. 2002. "Labor Market Vulnerability in Indonesia: A and the Process of Global Integration." Brookings Papers on Synthetic Cohort Panel Simulation Exercise." World Bank, Economic Activity No. 1. Brookings Institution, Washing- Washington, D.C. Processed. ton, D.C. AUTHORS AND THEIR AFFILIATIONS Kathie Krumm Robin Carruthers The World Bank The World Bank Homi Kharas Jitendra N. Bajpai The World Bank The World Bank William J. Martin David Hummels The World Bank Purdue University, United States Deepak Bhattasali Manjula Luthria The World Bank The World Bank Shantong Li Keith E. Maskus Development Research Centre of the State Council, University of Colorado at Boulder, China United States Mari Pangestu Shaohua Chen Center for Strategic and International Studies, The World Bank Indonesia Sudarshan Gooptu Martin Ravallion The World Bank The World Bank Bijit Bora Jehan Arulpragasam World Trade Organization The World Bank 189 190 East Asia Integrates Francesco Goletti François Bourguignon Agrifood Consulting, United States DELTA, France Tamar Manuelyan Atinc Chor-ching Goh The World Bank The World Bank Vera Songwe The World Bank INDEX A AMF (Asian Monetary Fund), 43 AFTA (ASEAN Free Trade Area) antidumping provisions and China's accession to WTO, extent of regional interdependence, xviii xv, 5­6, 17 market access barriers, poverty, and agricultural APEC. see Asia-Pacific Economic Cooperation (APEC) products, 63, 68, 72 apparel. see textiles RTAs, 41­42, 45, 52 Argentina age customs clearance, 83 household-level effects of China's accession to WTO, dumping, 6 137, 138, 139 environmental standards, 127 vulnerability of trade and labor market in Indonesia, labor standards and employment, 119 Republic of Korea, and Thailand, 177, 178 logistics, 75 Agreement on Textiles and Clothing (ATC), Uruguay vulnerability of trade and labor market, 166 Round, xxi Arulpagasam, Jehan, 149­169 Agreement on Trade-Related Aspects of Intellectual ASEAN. see Association of South East Asia Nations Property Rights (TRIPS), xxvii, 91 (ASEAN) China's accession to WTO, 6­7 Asia-Pacific Economic Cooperation (APEC) copyright in Indonesia, 102 environmental standards, 125­126 patents in Korea, 99, 108 extent of regional interdependence, xviii traditional knowledge protections, 103, 105, 107 logistics, xxiii, xxiv, 74, 87 agriculture RTAs, 37, 39, 43, 45 cashmere (see cashmere in Mongolia) Asian Monetary Fund (AMF), 43 China's accession to WTO, xiv­xv, 7­9, 18, 23, Association of South East Asia Nations (ASEAN) 139­140 AFTA [see AFTA (ASEAN Free Trade Area)] market access issues (see market access barriers, China and (see China) poverty, and agricultural products) logistics, 80, 87 rice market access barriers, poverty, and agricultural Cambodia and Vietnam (see rice in Cambodia products, 61, 63, 70, 72 and Vietnam) MRAs, xxii global trade in, 63, 65 policy recommendations, xxxiv significant source of income for low and middle- RTAs, xviii­xix, 37, 39, 51­52 income countries, 59­60 ASEAN-China Free Trade Area, 37, 41, 42, 47­52 technical requirements and standards, xxii current and proposed agreements, 37, 39, 41­43 traditional knowledge protections, 102, 103 financial cooperation on regional basis, 43­44 191 192 East Asia Integrates trade trends and opportunities, xiii­xiv value chain analysis, xxxii, 149­152, 163­166, ATC (Agreement on Textiles and Clothing), Uruguay 167­168 Round, xxi CBD (Convention on Biodiversity), 104­105 Atinc, Tamar Manuelyan, 149­169 CER (Australia-New Zealand Closer Economic Rela- Australia, 7, 37, 40, 41, 45 tions) and East Asia RTAs, 37, 41, 45 authors' rights. see copyright in Indonesia CGE (comparable general equilibrium) modeling, 133 charts. see tables, charts, and figures B chemicals patents in Korea, 97 Bajpal, Jitendra N., 73­89 Chen, Shaohua, 131­142 Basel Convention, 129 Chiang Mai initiative (CMI), x, 43 "behind the border" reform issues, xxi­xxii child labor, 117 Bhattasali, Deepak, 3­20 Chile, 40, 125, 127, 166 bilateral investment treaties, East Asian countries, China. see also Hong Kong (China); Taiwan (China) xxvi­xxvii accession to WTO (see China's accession to WTO) biological research and traditional knowledge protec- ASEAN and China tions. see traditional knowledge protections ASEAN-China Free Trade Area, 37, 41, 49­51 Bolivia, 166 China-ASEAN framework agreement, xviii books and copyright in Indonesia, 102 "Early Harvest" provisions, China-ASEAN agree- Bora, Bijit, 69­73 ment, xix, 28, 33, 49, 50 Bourguignon, François, 165­181 RTAs, xx, 37, 41, 42, 47­52 box features. see tables, charts, and figures WTO, China's accession to, 22, 28, 33, 47­48 Brazil bilateral investment treaties, xxvii customs clearance, 83 customs clearance, 83 environmental standards, 126, 127 economic/income gap, ix, xxx intellectual property rights, xxvii environmental standards, 123, 125, 126, 127 labor standards and employment, 119 intellectual property rights, xxvii, 108 logistics, 75 labor standards and employment, 117, 119 vulnerability of trade and labor market, 166 logistics, xxiii, xxiv, 12, 74­78, 80­86 Brenton, Paul, 69­73 market access barriers, poverty, and agricultural Brunei, 66­67 products, 60, 63, 66­67, 69, 70 regional role of, xiv­xvii C regulatory and pro-competitive reforms, xxvi Cambodia RTAs China's accession to WTO, xvii, 33­34 ASEAN and China, 37, 41, 42, 47­52 economic/income gap, ix, xxix current and proposed, 40, 41 labor standards and employment, 117 financial cooperation on regional level, 43­44 logistics, 75, 76, 77­78, 79, 83, 86 reasons for new resurgence of, 38 market access barriers, poverty, and agricultural services liberalization, xxv products, 61, 63, 69, 70 technical requirements and standards, xxii MFN status with China, xix textiles rice (see rice in Cambodia and Vietnam) easing adjustments in industry, xxxv textiles, xxi global integration, xx­xxi trade trends and opportunities, xi, xiv WTO, accession to, xv, 4­5, 11, 14­15, 26, 33, 48 Canada, 11, 24, 40, 66­69, 72 third country markets, China's trade in, xiv Carruthers, Robin, 73­89 trade trends and opportunities, xi­xiv cashmere in Mongolia, xxxi­xxxii, 149­151, 161­168. see China's accession to WTO, x, xiv­xvii. see also tables, also tables, charts, and figures charts, and figures exports, 161­163, 166 agriculture, xiv­xv, 7­9, 18, 23, 139­140 herders, 163, 164­165 ASEAN and China, 22, 28, 33, 47­48 household level effects of cashmere trade, 163, 167 educational opportunities, expansion of, 17 importance of cashmere to economy and society, employment, 15, 18 161­163 exports, 14­15, 18, 23­24 policy and policy reform, 151­152, 163­164, FDI, 24­26 166­167 financial issues processors, 165­166 economic reforms, 13­15 quality issues, 163 financial services, 13 share of cashmere products in total cashmere investment patterns, 24­26, 27 exports, 166 household-level impact (see also tables, charts, and traders, marketers, and distributors, 165 figures) Index 193 agricultural households, 139­140 corporate cartels, xxvii characteristics of households affecting, 136, cost/benefit analysis for RTAs, 44­47 137­140 Council for Mutual Economic Assistance (COMECON), geographic variations, 131, 136, 140 collapse of, 162­163 methodology of analysis, 133­134 Croatia, 41 policy and policy reforms, 131, 132, 140 cross-border facilitation of logistics, xxv, 86­87 poverty and income, 16­17, 131, 132, 134­140 customs clearance, xxv, 82, 83 results, 134­140 Czech Republic, 41 rural vs. urban households, 15­16, 131, 136­140 welfare impact of accession, 15, 131­132, 134­137 D imports, 15, 18, 22­23 design and traditional knowledge protections, 102 investment patterns, 24­26, 27 diversification of products and logistics, 80 Japan and (see Japan) Doha Round, Declaration, and Development Agenda, manufacturing, xv, 9­11, 17, 23 xx, xxii, 9, 45, 116, 129­130 motor vehicles, xv, 10­11, 26­27 drugs, 97, 102 national impact, 3­20 dumping/antidumping provisions and China's accession economic reforms, 13­15 to WTO, xv, 5­6, 17 household-level effects (see subhead "household- level impact," this entry) E policy reforms, xv, 4­7, 11 e-commerce and logistics, 82, 84 sectoral changes, xiv­xv, 7­13 "Early Harvest" provisions, China-ASEAN agreement, ways of doing business, changes in, xvi xix, 28, 33, 49, 50 poverty and income effects, 16­17, 131, 132, 134­140 earnings stability. see vulnerability of trade and labor regional impact, xvi­xvii, 21­38 market in Indonesia, Republic of Korea, and channels of, 22­26 Thailand low-income countries, 33­36 East Asia currency and financial crisis, 1997-1998, x, middle-income developing countries, 27­33 xxxii, 123­124, 179 NIEs, 26­27 East Asia integration, ix­xxxviii. see also more specific services, xvi, 10­13, 23 topics social protection measures, 17, 18 "behind the border" reform issues, xxi­xxix technical requirements and standards, xxii benefits, scope of, xviii­xix textiles, xv, 4­5, 11, 14­15, 26, 33, 48 China third-country market competition, 24 accession to WTO, regional impact of (see China's United States and (see United States) accession to WTO) welfare impact, 15, 131­132, 134­137 role in region, xiv­xvii (see also China) CITES (Convention on International Trade in Endan- extent of regional interdependence, xvii­xviii gered Species), 129 global integration (see global integration) clothing. see textiles logistics and transport issues, regional, 81­85, 87 (see CMI (Chiang Mai initiative), x, 43 also logistics) Colombia, 119, 127, 166 reasons for current importance of, x COMECON (Council for Mutual Economic Assistance), trade trends and opportunities, xi­xiv collapse of, 162­163 EBA (Everything But Arms) Agreement, 68 comparable general equilibrium (CGE) modeling, 133 education competition China's accession to WTO, 17 environmental standards, xxviii­xxix, 115­116, vulnerability of trade and labor market in Indonesia, 126­129 Republic of Korea, and Thailand, xxxiii, 177­178, investment and competition policies, xxvi­xxvii 179 labor standards, xxviii­xxix, 115­116, 118­123 EFTA (European Free Trade Area), 41 regulatory and pro-competitive reforms, xxv, xxvi Egypt, 119, 127 Convention on Biodiversity (CBD), 104­105 employment. see labor standards and employment Convention on International Trade in Endangered environmental standards, xxviii­xxix, xxxv, 115­116, Species (CITES), 129 123­131. see also tables, charts, and figures copyright in Indonesia, 92, 99­102, 108 competition, xxviii­xxix, 115­116, 126­129 film, 101­102 Doha Round, 129­130 legal structures and enforcement, 100 East Asia currency and financial crisis, 1997-1998, music recordings, 100­101 123­124 pirating, 99, 100 exports of pollution-intensive products, 127­129 print publishing, 102 FDI and, 115­116, 124­126 software, 100­101 logistics in urban environments, 87 194 East Asia Integrates policy and policy reforms, 130­131 gender and vulnerability of trade and labor market in political freedom index, 128 Indonesia, Republic of Korea, and Thailand, trade liberalization and openness, effect of, 124­126 177­178, 179 EPZs (export processing zones), 120 General Agreement on Tariffs and Trade (GATT) European Free Trade Area (EFTA), 41 ATC (Agreement on Textiles and Clothing), Uruguay European Union (EU) Round, xxi China's accession to WTO, 11, 24, 28 China's accession to WTO, 4, 6, 9, 17 compared to East Asia, xvii environmental standards, 129 environmental standards, 127 Japan and, 4 logistics, 85 General Agreement on Trade in Services (GATS) market access barriers, poverty, and agricultural China's accession to WTO, 12, 13, 17 products, 61, 64, 65, 66­67, 68, 69, 72 investment and competition policies, xxvi RTAs, 41, 45 logistics, 86 trade trends and opportunities, xi general equilibrium analysis, 133 Everything But Arms (EBA) Agreement, 68 Generalized System of Preferences (GSP), 63, 68­70 export processing zones (EPZs), 120 genetic resources, traditional knowledge regarding. see exports. see also tariffs traditional knowledge protections agricultural and agroprocessed exports, 60­61 (see Germany, 83, 94, 96 also market access barriers, poverty, and agricul- Gini coefficient for interpersonal income inequality, xxx tural products) global integration, xix­xxi cashmere in Mongolia, 161­163, 166 rice, global trade in, 63, 65 China's accession to WTO, 14­15, 18, 23­24 traditional knowledge protections, 104­105 environmental standards and exports of pollution- Global Trade Analysis Project (GTAP), 14, 133, 134 intensive products, 127­129 Goh, Char-ching, 165­181 labor standards and employment, 120­123 Goletti, Francesco, 149­169 rice from Cambodia and Vietnam, 153­155, 159 Gooptu, Sudarshan, 21­38 Greece, 83 F GSP (Generalized System of Preferences), 63, 68­70 FA (freedom of association) rights, 121, 122 GTAP (Global Trade Analysis Project), 14, 133, 134 farming. see agriculture FDI. see foreign direct investment (FDI) H figures. see tables, charts, and figures health and safety standards, xxii film and copyright in Indonesia, 101­102 high-tech products, ASEAN and China, 48 financial issues Hong Kong (China) China's accession to WTO bilateral investment treaties, xxvii economic reforms, 13­15 China's accession to WTO, effect of, 6, 26 financial services, 13 customs clearance, 83 investment patterns, 24­26, 27 economic/income gap, ix investment (see investment) labor standards and employment, 117, 121 RTAs and financial cooperation on regional level, logistics, 74, 75, 76, 81, 85 43­44 market access barriers, poverty, and agricultural financial services and China's accession to WTO, 13 products, 63, 66­67 folklore and traditional knowledge protections, 102 RTAs, 40, 41 foreign direct investment (FDI) household-level effects China's accession to WTO, effect of, 24­26 cashmere trade in Mongolia, 163, 167 environmental standards, 115­116, 124­126 China's accession to WTO (see China's accession to labor standards, 115­116, 120­123 WTO) France, 75, 83, 94 trade reforms, impact of, xxx­xxxi free trade hub and spoke arrangements, 46­47 ASEAN-China Free Trade Area, 37, 41, 49­51 hukou system and China's accession to WTO, 15, 16, 18 current and proposed free trade areas, 41 Hummels, David, 73­89 regional blocs, effects of, 45­46 freedom of association (FA) rights, 121, 122 I Ianchovichina, Elena, 21­38 G Ikezuki, Takako, 69­73 GATS. see General Agreement on Trade in Services ILO (International Labour Organization), xxix, 117, 121, (GATS) 122 GATT. see General Agreement on Tariffs and Trade imports. see also tariffs (GATT) China's accession to WTO, xvi, 15, 18, 22­23 Index 195 market access barriers, poverty, and agricultural J products, 63 Japan India bilateral investment treaties, xxvii China's accession to WTO, 26 China's accession to WTO customs clearance, 83 national impact, 4, 5, 7, 9 dumping, 6 regional impact, 22, 24, 26, 27, 28 environmental standards, 127 environmental and labor standards, 115­116 intellectual property rights, xxvii logistics, 75, 85 labor standards and employment, 119 market access barriers, poverty, and agricultural market access barriers, poverty, and agricultural products, 61, 63, 64, 65, 66­67, 69, 72 products, 61 market-driven integration or regionalization, 40 RTAs, 41 patents, 95 Indonesia RTAs, xviii, 37, 40, 41, 43 bilateral investment treaties, xxvii Thailand, 32 China's accession to WTO, xvii, 27, 28­30 CMI, 43 K copyright (see copyright in Indonesia) Khuras, Homi, ix­xxxviii customs clearance, 83 Korea. see Republic of Korea environmental standards, 124, 127 Krumm, Kathie, ix­xxxviii labor standards and employment, 117, 119, 121 logistics, xxiv, 75, 76­77, 85 L market access barriers, poverty, and agricultural labor standards and employment, xxviii­xxix, xxxv, products, 63, 66­67, 69 115­123. see also tables, charts, and figures regulatory and pro-competitive reforms, xxvi child labor, 117 trade trends and opportunities, xii China's accession to WTO, xxxi, 15, 18 vulnerability of trade and labor market (see vulnera- competition, xxviii­xxix, 115­116, 118­123 bility of trade and labor market in Indonesia, determinants of standards, 118 Republic of Korea, and Thailand) EPZs, 120 industrial inventions. see intellectual property rights export performance, effect on, 120­123 industrial products. see manufacturing FA (freedom of association) rights, 121, 122 informal employment, 117­118 FDI, 115­116, 120­123 information technology and logistics, 82, 84 household-level effects of trade reforms, xxxi intellectual property rights, xxvii­xxviii, 91­110. see also ILO conventions, ratification of, 117, 121, 122 tables, charts, and figures informal employment, 117­118 China's accession to WTO, 6­7 minimum wage laws, 116­117 copyright (see copyright in Indonesia) policy and policy reform, 117, 130­131 defined, 91 political freedom index, 122 patents (see patents in Korea) poverty, 118, 119 pirating, 99, 100 ranking standards, 118, 119 policy recommendations, xxxv rights of workers, 116, 117 traditional knowledge (see traditional knowledge social protection measures, 117 protections) tariffs and trade liberalization, 118­120 TRIPS [see Agreement on Trade-Related Aspects of vulnerability to trade liberalization (see vulnerability Intellectual Property Rights (TRIPS)] of trade and labor market in Indonesia, Republic International Conventions for the Protection of New of Korea, and Thailand) Varieties of Plants (UPOV), 104 Lao Peoples' Democratic Republic (Lao PDR) International Labour Organization (ILO), xxix, 117, 121, China's accession to WTO, xvii, 33, 34­35 122 economic/income gap, ix, xxix intraregional trade trends, xiii logistics, xxiv, 75, 76, 77­78, 78, 79, 86, 87 inventions. see intellectual property rights market access barriers, poverty, and agricultural inventory costs and logistics, 79­80 products, 61, 69, 70 investment MFN status with China, xix bilateral investment treaties, East Asian countries, poverty and accessibility, 78 xxvi­xxvii trade trends and opportunities, xi, xiv China's accession to WTO, effect of, 24­26, 27 Li, Shantong, 3­20 FDI [see foreign direct investment (FDI)] logistics, xxii­xxv, xxxiv­xxxv, 73­89. see also tables, policies on investment and competition, charts, and figures xxvi­xxvii air freight, 85 IPRs. see intellectual property rights benefits from improvement to, 79­81 196 East Asia Integrates China's accession to WTO, 12 NTMs, 70, 71 commodity structure of trade, 74­76 preferential rates, 68­70 country groupings, 74­78 rice, global trade in, 63, 65 cross-border facilitation, xxv, 86­87 significance of agricultural income to middle and customs clearance, xxv, 82, 83 lower-income countries, 59­60 development of markets and product diversification, tariff barriers, 63­65 80 tariff escalation, 65­68 domestic integration of systems, 85­86 tariff rates, 66­67 inventory costs, 79­80 trade policy recommendations, 72 land access to ports, 83­85 trade policy structure of EU, Japan, and United States less open and accessible (Group 3) countries, 77­78 for East Asian imports, 64­65, 68 maritime transport, 81­82 market development and logistics, 80 multimodal transport, 82, 84 market-driven integration or regionalization, 39­40 outward-oriented, accessible (Group 2) countries, market opening and China's accession to WTO, 4­5 76­77 marketing margins and costs for rice in Cambodia and outward-oriented, highly accessible (Group 1) coun- Vietnam, 158­161 tries, 76 Martin, Will, 3­20 per capita income levels, 74­76 Maskus, Keith E., 91­110, 115­134 policy recommendations, 85­87 MEAs (multilateral environmental agreements), 129 pricing differentials, 79 medical research and traditional knowledge protections. private sector involvement, 86 see traditional knowledge protections regional cooperation, 87 Mekong River Commission and Greater Mekong subre- regional price fluctuations, 79 gion, 87 regional transport issues, 81­85 Mexico regulatory environment for, 86­87 customs clearance, 83 remote regions within countries, 78­79 dumping, 6 security issues, 87 environmental standards, 126, 127 trade and logistics nexus, 74 labor standards and employment, 119 urban land use and management, 87 logistics, 75 value chain, position on, 80­81 RTAs, 40, 43 low-income countries vulnerability of trade and labor market, 166 agriculture as significant source of income for, 59­60 MFA (Multi-fiber Arrangement), 4­5 China's accession to WTO, effect of, 33­36 MFN. see most favored nation (MFN) status intellectual property rights, xxviii middle-income developing countries Luthria, Manjula, 91­110 agriculture as significant source of income for, 59­60 China's accession to WTO, regional impact of, 27­33 M intellectual property rights, xxviii Malaysia trade trends and opportunities, xi bilateral investment treaties, xxvii minimum wage laws, 116­117 China's accession to WTO, effect of, 27, 30­31 Mongolia, xxiv, xxxi, 75­79, 87. see also cashmere in customs clearance, 83 Mongolia economic/income gap, xxix Montreal Protocol, 129 environmental standards, 125, 127 Morocco, 166 labor standards and employment, 117, 119, 121 most favored nation (MFN) status logistics, xxiv, 75, 76­77, 81 China's accession to WTO, 4­5 market access barriers, poverty, and agricultural "Early Harvest" provisions, China-ASEAN agree- products, 66­67, 69 ment, xix, 33 regulatory and pro-competitive reforms, xxvi market access barriers, poverty, and agricultural manufacturing industries and China's accession to products, 63, 64, 68, 70 WTO, xv, 9­11, 17, 23 market opening principle of GATT and WTO, 4 maritime transport, 81­82 motor vehicle industry and China's accession to WTO, market access barriers, poverty, and agricultural prod- xv, 10­11, 26­27 ucts, xxi, 59­73. see also tables, charts, and figures movies and copyright in Indonesia, 101­102 agricultural liberalization, importance of, 60 Mozambique, 83 domestic support, 63, 64 MRAs (mutual recognition agreements), xxii exports, agricultural and agroprocessed, 60­61 Multi-fiber Arrangement (MFA), 4­5 exposure of poor to international markets, 60­61 multilateral environmental agreements (MEAs), 129 imports, 63 music key markets for agroproducts, 61­63 copyright in Indonesia, 100­101 Index 197 traditional knowledge protections, 102 labor standards and employment, 117, 119, 121 mutual recognition agreements (MRAs), xxii logistics, xxiv, 75, 76­77, 78, 85 Myanmar, xiv, 61, 66­67, 69 market access barriers, poverty, and agricultural products, 65, 66­67, 69 N regulatory and pro-competitive reforms, xxvi NAFTA (North American Free Trade Agreement), xi, RTAs, 41 xvii, 122, 128 trade trends and opportunities, xii, xiv Netherlands, 83 traditional knowledge protections, 107 New Zealand, 37, 40, 41, 45 pirating of copyright material, 99, 100 newly industrializing economies (NIEs) Poland, 75 China's accession to WTO, impact of, 26­27, 28 policy and policy reforms, ix, x, xxxiii­xxxvi market-driven integration or regionalization, 40 "behind the border" reform issues, xxi­xxix trade trends and opportunities, xi cashmere in Mongolia, 151­152, 163­164, 166­167 nondiscrimination principle and China's accession to China's accession to WTO, household-level impact WTO, 4 of, 131, 132, 140 nontariff measures (NTMs), 70, 71 environmental standards, 130­131 North American Free Trade Agreement (NAFTA), xi, household-level effects, xxx­xxxi xvii, 122, 128 labor standards, 117, 130­131 Norway, 11 logistics, policy recommendations for, 85­87 NTMs (nontariff measures), 70, 71 market access barriers, poverty, and agricultural products O trade policy recommendations, 72 Organisation for Economic Co-operation and Develop- trade policy structure of EU, Japan, and United ment (OECD), 9, 120, 126 States for East Asian imports, 64­65, 68 predictability of trade policy and China's accession to P WTO, 5 Pakistan, 61, 119, 127 regulatory and pro-competitive reforms, xxv, xxvi Panama, 41 rice in Cambodia and Vietnam, 151­152, 155, Pangestu, Mari, 21­38 157­161 Papua New Guinea, 76, 77­78 vulnerability of trade and labor market to (see vul- Paris Convention for the Protection of Industrial Prop- nerability of trade and labor market in Indonesia, erty, 95 Republic of Korea, and Thailand) partial equilibrium analysis, 133 WTO, associated with accession to, xv, 4­7, 11 patents in Korea, 92­99, 106. see also tables, charts, and political freedom index, environmental and labor stan- figures dards, 122, 128 citations of, 96­97, 98 ports, 81­82. see also logistics comparison with other countries, 93, 94, 96­99 container movements at selected ports, 81 corporate management push, 95­96 land access to, 83­85 demand and supply side factors contributing to rise poverty. see also tables, charts, and figures in, 93 China's accession to WTO, 16­17, 131, 132, 134­140 fertile technology pull, 93­95 economic/income gap, ix, xxix­xxx friendly court pull, 94­95 labor standards, 118, 119 industrial upgrading push, 95 logistics, poverty, and accessibility in Lao PDR, 78 R&D activity, 95­96 market access barriers (see market access barriers, semiconductor patents poverty, and agricultural products) comparison charts, 94 micro-interventions to relieve, xxxi fertile technology pull, 94­95 reduction efforts, xxix, xxx, xxxi, xxxv reasons for prevalence of, 97­99 trade in sectors important to the poor (see cashmere TRIPS, 99, 108 in Mongolia; rice in Cambodia and Vietnam) U.S., Korean patents granted in, 95, 98 vulnerability to, indirect estimation using cross-sec- Peru, 119, 127, 166 tional data, 175­178 pharmaceuticals, 97, 102 preferential trade agreements, 37­38, 45, 63. see also Philippines regional trade agreements (RTAs) bilateral investment treaties, xxvii preferential treatment for developing countries, 6, 68­70 China's accession to WTO, 26 pricing differentials and logistics, 79 China's accession to WTO, effect of, 27, 31­32 print publishing and copyright in Indonesia, 102 customs clearance, 83 private sector business, 86, 104 economic/income gap, ix, xxx product diversification and logistics, 80 environmental standards, 124, 127 publishing and copyright in Indonesia, 102 198 East Asia Integrates R Vietnam, issues specific to, 157 R&D. see research & development (R&D) activity rights of workers, 116, 117 Ravallion, Martin, 131­142 RTAs. see regional trade agreements (RTAs) recorded music and copyright in Indonesia, 100­101 rural households and China's accession to WTO, 15­16, regional trade agreements (RTAs), xvii­xix, 37­55. see 131, 136­140 also tables, charts, and figures Russia, 83 AFTA [see AFTA (ASEAN Free Trade Area)] ASEAN [see Association of South East Asia Nations S (ASEAN)] safeguards and China's accession to WTO, xv, 5­6, 17 blocs vs. free trade, 45­46 Samoa, 76, 77­78 China (see China) Sangwe, Vera, 149­169 costs and benefits, 44­47 sanitary and phytosanitary (SPS) standards, xxii current and proposed agreements, 39­43 security issues and logistics, 87 financial cooperation on regional level, 43­44 semiconductor patents in Korea. see patents in Korea forms of, 38 September 11 attacks, 87 hub and spoke arrangements, 46­47 services sector, xvi, xxv­xxvi, 10­13, 23 market access barriers, poverty, and agricultural sex and vulnerability of trade and labor market in products, 63 Indonesia, Republic of Korea, and Thailand, reasons for new resurgence in, 38­39 177­178, 179 regional plus agreements, 37, 41 Singapore regionalization vs. regionalism, 39 bilateral investment treaties, xxvii regulatory and pro-competitive reforms, xxv, xxvi China's accession to WTO, 26 Republic of Korea customs clearance, 83 agricultural protections, 9 intellectual property rights, xxvii bilateral investment treaties, xxvii labor standards and employment, 117 China's accession to WTO, effect of, 26 logistics, xxiv, 74, 75, 76, 81, 85 corporate cartels, xxvii market access barriers, poverty, and agricultural economic/income gap, xxix products, 66­67 environmental and labor standards, 115­116 regulatory and pro-competitive reforms, xxvi environmental standards, 123, 127 RTAs, 37, 40, 41 intellectual property rights, xxvii size of household and effects of China's accession to labor standards and employment, 117, 119, 121 WTO, 138 logistics, xxiv, xxv, 75, 76, 80, 85 social protection measures, 17, 18, 117 market access barriers, poverty, and agricultural social stability, reinforcing, xxix­xxxiii products, 63, 66­67 software and copyright in Indonesia, 100­101 patents (see patents in Korea) South Africa, 6, 119, 127 regulatory and pro-competitive reforms, xxvi South Korea. see Republic of Korea RTAs, 37, 40, 42, 43 Spain, 83 services liberalization, xxv standards requirements, xxii textiles, xxi Suthiawart-Narueput, Sethaput, 21­38 trade trends and opportunities, xi, xii Sweden, 83 vulnerability of trade and labor market (see vulnera- bility of trade and labor market in Indonesia, T Republic of Korea, and Thailand) tables, charts, and figures research & development (R&D) activity bilateral investment treaties, East Asian countries, patents in Korea, 95­96 xxvii traditional knowledge protections, 103­104 cashmere in Mongolia rice, global trade in, 63, 65 economic role of, 162 rice in Cambodia and Vietnam, xxxi­xxxii, 149­161, share of cashmere products in total cashmere 167­168. see also tables, charts, and figures exports, 166 aggregate effects of policy options, 158 value chain analysis, 152 Cambodia, issues specific to, 156­157 China's accession to WTO different participants, effect of policy on, 158­161 agriculture protections, xv, 8 exports of rice, 153­155, 159 exports, 24 importance of rice to economy and society, household-level effects (see subheading "house- 151­153 hold-level effects of China's accession to marketing margins and costs, 158­161 WTO," this entry) policy and policy reform, 151­152, 155, 157­161 imports, xvi, 22, 23 value chain analysis, 149­152, 155­161, 167­168 low-income countries, effect on, 33­35 Index 199 manufacturing protections, 10 trade policy structure of EU, Japan, and United middle-income developing countries, effect on, States for East Asian imports, 68 27­30 marketing margins of farmers in Thailand, United NIEs, effect on, 26 States, Cambodia, and Vietnam, 161 reductions in protection, impact of, 14 patents in Korea e-commerce, 84 citations of Korean patents, 97 "Early Harvest" provisions, China-ASEAN agree- research productivity of selected countries, 96 ment, 50 self-citation rates of Korean private corporations, environmental standards 98 impact on pollution-intensive exports, 128 semiconductor patent comparisons, 94 policy indexes, 127 trend comparisons, Korea and major developing sustainability index, xxviii, 127 countries, 93 Gini coefficient for interpersonal income inequality, U.S., Korean patents granted in, 98 xxx poverty household-level effects of China's accession to WTO, declining rate of, xxx xxxi Gini coefficient for interpersonal income inequal- mean gains, 137­139 ity, xxx percentage of gainers, 139 market access barriers to agricultural products predicted aggregate impacts, 135 (see subhead "market access barriers, poverty, rural poverty incidence curves, 135 and agricultural products," this entry) urban poverty incidence curves, 136 regional trade trends, xii ILO conventions, ratification of, 117 regulatory and pro-competitive reforms, xxvi intellectual property rights rice, global trade in, 65 patents in Korea (see subhead "patents in Korea," rice in Cambodia and Vietnam this entry) economic role of, 153 research productivity of selected countries, 96 income of poor and, 154 traditional knowledge protections in the Philip- land size and rice profit, 154 pines, 107 marketing costs and margins, 160, 161 intraregional trade trends, xiii profit shares, 155 labor standards and employment profit structure, 161 exports, effect on, 122 strategic policy options, effects and analysis of, ILO conventions, ratification of, 117 158, 159 index of countries, xxviii value chain analysis, 152 ranking of standards, 119 RTAs logistics average tariff rates, xx air freight, 85 current and proposed agreements, 40­41 APEC, expansion of exports to, xxiv "Early Harvest" provisions, China-ASEAN agree- availability of transport infrastructure, 75 ment, 50 country groupings by openness, 75 economic welfare effects of proposals, xviii customs clearance times, 83 forms of, 38 e-commerce, 84 multidimensional nature of, xx land access to ports, 84 third country markets, China's trade in, xiv manufactured exports by country, structure of, traditional knowledge protections in the Philippines, xxiii 107 poverty and accessibility in Lao PDR, 78 value chain analysis, 152 producer services, costs of, xxiv vulnerability of trade and labor market in Indonesia, remote regions, xxiii, 80 Republic of Korea, and Thailand selected ports, container transport from, 81 annual wage growth, xxxiii market access barriers, poverty, and agricultural average monthly hours worked, 174 products average tariffs and percent change in tariffs, 167 domestic support levels, 64 core nontariff barrier measures, 169 key markets, 62 decade average shares of import tax revenue, 169 NTMs, 70, 71 decade averages of import tariffs for manufactur- real income gains from agricultural liberalization, ing industries, 168 60 demographic factors, 177, 178, 179 rice, global trade in, 65 distribution of employment status and sectors, 174 tariff escalation, 69 education in Korea, xxxiii tariff rates, 66­67 log of real wages, 171, 172, 173 top export products, 61 standard deviation of wage growth, xxxiii 200 East Asia Integrates trade sectors, 178 WTO, 107 Taiwan (China) transparency of trade policy and China's accession to bilateral investment treaties, xxvii WTO, 5 China's accession to WTO, effect of, 26 transport systems. see logistics customs clearance, 83 TRIPS. see Agreement on Trade-Related Aspects of Intel- environmental and labor standards, 115­116 lectual Property Rights (TRIPS) logistics, 75, 76, 80 Turkey, 119, 127 market access barriers, poverty, and agricultural products, 63, 66­67, 69 U RTAs, 40, 41 undistorted trade and China's accession to WTO, 5­6 tariffs United Kingdom (UK), 94 labor standards and employment, 118­120 United Nations, xxii market access barriers, poverty, and agricultural United States products China's accession to WTO tariff barriers, 63­65 national impact, 4, 5, 7, 11 tariff escalation, 65­68 regional impact, 22, 24, 26, 28 tariff rates, 66­67 customs clearance, 83 vulnerability of trade and labor market in Indonesia, environmental standards, 127 Republic of Korea, and Thailand, 167­169 intellectual property rights, xxvii technical requirements and standards, xxii Korean patents granted in, 95, 98 telecommunications industry and China's accession to labor standards and employment, 120 WTO, 12­13 logistics, 75, 80, 85, 87 textiles market access barriers, poverty, and agricultural ATC, xxi products, 61, 64, 65, 66­67, 68, 69, 72 China marketing margins of farmers, 161 easing adjustments in industry, xxxv RTAs, 40, 41, 43 global integration, xx­xxi Thailand, 32 WTO, accession to, xv, 4­5, 11, 14­15, 26, 33, 48 third country markets, China's trade in, xiv easing adjustments in industry, xxxv UPOV (International Conventions for the Protection of global integration, xx­xxi New Varieties of Plants), 104 Thailand urban environments bilateral investment treaties, xxvii cashmere in Mongolia, 163 China's accession to WTO, xvii China's accession to WTO, household-level effects China's accession to WTO, effect of, 11, 27, 28, 29, of, 15­16, 131, 136­138, 140 32­33 logistics, 87 customs clearance, 83 environmental standards, 123­124, 127 V labor standards and employment, 117, 119, 121 value chain logistics, xxiv, 75, 76­77, 82, 83, 85 cashmere in Mongolia, xxxii, 149­152, 163­166, market access barriers, poverty, and agricultural 167­168 products, 65, 66­67, 69 logistics, 80­81 marketing margins of farmers, 161 rice in Cambodia and Vietnam, xxxii, 149­152, RTAs, 40, 41, 42, 43 155­161, 167­168 technical requirements and standards, xxii Venezuela, 119, 127 textiles, xxi VERLS labor standards, 121, 122 trade trends and opportunities, xii Vietnam vulnerability of trade and labor market (see vulnera- bilateral investment treaties, xxvii bility of trade and labor market in Indonesia, China's accession to WTO, xvii Republic of Korea, and Thailand) China's accession to WTO, effect of, 26, 27, 33, 35­36 trade policy and reform. see policy and policy reforms customs clearance, 83 traditional knowledge protections, 92, 102­108 economic/income gap, ix, xxx evolution of, 104­105 household-level effects of trade reforms, xxxi global policy development, 104­105 labor standards and employment, 117 Philippines, 107 logistics, xxiv, 75, 76, 77­78, 79, 86 practical issues and problems, 106­107 market access barriers, poverty, and agricultural property rights products, 61, 68, 69, 70 history of, 104­105 rice (see rice in Cambodia and Vietnam) possible options for, 105­106 textiles, xxi TRIPS, 103, 105, 107 trade trends and opportunities, xi, xiv Index 201 vulnerability of trade and labor market in Indonesia, welfare. see social protection measures Republic of Korea, and Thailand, xxxii­xxxiii, welfare impact of China's accession to WTO, 15, xxxv, 165­181. see also tables, charts, and figures 131­132, 134­137 cross-sectional data, indirect estimation of changes WHO (World Health Organization), xxii in vulnerability from, 175­178 World Bank, xxii, 87 demographic factors, 177­178 World Health Organization (WHO), xxii education, xxxiii, 177­178 World Trade Organization (WTO) GDP (gross domestic product) growth rates, volatil- bound tariff rates, 63 ity of, 169, 170 China's accession to (see China's accession to synthetic panels, earnings and employment trends, WTO) 170, 171­174 Doha Round, Declaration, and Development tariffs, 167­169 Agenda, xx, xxii, 9, 45, 116, 129­130 trade liberalization five basic principles, 4 contested effects of, 165­166 intellectual property rights, xxvii labor market volatility, 165­171 logistics, 74, 86 measures of, 167­169 RTAs, 37, 40, 45 volatility of aggregate indicators, 169­171 technical requirements and standards, xxii wage growth rates, volatility of, 169, 170 traditional knowledge protections, 107 W Z wage stability. see vulnerability of trade and labor mar- Zhao, Min, 21­38 ket in Indonesia, Republic of Korea, and Thailand Zimbabwe, 83 T R A D E A N D D E V E L O P M E N T S E R I E S E merging East Asian economies have seen their share of world exports more than triple during the past quarter-century, and intraregional trade has driven this growth. Broad measures of development in East Asia have improved at the same headlong pace. Why push further integration now? Two economic events of historic proportions provide the context: strategic rethinking of development in the region following the East Asian financial crisis of 1997­98 and the accession of China to the World Trade Organization. Policymakers interested in a stable, prosperous region are concerned by mildly rising inequality within countries and a widening income gap between richer economies and the poorest economies. Increasingly, the development agenda in the region--with its focus on growth, jobs, and social stability--and the trade policy agen- da--with its focus on market access and competitiveness--have become intertwined. East Asian policymakers seek to develop a coherent set of economic policies that can deliver stability, growth, and regional integration. Without attempting to be compre- hensive, East Asia Integrates offers fundamental strategies that promote cross-border flows of trade, globally and regionally, along with domestic policies on logistics, trade facilitation, standards, and institutions to maximize the impact of these flows on devel- opment and to distribute the gains from trade widely. As the authors demonstrate, multilateral and regional trade initiatives must provide a compelling vision of how integration can deliver broadly shared growth and prosperity if they are to succeed. In addition, they must use the momentum offered by trade agreements to address the links between trade on the one hand, and social stability, poverty reduction, and growth on the other. THE WORLD BANK ISBN 0-8213-5514-7