Valuing Services in Trade Valuing Services in Trade A Toolkit for Competitiveness Diagnostics Sebastián Sáez, Daria Taglioni, Erik van der Marel, Claire H. Hollweg, Veronika Zavacka © 2014 International Bank for Reconstruction and Development / The World Bank 1818 H Street NW, Washington DC 20433 Telephone: 202-473-1000; Internet: www.worldbank.org Some rights reserved 1 2 3 4 17 16 15 14 This work is a product of the staff of The World Bank with external contributions. The findings, interpretations, and conclusions expressed in this work do not necessarily reflect the views of The World Bank, its Board of Executive Directors, or the govern- ments they represent. The World Bank does not 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 any judgment on the part of The World Bank concerning the legal status of any territory or the endorsement or acceptance of such boundaries. Nothing herein shall constitute or be considered to be a limitation upon or waiver of the privileges and immunities of The World Bank, all of which are specifically reserved. Rights and Permissions This work is available under the Creative Commons Attribution 3.0 IGO license (CC BY 3.0 IGO) http://creativecommons .org/licenses/by/3.0/igo. Under the Creative Commons Attribution license, you are free to copy, distribute, transmit, and adapt this work, including for commercial purposes, under the following conditions: Attribution—Please cite the work as follows: Sáez, Sebastián, Daria Taglioni, Erik van der Marel, Claire H. Hollweg, and Veronika Zavacka. 2014. Valuing Services in Trade: A Toolkit for Competitiveness Diagnostics. Washington, DC: World Bank. doi:10.1596/978-1-4648-0155-6. License: Creative Commons Attribution CC BY 3.0 IGO Translations—If you create a translation of this work, please add the following disclaimer along with the attribution: This translation was not created by The World Bank and should not be considered an official World Bank translation. The World Bank shall not be liable for any content or error in this translation. Adaptations—If you create an adaptation of this work, please add the following disclaimer along with the attribution: This is an adaptation of an original work by The World Bank. Views and opinions expressed in the adaptation are the sole responsibility of the author or authors of the adaptation and are not endorsed by The World Bank. Third-party content—The World Bank does not necessarily own each component of the content contained within the work. The World Bank therefore does not warrant that the use of any third-party-owned individual component or part contained in the work will not infringe on the rights of those third parties. The risk of claims resulting from such infringement rests solely with you. If you wish to re-use a component of the work, it is your responsibility to determine whether permission is needed for that re-use and to obtain permission from the copyright owner. Examples of compo- nents can include, but are not limited to, tables, figures, or images. All queries on rights and licenses should be addressed to the Publishing and Knowledge Division, The World Bank, 1818 H Street NW, Washington, DC 20433, USA; fax: 202-522-2625; e-mail: pubrights@worldbank.org. ISBN (paper): 978-1-4648-0155-6 ISBN (electronic): 978-1-4648-0156-3 DOI: 10.1596/978-1-4648-0155-6 Cover design: Debra Naylor, Naylor Design, Inc. Library of Congress Cataloging-in-Publication Data Sáez, Sebastián. Valuing services in trade: a toolkit for competitiveness diagnostics / Sebastián Sáez, Daria Taglioni, Erik van der Marel, Claire H. Hollweg, and Veronika Zavacka. pages cm ISBN 978-1-4648-0155-6 — ISBN 978-1-4648-0156-3 1. Service industries. 2. International trade. 3. Competition. I. World Bank. II. Title. HD9980.5.S24 2014 382’.45—dc23 2014008433 CONTENTS Foreword xi Preface xiii Acknowledgments xv About the Authors xvii Abbreviations xix Balance of Payment (BOP) Codes Used in This Book xxi Glossary xxiii Introduction 1 The Role of Services in Trade and Competitiveness: Key Issues and Debates 1 The Role of Regulation 5 Organization of the Toolkit 8 Notes 9 References 9 Module 1 Assessing Services Trade and Competitiveness Outcomes 11 Size of Trade in Services 11 Comparing a Country’s Trade in Services with Peers and the World 18 Notes 46 References 46 Module 2 Services as a Source of Competitiveness in the Whole Economy 47 Services in the Domestic Economy 47 Linkages to the Rest of the Economy 55 Services Indirect Linkages 66 Note 70 References 70 Module 3 Assessing the Potential for Trade in Services 71 Assessing Tradability by Modes 71 Assessing Tradability Based on Production 75 Assessing Tradability Based on Comparative Advantage 80 Assessing Tradability through Gravity Models 85 Annex 88 Notes 91 References 91 Module 4 Policy Options for Increasing Competitiveness and Trade in the Services Sector 93 Reasons for Regulatory Policies 93 Trade Policy Barriers 103 Domestic Factors that Enable Trade in Services 111 Notes 122 References 122 v vi Contents Appendix A Export of Value Added Database 125 Appendix B Trade in Services Database 135 Boxes 1.1 Modes of Services Trade and Data on Trade in Services 12 1.2 How Can Developing Countries Join and Climb the Ladder of Global Value Chains? 19 1.3 What Kind of Firms Export Services? 21 1.4 Using Firm-Level Data to Estimate Concentration Ratios: Evidence from Romania 34 1.5 Services Sophistication: Handle with Care 37 1.6 What Is a Gravity Model? 45 2.1 Does India Defy the General Pattern of Ladders of Comparative Advantage? 51 2.2 What Is the Structure of the Services Sector? Evidence from Romania 55 2.3 Where Do Services Inputs Go? Evidence from Ghana 60 2.4 Output, Value Added, and the Importance of the Wholesale Sector for Exports of Downstream Goods: Evidence from Romania 68 3.1 Remittances are not a measure of Mode 4 74 3.2 Quantifying Tradability Potential: Evidence from Romania 79 3.3 The Updated Tradability Index of Gervais and Jensen 80 3.4 How Are Factor Intensities Computed? 82 3.5 What Does Firm Productivity Say about the Likelihood of Exporting? Evidence from Romania 84 4.1 Weak Policies, Weak Investment Climate, and Weak Integration into World Markets in the Middle East and North Africa 94 4.2 Good Policy Practice: Increasing the Transparency of Regulations Governing Lawyers Working in Asia-Pacific Economic Cooperation (APEC) Member Economies 97 4.3 Good Policy Practice: Achieving Universal Postal Coverage in Trinidad and Tobago 98 4.4 Good Policy Practice: Increasing Access to Health Services in Thailand through Regulation Rather Than Provision 99 4.5 What Are the Effects of Planning and Zoning Regulations? 102 4.6 How Can Developing Countries Increase the Technological Capabilities of Their Services Sector? 104 4.7 The World Bank’s Services Trade Restrictiveness Index (STRI) Database 107 4.8 Domestic Regulations and the World Trade Organization’s Necessity Test 108 4.9 The Australian Productivity Commission’s Trade Restrictiveness Index 109 4.10 Static and Dynamic Impact of Restrictions on Entry and Operations 109 4.11 Challenges Facing Potential Services Exporters in Kenya 113 4.12 Has the Philippine Economic Zone Authority (PEZA) Boosted Services Exports? 116 4.13 What Are Developing Countries Doing to Develop Skilled Labor in High-Tech Services Sectors? 118 4.14 Assessing the Knowledge Economy 120 A.1 Export of Value Added Database Variables 127 Figures I.1 Services Trade as Share of GDP, by Region 3 I.2 Services Involved in the Internationalization of Production 4 I.3 Value Added of Services as Share of Value Added of Manufacturing, by Sector, 2009 5 I.4 From Sector-Based to Task-Based Development Strategies 5 I.5 Cross-Country Regression of Services Exports versus Gross Tertiary Enrollment and Internet Penetration 8 1.1 Shares of Exports of Goods and Services, by Country Income Level, 2011 13 1.2 Index of Services Exports, by Country Income Level, 2000–10 13 1.3 Contribution to World Services Exports, by Income Level, 1990–2010 13 1.4 Services Trade Balance as Percent of GDP, by Income Group and in Selected Countries, 2001–10 14 1.5 Trade Balance for Goods and Services, by Income Group, 2000–10 15 1.6 Trade Balance in Kazakhstan, the Kyrgyz Republic, and Tajikistan, as Percent of GDP, 2000–10 16 Contents vii 1.7 Per Capita Trade in Services and Per Capita GDP, 2002–04 and 2008–10 17 1.8 Per Capita Trade in Services and Ordinary Least Squares Residuals, 2002–04 and 2008–10 18 1.9 Per Capita Exports in Services and Per Capita GDP, 2002–04 and 2008–10 19 1.10 Size of Sector and Exports of Sector as Share of GDP in Selected Economies, 2008–10 20 B1.3.1 Number of Employees of Exporting and Nonexporting Firms, 2009 21 1.11 Commercial Services Exports by Turkey and Selected Comparator Countries, 2010 21 1.12 Services Exports and Diversification in Selected Economies, 2008–10 22 1.13 Annual Growth of Exports of Services by India, Turkey, and Regional and Income-Group Comparators, 1994–2010 23 1.14 World Demand for Imported Services and Exports of Services by India, Malaysia, and Mexico, 2002–09 24 1.15 Services Exports by India, by Sector and Destination, 2000–08 25 1.16 Balance of Trade in Services, by Country Income Level and Sector, 2000–10 26 1.17 Balance of Trade in Services in Selected Lower-Middle-Income Countries, by Sector, 2001–11 27 1.18 Sectoral Shares of Services Exports by Selected Countries in Central Asia, 2002–10 28 1.19 Sectoral Shares of Services Exports and GDP Per Capita in Selected Countries, 2009 29 1.20 Trade Balance and Level of Development in Selected Upper- and Lower-Middle-Income Countries, 2009 30 1.21 Composition of Services Exports by India and Turkey, 2001–09 31 1.22 Value of Services Exports in Malaysia and the United Kingdom, by Sector, 2000–10 32 1.23 Concentration and Diversification Indexes for Selected Countries, 2002 and 2010 33 B1.4.1 Herfindahl-Hirschman Index for Romania, by Sector, 2008 34 B1.4.2 Share of Output Produced by Top Five Firms in Romania, by Sector, 2008 34 1.24 Relationship between Theil Entropy Index and Level of Development, 2010 35 1.25 Revealed Comparative Advantage Index for Costa Rica, 2000, 2005, and 2009 36 B1.5.1 Export Sophistication and GDP in Selected Countries, 2009–11 38 1.26 Sophistication of Services Exports and Per Capita GDP in Selected Countries, 2001 and 2009 39 1.27 Sophistication of Services Exports by Six Countries in Latin America, 2001–09 39 1.28 Revealed Factor Intensities for Human and ICT-Related Capital in Selected Services Sectors in South Africa and Turkey, 2009 40 1.29 Intensive Margin of Turkey, 2002 and 2010 41 1.30 Extensive Margins of Mexico and the Slovak Republic, 1994–2010 42 1.31 Extensive and Intensive Margins of Mexico, the Slovak Republic, and Turkey, 2010 43 1.32 Trade Complementarities Index between Malaysia and Selected Trade Partners, 2004–10 44 1.33 Trade Intensity Index for Turkey and Selected Trade Partners, 2008 44 1.34 Gravity Model for Malaysia and Selected Trade Partners, 2008 45 1.35 Trading Partners and Development, 2008 45 2.1 Per Capita Services Value Added and Per Capita GDP in Selected Countries, 2000–02 and 2008–10 48 2.2 Services Value Added and Structural Factors in Selected Countries, 2000–02 and 2008–10 49 2.3 Per Capita Value Added of Agriculture, Industry, and Services versus Per Capita GDP, 2008–10 50 2.4 Sectoral Share of Value Added, by Income Group, 1990, 2000, and 2010 51 2.5 Value Added of Agriculture, Industry, and Services as Share of GDP in Central Asian Countries, 1990, 2000, and 2010 52 2.6 Value Added of Agriculture, Industry, and Services as Share of GDP in Selected Countries in the Middle East and North Africa, 1990, 2000, and 2010 52 2.7 Employment in Agriculture, Industry, and Services, by Income Group, 1994, 2000, and 2005 53 2.8 Employment in Agriculture, Industry, and Services in Selected Low- and Lower-Middle-Income Countries, 2006 53 2.9 Share of Value Added and Share of Employment in Selected Countries, by Sector, 2008–10 54 B2.2.1 Concentration of Employment, Value Added, Output, and Exports in Romania, by Sector, 2008 55 2.10 Knowledge and Services Exports Contribution to Total Exports, 2007 56 2.11 Gross Exports and Value Added of Exports of Machinery and Business Services, by Income Level, 2007 57 2.12 Ratio of Value Added to Gross Value and Development for Machinery and Business Services in Selected Economies, 2007 58 viii Contents 2.13 Forward and Backward Linkages for Machinery and Business Services, by Income Group, 2007 61 2.14 Backward and Forward Linkages for Goods and Services in Selected Countries, 2007 61 2.15 Direct Value Added plus Linkages/Gross Share and Per Capita GDP in Selected Economies, 2007 62 2.16 Direct Value Added plus Linkages/Gross Share for Transport and Trade Services in Selected Economies, 1992–2007 63 2.17 Relationship between Value-Added Linkages and Development in Selected Economies, 2007 64 2.18 Ratio of Exported Forward to Backward Linkages of Business Services in Other Sectors in Romania, Malaysia, and the United States, 2001 and 2007 67 2.19 Sectors Exporting Business Services Value Added as Share of Total Exported Business Services Value Added based on Forward Linkages in Mexico and the United States, 2007 68 B2.4.1 Exports and Value-Added Shares of Selected Services Subsectors in Romania 69 3.1 The Four Modes of Trade in Services 72 3.2 Number of Exporting Firms in Romania, by Sector, 2009 75 3.3 Tradable versus Nontradable Services Subsectors 78 B3.2.1 Number of Firms and Percentage of Exporters in Romania, by Sector, 2009 79 3.4 Services Trade, High-Skilled Labor Supply, and the Rule of Law in Selected Countries, 2009 82 3.5 Index of Supply of High-Skilled Labor, Legal Institutions, and Internet in Selected Countries, 2009 83 B3.5.1 Relative Productivity of Selected Export Sectors in Romania, 2009 84 3.6 Residuals and Fitted Values of Gravity Model, Estimated with GDP and Fixed Effects, 2008–09 86 3.7 Predicted Trade between Indonesia and Other Countries, Estimated with GDP and Fixed Effects, 2008–09 87 3.8 Actual Trade and Predicted Trade of Gravity Model, Estimated with Fixed Effects, 2008–09 87 4.1 Types of Market Failures 95 4.2 Options, Constraints, Levers, and Risks Associated with Regulating Natural Monopoly or Oligopoly 97 4.3 Options, Constraints, Levers, and Risks Associated with Regulating Exclusive or Controlled Access to Natural Resources 99 4.4 Options, Constraints, Levers, and Risks Associated with Regulating Public Goods 100 4.5 Options, Constraints, Levers, and Risks Associated with Addressing Health and Safety Concerns in the Presence of Asymmetric Information 101 4.6 Options, Constraints, Levers, and Risks Associated with Dealing with Asymmetric Information between Providers and Consumers of Financial Services 102 4.7 Options, Constraints, Levers, and Risks Associated with Promoting Technology Transfer 103 4.8 Complexity of Services versus the Rule of Law in Selected Countries, 2010 114 4.9 Services Trade Barriers versus Regulatory Quality in Selected Countries, 2008–10 115 4.10 Mode 3 Barriers versus Modes 1 and 2 Trade, 2008–10 117 4.11 Human Capital versus Services Exports in Selected Countries, 2010 119 4.12 Commercial Services Exports versus Management Score in Selected Countries, 2006–09 120 4.13 Value Added of Services as Percent of GDP versus Logistics Performance Index in Selected Countries, 2010 121 4.14 Internet Penetration versus Trade in Services in Selected Countries, 2010 122 Tables I.1 Share of Total Trade in Services, by Mode of Supply, 2005 2 I.2 Types and Examples of Trade Barriers 6 I.3 Domestic Enabling Factors 7 1.1 Results of Regression of Trade in Services against Development and Population 18 1.2 Growth of Exports of Goods and Services in Central Asia, by Country, 2000–10 23 1.3 Growth Rate of Services Exports by Selected Countries in Central Asia, South Asia, and the Caucasus Region, by Sector, 2000–10 27 1.4 Average Annual Growth of Services Exports, 1994–2007 and 2008–09 31 1.5 Value, Export Share, and Revealed Comparative Advantage Index of Services Exports by Argentina, 2002 and 2010 36 B1.5.1 Revealed Comparative Advantage Index of Selected Countries for Goods and Services Exports, 2009–11 37 1.6 Factor Intensities of Selected Services Exports 40 Contents ix 1.7 Regression Results of Future Growth between 1999 and 2008 on Revealed Human Capital Index (RHCI) and Revealed ICT-Related Capital Index (RICTI) 41 2.1 Cross-Country Regression of Size of Services Sector against Specialization in Services Export 48 2.2 Cross-Country Regression of Per Capita Services Value Added against Level of Economic Development 49 2.3 Value Added of Domestic Production and Exports in Albania, by Sector, 2007 59 B2.3.1 Value Added of Exports and Domestic Production in Ghana, by Sector, 2007 60 2.4 Domestic Services Content of Exported Goods and Services: Issues for Discussion 65 2.5 Impact of Broad Categories of Foreign Services Inputs on Manufacturing Exports 65 2.6 Impact of Disaggregated Foreign Services Inputs on Manufacturing Exports 66 B2.4.1 Unit Values of Direct Exports versus Exports through Intermediaries 69 3.1 Assessing Tradability by Modes: Issues for Discussion 72 3.2 Examples of Services Trade by Mode 73 3.3 Dominant Modes of Services Exports by Sector 73 3.4 Tradability of U.S. Services through Modes 1, 2, and 3, 2009 75 3.5 Assessing Tradability Based on Production: Issues for Discussion 76 3.6 Gini Classes 77 3.7 Share of Trade and Tradability Index by Services Sector in Malaysia and the United States, 2009 78 B3.2.1 Services Tradability Index for Romania, 2008 79 B3.2.2 Correlation between Sectoral Tradability and Firm Characteristics in Romania, 2009 80 3.8 Assessing Tradability Based on Comparative Advantage: Issues for Discussion 81 3.9 Rank of Factor Intensities for High-Skilled Labor, Information and Communications Technology (ICT) Capital, and Complexity 82 3.10 World Tradability Index for Selected Services Sectors 84 3.11 Regression Results of Gravity Model of Trade in Services, 2008–09 85 3A.1 Gini Class and Tradability of Selected Services Sectors 88 3A.2 Probability Results from Comparative Advantage Regressions 90 4.1 Instruments for Addressing Market Failures Affecting Trade in Services 96 4.2 Examples of Restriction on Trade in Services, by Mode 104 4.3 Policies Affecting Trade in Services 105 4.4 Typology of Barriers to Trade in Services 106 B4.7.1 Levels of Trade Openness in the World Bank’s Services Trade Restrictiveness Index (STRI) Database 107 4.5 Tariff Equivalents of Barriers to Services Trade in Selected Countries, 2009–11 111 4.6 Domestic Factors that Enable Trade in Services 112 A.1 Value Added in Final Output in Turkey, by Sectors, 2007 128 A.2 Economies in the Export of Value Added Database 129 A.3 Sectors in the Export of Value Added Database 131 A.4 Service Sectors Mapping to International Standard Industrial Classification (ISIC) 132 B.1 Variable Descriptions 136 B.2 Services Sector Components, by Balance of Payments (BOP) Code and Chapter 136 B.3 Economies Covered in the Trade in Services Database 139 Foreword Services trade today represents more than a fifth of global trade measured in volumes and 50 percent if measured in value addition. For some countries, including developing countries, its share in total exports is even bigger. Services are traded along with inputs into other economic activities, with both playing a key role in a country’s competitiveness and participation in global value chains. Assessing exactly how competitive these sectors are and how their performance can be improved is therefore critical in designing and implementing trade policies that can increase services’ contribution to economic development and growth. Valuing Services in Trade: A Toolkit for Competitiveness Diagnostics is a welcome con- tribution to the empirical analysis of countries’ services performance. Even before our era of “servicification,” the importance of services trade for the global economy and country welfare has been widely accepted since the early 1980s. The inclusion of services in the multilateral trading system, which created the World Trade Organization (WTO) in 1995, was a significant step forward. Opening services trade now had a platform. This became clear following the conclusion of the WTO financial services and telecommunication negotiations in 1997. These were important milestones. The role of services in the global trade agenda has been increasing ever since. Today, services trade has become one of the most important and complex subjects in regional trade negotiations, espe- cially in South-South agreements. Liberalization of services trade, however, has proven very difficult. A lack of data and a limited number of analytical tools for conducting meaningful, rigorous research are major constraints. This problem is particularly severe in some developing and least-developed countries where statistics and knowledge are limited. As Director General of the WTO, one of my key concerns was to help countries overcome these constraints. Through mutual efforts and close collaboration with World Bank and other trade experts, countries are coming to understand the benefits and challenges of a complex international trade agenda and the importance of a rules-based multilateral trading system. Our support here has been highly valuable for developing country members, but more can still be done. This book is a significant step forward in filling the existing knowledge and analytical gaps in two important dimensions of the trade in services debate: the role of services as a source of export diversification and its role as a source of competi- tiveness for the whole economy. It provides a clear, coherent, and innovative framework that policy makers, academics, experts, and anyone interested in trade policy matters will be able to use to analyze the services performance of developing countries. The book begins by providing broad indicators of the services sector, which situate the country in the global context. These indicators can be used to compare a country’s performance against its peers’, helping to identify potential issues that may explain any differences in performance. It then digs down into the microeconomic indicators, to provide an increasingly precise characterization of the ability and potentialities of a country to be competitive in the export of the services sector and in its ability to use the domestic services industry to enhance the competitiveness of other export sectors. The book also illustrates the positive contribution of firm-level analysis. The increasing use of firm-level data in service- related analytical work is giving a clearer picture of the characteristics of firms—the real protagonists engaged in services trade. This is helping, in particular, to better understand the size of firms, the skills of employees, the markets where firms export, and the constraints these firms face. Digging down to this level of analysis allows us to compare services firms with firms engaged in goods trade, again helping to identify relevant policy issues and those that could boost services’ contribu- tions to trade. Firm-level analysis is just one example of the refreshing use of new data found in this book. The fragmentation of pro- duction in different locations, within countries and abroad, is changing trade patterns. Global value chains, which connect these increasingly fragmented tasks, are creating complex data challenges. The authors recognize these challenges and have provided this book as a toolkit to meet them. The book draws on a new database built specifically to help countries assess xi xii Foreword the role of services in trade and other economic activities by using value-added data. This database covers a wide range of countries, including least-developed countries. It is in itself a valuable contribution to the empirical work on services trade. Policies applied by countries and trading partners are particularly important in explaining why performance differs. The toolkit provided in this book highlights the main policies affecting trade and their influences on performance. The authors distinguish three broad categories: regulatory policies, which primarily restrict trade; policies aimed at address- ing domestic market failures; and domestic enabling factors, including supply-side factors, such as skills availability, and services-related infrastructure, such as telecommunications and physical infrastructure. The authors successfully achieve a difficult balance. The user will find a thorough discussion of trade theory, a critical review of the relevant empirical literature, a clear explanation of the indicators used and their relevance, and highly infor- mative case studies where the toolkit has been used. This book is another important contribution from the World Bank to the empirical work on services trade. It will be of particular help to policy makers, trade experts, academics, and all of us interested in trade, growth, and development. Pascal Lamy Former Director General World Trade Organization Preface Increasingly, as we forge ahead in the 21st century, services will define the ability of countries and their firms to compete on international markets. This toolkit is an effort of the World Bank Group’s new Trade and Competitiveness Global Practice to address the needs of countries facing this rising trend, and to help our clients make informed policy choices to better their chances of benefiting from the increasing prominence of services in international trade. There are two aspects to services’ rise. First, thanks to improvements in information and communications technology, services have become more and more tradable; service providers, such as computer programmers or call-center operators, no longer need to be close to their customers. They can operate from a location that provides skilled labor at low costs. But there’s a second component. Not only are more services crossing borders, domestically produced services are also emerging as vital inputs to the production of traded goods and services. Neither a factory nor a call center can operate without reliable electricity and water, for example. And as more and more production enters the just-in-time value chain model, crossing multiple countries in the transition from raw material to finished good, reliable services can mean the dif- ference between a firm’s profit and loss. The level of services can also make the case for a firm’s investment in one country over another. This phenomenon, which touches middle-income as well as low-income countries, has important implications for development and for the World Bank Group’s twin goals of eradicating poverty and improving the incomes of the bottom 40 percent of populations everywhere. Already, many developing countries export a diverse range of services. They might provide transport services to their neighbors and back-office business services to developing countries, for example. The falling costs of information technology and increasing access to the internet suggest that this trend will only grow, offering the possibility that more and more jobs in the developing world will be in the services sector. Just as they want to see their exported services thrive, policy makers in many developing countries would like to do more to ensure that all of their industries—producing both goods and services—can be competitive on the world market. Growing and hiring domestic industries, as well as foreign investment, are key components of many countries’ strategies for economic growth, job creation, and, ultimately, poverty reduction. This toolkit is designed to help countries assess their past performance in services as well as areas in which they have comparative advantage. It allows policy makers to improve their understanding of the size, scope, and potential of services production and exports in their countries. It also allows them to identify obstacles to the competitiveness of their services sectors. It is a key pillar of our efforts in the World Bank’s Trade and Competitiveness Global Practice to serve our clients and help them have cutting-edge tools in an ever-changing global economy. Anabel Gonzalez Senior Director Trade and Competitiveness Global Practice The World Bank xiii Acknowledgments Valuing Services in Trade was prepared by the World Bank’s International Trade Unit in Poverty Reduction and Economic Management Network. Sebastián Sáez and Daria Taglioni were team coleaders. Erik van der Marel, Claire H. Hollweg, and Veronika Zavacka (consultants) provided invaluable inputs to the report. The authors would like to thank Michael Engman and Ana Margarida Fernandes, who peer-reviewed the concept note and provided useful comments and suggestions that greatly improved the original idea. The authors are especially grateful to Enrique Aldaz-Carrol, Michael Ferrantino, and Bernard Hoekman, who peer-reviewed the first draft and provided valuable suggestions. Sven Blank, Dale Honeck, Suhail Kassim, and Ivan Rossignol also provided comments on the first draft. The team would also like to thank Joseph Francois, Miriam Manchin, Olga Pindyuk, and Patrick Tomberger. They compiled the two new cross-country datasets widely used for the applications of the toolkit based on their research on services trade and trade in value added. The team also thanks the staff and consultants of the International Trade Unit who contributed and supported the work at different stages of the project, in particular, Tom Farole, Charles Kunaka, Martin Molinuevo, José-Daniel Reyes, and Jose Guilherme Reis for their discussions and suggestions on how best to tackle the project. During the development and piloting of parts of the toolkit in different countries and regions across the World Bank, we benefited from the support, comments, and suggestions of Ahmed Ahsan, Julian Latimer Clarke, Sebastian Eckardt, David Gould, Sibel Kulaksiz, Dorsati Madani, Hannah Messerli, Richard Record, Frederico Gil Sander, Rashmi Shankar, John Speakman, Ekaterine Vashakmadze, and Ekaterina Vostroknutova. The team acknowledges the collaboration of experts and consultants who helped to shape the analytical framework and messages of the toolkit, in particular, Guillermo Arenas, Matej Bajgar, Christina Busch, Jose Fernandez, Arti Grover, Beata Javorcick, Carolina Lennon, Saurabh Mishra, Anasuya Raj, Ben Shepherd, and Deborah Winkler. Shienny S. Lie provided excellent administrative assistance and helped to format the volume. Amir Alexander Fouad pro- vided support during the toolkit’s preparation and coordinates the International Trade Unit’s publication program. We also thank Stephen McGroarty, Andrés Meneses, Paola Scalabrin, and Janice Tuten of the World Bank’s Publishing and Knowledge Division for their valuable advice and guidance during the publication. The toolkit benefited from the pro- fessional editorial work of Barbara Karni. This project was supported in part by the governments of Finland, Norway, Sweden, and the United Kingdom through the Multidonor Trust Fund for Trade and Development. Finally, the project and the book would not have been possible without the continuous support and guidance of the staff of the International Trade Unit and its management team, in particular Mona Haddad (Sector Manager) and Bernard Hoekman (former Director), and Jeff Lewis (Director). xv About the Authors Sebastián Sáez is a senior trade economist in the International Trade Unit at the World Bank, where he has been respon- sible for the trade in services agenda since 2009. From 1990 to 1994, he was involved in the General Agreement on Tariffs and Trade’s Uruguay Round negotiations as adviser to the Chilean Minister of Finance. He subsequently served as deputy permanent representative in the Chilean Mission to the World Trade Organization (WTO) and as head of the Free Trade Agreement of the Americas Department in the Chilean Ministry of Foreign Affairs. Beginning in 2001, he served as head of the Chilean Department of Foreign Trade, Ministry of Economy, where he participated in trade negotiations with the European Union, Korea, and the United States. Prior to joining the World Bank, he worked in the International Trade and Integration Division at the United Nations Economic Commission for Latin America and the Caribbean (UN–ECLAC). Claire H. Hollweg is a consultant at the International Trade Unit at the World Bank. She holds a PhD and MA in econom- ics from the University of Adelaide. Prior to studying economics, she worked as a journalist in newspaper and radio, and holds a BS in journalism from the University of Colorado at Boulder. She also has work experience with the Government of South Australia and the Pacific Economic Cooperation Council (PECC) in Singapore. Her research interests include development economics with a recent focus on trade and labor, trade in services, and household responses to shocks. Daria Taglioni is a senior economist in the International Trade Unit at the World Bank. She has more than 10 years of experience in economic policy analysis on issues of international competitiveness, globalization, and the links between financial markets and trade. Prior to joining the World Bank, she worked at the European Central Bank and at the Organ- isation for Economic Co-operation and Development (OECD). She holds a PhD in international economics from the Graduate Institute for International Studies in Geneva. Erik van der Marel is a senior economist at the European Centre for International Political Economy (ECIPE). His main areas of expertise are in services trade, the political economy of services trade policy, the link between total factor produc- tivity (TFP) and regulation, and trade policy in developing countries. Prior to his appointment at ECIPE, he held posts as a lecturer at the London School of Economics and as a research fellow at the Groupe d’Économie Mondiale (GEM) insti- tute in Paris at Sciences-Po, and was a consultant at the Asia-Pacific Economic Cooperation (APEC) forum, the European Commission, the OECD, and the World Bank. He holds a PhD in international economics from Sciences-Po in Paris. Veronika Zavacka is a board operations officer at the International Monetary Fund (IMF). The focus of her research is on the behavior of international trade under exceptional circumstances, such as financial crises and uncertainty shocks. Prior to coming to the IMF she worked for several other international institutions, including the World Bank, the European Bank for Reconstruction and Development, European Central Bank, and United Nations Development Programme (UNDP). She holds a doctorate in international economics from the Graduate Institute for International Studies in Geneva. xvii ABBREVIATIONS ASEAN Association of Southeast Asian Nations BPO business process outsourcing BRIC Brazil, Russia, India, China EBOPS Extended Balance of Payments Services EXPY (revealed) income content of export basket FATS Foreign Affiliates Trade Statistics FDI foreign direct investment GATS General Agreement on Trade in Services GDP gross domestic product GTAP Global Trade Analysis Project GVC global value chain ICT information and communication technology IMF International Monetary Fund IT information technology ITES information technology–enabled services KAM Knowledge Assessment Methodology NACE Nomenclature statistique des activités économiques dans la Communauté européenne (Statistical Classification of Economic Activities in the European Community) NAICS North American Industry Classification System OECD Organisation for Economic Co-operation and Development PPP purchasing power parity PRODY (revealed) income content of product basket RCA revealed comparative advantage RHCI Revealed Human Capital Index RICTI Revealed Information and Communications Technology–Related Capital Index STCD Services Trade Competitiveness Diagnostic STRI Services Trade Restrictiveness Index TCI Trade Complementarities Index TFP total factor productivity TII Trade Intensity Index TiVA Trade in Value Added UNCTAD United Nations Conference on Trade and Development WDI World Development Indicators Note: All dollar amounts are U.S. dollars ($) unless otherwise indicated. xix Balance of Payment (BoP) Codes Used in This Book The balance of payments coding system, available at http://www.imf.org/external/np/sta/bopcode/, has been devel- oped by the International Monetary Fund in cooperation with the Statistical Office of the European Union (Eurostat), the Organisation for Economic Co-operation and Development (OECD), and the European Central Bank (ECB). The coding system is designed to facilitate the exchange of data on balance of payments, international investment position, international trade in services, and foreign currency liquidity among these organizations, their member states, and other interested organizations or entities. Total services 200 5.3 Other direct insurance 256 1. Transportation 205 5.4 Reinsurance 257 1.1 Sea transport 206 5.5 Auxiliary services 258 1.2 Air transport 210 6. Financial services 260 1.3 Other transport 214 7. Computer and information services 262 1.4 Space transport 218 7.1 Computer services 263 1.5 Rail transport 219 7.2 Information services 264 1.6 Road transport 223 8. Royalties and license fees 266 1.7 Inland waterway transport 227 8.1 Franchises and similar rights 891 1.8 Pipeline transport and electricity 231 8.2 Other royalties and license fees 892 transmission 9. Other business services 268 1.9 Other supporting and auxiliary 232 9.1 Merchanting and other trade-related 269 transport services services 2. Travel 236 9.2 Operational leasing services 272 2.1 Business travel 237 9.3 Miscellaneous business, professional, 273 2.2 Personal travel 240 and technical services 3. Communications services 245 10. Personal, cultural, and recreational 287 services 3.1 Postal and courier services 246 10.1 Audiovisual and related services 288 3.2 Telecommunications services 247 10.2 Other personal, cultural, and 289 4. Construction services 249 recreational services 4.1 Construction abroad 250 11. Government services, not included 291 4.2 Construction in the compiling 251 elsewhere economy 11.1 Embassies and consulates 292 5. Insurance services 253 11.2 Military units and agencies 293 5.1 Life insurance and pension funding 254 11.3 Other government services 294 5.2 Freight insurance 255 xxi Glossary asymmetric information. Situation in which one party to a transaction has better information than another, allowing it to exploit the other party’s incomplete knowledge. backward linkage. A “pull” impact of an increase in final demand in a sector on input suppliers to that sector in terms of value added; reveals demand response of a sector from all other (upstream) sectors’ supply of inputs. business process outsourcing (BPO). Contractual agreement of a service to completely manage, deliver, and operate one or more (typically IT–intensive) business processes or functions. computable general equilibrium model. Model that captures interactions among markets, production, and consumption and which allows estimation of welfare and resource allocation effects resulting from changes in trade policy. direct method. Assessment of the impact of service regulation on economic performance such as prices and costs on the basis of real information on restrictive policies. domestic enabling factors. Policies related to the domestic economic and institutional situation within a country, such as institutions, governance, the business environment, labor skills, management and entrepreneurial skills, and trade-related infrastructure. dyadic gravity equation. Specification for a gravity equation in which the economic mass variable is picked up by importer and exporter fixed effects. This specification controls for all country-specific factors that affect bilateral trade flows and corrects for unobservable omitted variables that could be present in the error term of a specification without fixed effects, potentially biasing the point estimates. EU KLEMS. Project, funded by the European Commission between 2003 and 2008, that created a database of measures of economic growth, productivity, employment, capital formation, and technological change at the industry level for all member states of the European Union and several other OECD countries from 1970 onward. Export of Value Added Database. Database of gross and net production and export figures constructed by the World Bank. Data come from the Global Trade Analysis Project (GTAP). export sophistication/EXPY. Measure of the income content of a country’s export basket. Captures whether a country’s export basket consists primarily of services typically exported by high-income or low-income economies. extensive margin. Indicator that assesses whether exports are growing because a country is exporting new goods and services or reaching new markets rather than exporting more of the same goods and services to the same markets. Foreign Affiliates Trade Statistics (FATS). Indicators of Mode 3 trade in services. Measures the extent to which multina- tionals sell their service through an affiliate established in a foreign country. forward linkage. Measure of the contribution of a sector as an input in terms of value added in other sectors; reveals supply response of a sector to all other (downstream) sectors’ demand for more inputs. free-rider problem. Situation in which some individuals or firms either consume more than their fair share or pay less than their fair share of the cost of a common resource. Free-riding can lead to the underproduction of a public good or excessive use of a common property resource. xxiii xxiv Glossary Gini coefficient. Measure of income distribution or industrial concentration. global value chain. The full range of activities executed through different stages of production, which are located in various geographical locations, that bring a product or service from conception to delivery to final consumers and disposal after use. gravity equation. Equation that relates countries’ bilateral trade flows to structural determinants such as GDP, geographic distance, and other factors that affect trade costs. gravity model. Model for predicting the level of interaction between two countries. Herfindahl-Hirschman Index. Standard concentration index. Measures the size of exports of each sector in relation to each other within a geographical unit. horizontal measures. Measures that affect services providers across the board. These include measures such as conditions on making foreign investments, regulations on accessing foreign currency, and laws on the entry and stay of individual travelers that affect (services) providers across the board. indirect method. Assessment of the level of openness, restrictiveness, or contestability of a market based on comparison against a benchmark. Often such assessment is executed by lack of any direct method. intensive margin. Indicator that assesses whether exports are growing because a country is exporting more of the same goods and services to the same market rather than exporting new goods and services or reaching new markets. inward FATS. Indicator of operations of foreign-owned firms in the host economy under Mode 3 trade in services. joint venture. Company whose equity is owned by both a foreign and a domestic firm; form of business organization often required before a foreign firm is allowed to operate in a country. jointness in production. Requirement for proximity of the consumer and the producer. Knowledge Assessment Methodology (KAM). Interactive benchmarking tool created by the Knowledge for Development Program of the World Bank to help countries identify the challenges and opportunities they face in making the transition to the knowledge economy. knowledge economy. Economy in which knowledge is the main engine of economic growth. knowledge process outsourcing (KPO). Contractual agreement to completely manage, deliver, and operate one or more (typically IT–intensive) business processes or functions that are knowledge intensive, such as business consulting, business analytics, market intelligence, and legal services. ladder of comparative advantage. Stages of development in which countries move from an economy of production and trade largely based on agriculture, to one based on industry, and finally services following relative prices and adoption of new technologies. leapfrogging. Skipping of one or more of the sequential stages in a process. moral hazard. Tendency of a person to be more willing to take a risk if he or she knows that the potential costs of doing so will be borne, in whole or in part, by others. national treatment. Nondiscriminatory treatment of nationals, foreigners, investors, goods, and services. natural monopoly. Situation in which the long-run average cost of production decreases with output, implying that costs are lowest when a single firm produces a good or service, making participation by more than one firm unviable. Natural monopolies exist where investment costs are very high, creating huge economies of scale and high barriers to entry. Examples include public utilities. network externalities. Situation in which access to essential facilities is a critical ingredient and the overall benefit for the economy depends on how many other individuals or firms also have access to the facilities or knowledge provided. Sectors characterized by network externalities include telecommunications, transport and logistics services, and financial services. Glossary xxv outward FATS. Indicator of operations of affiliates of domestic firms established abroad under Mode 3 trade in services. positive externality. Benefit created for one party by an activity performed by another. Examples include the effect of one person’s education on others. PRODY. Measure of the income content of a product or service; income level associated with a product or service. prudential regulations. Regulations aimed at achieving a legitimate policy objective such as consumer protection without being protectionist. In the case of the financial services sector, they ensure financial soundness of licensed financial institu- tions to prevent financial system instability and losses of deposits by unsophisticated investors. Examples include capital adequacy requirements, liquidity requirements, and loan-loss provisioning mandates. public good. Good or service that is both nonexcludable (individuals cannot be effectively excluded from use) and non- rival (use by one individual does not reduce the availability of the good to others). Examples include national defense and lighthouses. public-private partnership. Private sector business venture that provides a service traditionally provided by the government and is funded and operated through a partnership of government and one or more private sector companies. Revealed Comparative Advantage (RCA) Index. Measure that compares the share of a sector’s exports of a country in the country’s total exports with the share of exports of all countries in that sector in the world’s total exports for all sectors. revealed factor intensity. Measure of intensities of factors, such as human and physical capital, in a sector. Revealed Human Capital Index (RHCI). Adjusted version of the RCA Index multiplied by a country’s human capital stock. Revealed Information and Communications Technology–Related Capital Index (RICTI). Adjusted version of the RCA Index multiplied by a country’s ICT-related capital stock. Revealed Physical Capital Index (RPCI). Adjusted version of the RCA Index multiplied by a country’s physical capital stock. Theil Entropy Index. Standard diversification measure. Measures the size of exports of each sector in relation to each other within a geographical unit. Tradability Index. Measure of the probability of trade taking place in a particular sector between partner countries. Trade Complementarities Index (TCI). Index indicating extent to which a country’s export profile is similar to a partner country’s import profile. Trade in Services Database. Data set of bilateral services flows aggregated by the World Bank. Data come from the Organisation for Economic Co-operation and Development (OECD), Eurostat, the United Nations, and the International Monetary Fund (IMF). trade-in-tasks paradigm. Paradigm in which trade patterns consist of flows of goods, services, investment, and know-how necessary to produce goods and services fragmented in multiple tasks located in different countries. Trade Intensity Index (TII). Index indicating the extent to which a country is present in the export market of a trading partner relative to other trade partner countries. World Development Indicators (WDI). World Bank’s primary collection of development indicators, compiled from officially recognized international sources. Indicators, which include national, regional, and global estimates, are the most current and accurate global development data available. Introduction Globalization continues to create opportunities, in par- What’s Special about Trade in Services? ticular for competitive developing countries. But it also Trade of goods and trade of services differ in several presents new challenges. Because the global economy is important ways (Francois and Hoekman 2010; Copeland changing rapidly, countries need to continuously reposi- and Mattoo 2008). Goods trade involves shipping tion themselves on the global trade and production map. goods from one country to another, whereas cross- border trade in services (known as Mode 1) is only one of the modes by which services are traded. Trade in ser- The Role of Services in Trade and vices also occurs when the consumer crosses the border Competitiveness: Key Issues and Debates (Mode 2). Trade also occurs when a producer estab- In an ever-more globalizing and competitive world, it lishes a commercial presence in the consumer’s country is important for countries to understand how they fare (Mode 3) or the producer temporarily moves across relative to competitors and to their own past export borders to provide the service (Mode 4). In services performance. trade, then, the international movement of factors of Trade indicators have been used extensively as a mea- production—capital, in the form of foreign direct invest- sure of competitiveness. Balassa (1965) used them to ment (FDI), and labor, in the form of temporary labor measure revealed comparative advantage. Hatsopoulos, mobility—are critical. Krugman, and Summers (1988) and Markusen (1992) tied How important are the different modes of supply? In trade balances to rising real income. Sharpe (1985, 1986), 2005, the most important mode of supply was commer- Krugman and Hatsopoulos (1987), Fagerberg (1988), and cial presence (Mode 3), suggesting that firms serve foreign Mandeng (1991) used them to estimate market shares or markets primarily through in-house offshored services increases in market share. activities (table I.1). In addition to the need for face-to-face Competition can also be measured by relative prices. interaction between the producer and the consumer of In competitive economies, equilibrium factor prices are many services, the fact that services often have a very high lower than those of international competitors. The real level of intellectual property content (and rights) may exchange rate and the real effective exchange rate are mea- explain this preference. sures of competitiveness based on relative prices. Durand Modes of supply may be either substitutes or comple- and Giorno (1987), Helleiner (1989), and Lipshitz and ments. When modes are perfect substitutes, liberalization McDonald (1991) use them to measure competitiveness. of one of them should be sufficient to fully reap the gains of Other studies (Buckley, Pass, and Prescott 1988; Porter liberalization. When modes are imperfect substitutes, their 1990; WEF 2004) use multidimensional definitions of competitiveness normally requires more than one set of competitiveness. Although these measures are unsuitable reforms and a framework that takes into account the mix for deriving robust quantitative measures, they have the of cross-border and locally supplied services. advantage of capturing several aspects of competitiveness. The mix of reforms needed is likely to change over Services are an important component of competitive- time and to be location specific. The evolving patterns ness in two respects. First, a country with a comparative of technology underpinning services trade and the new advantage in tradable services can export them, thereby possibilities they open require recognition of the fact diversifying the economy and expanding the country’s that policies need to adapt. A new technology that allows exports. Second, services can be a strategic driver of com- the cross-border provision of a wide range of profes- petitiveness for the whole economy. sional services, for example, makes nationality or local 1 2 Valuing Services in Trade Table I.1. Share of Total Trade in Services, by Mode of particularly in sectors with significant network externalities. Supply, 2005 Problems of imperfect and asymmetric information arise if Share of total trade in buyers and service providers face difficulties assessing the Mode Description services (percent) standing of providers or consumers. Regulation and infra- 1 Cross-border trade 25–30 2 Consumption abroad 10–15 structure, including competition laws and independent 3 Commercial presence 55–60 regulatory bodies, are necessary in sectors in which mar- 4 Presence of natural persons Less than 5 ket failures are prevalent. Regulating in a way that does not Source: Magdeleine and Maurer 2008. unnecessarily restrict trade is therefore complex. Services as a Source of Export Diversification presence requirements increasingly irrelevant and ineffec- tive (Francois and Hoekman 2010). Many developing countries export a diversified range of For many services (and goods), the relationship between services, both within their own regions and to high-income investment and trade is complex. For some services, there countries. Although services are intrinsically less tradable is strong complementarity: investment facilitates trade than most goods (because they typically require the face- (Helpman 2011). The empirical literature on services to-face presence of buyer and seller), the falling costs of indicates a robust complementary relationship between travel, communications, and information technology and trade and FDI that has increased over time. The sign and increasing access to the Internet are making it easier to pro- magnitude of the relationship differs across categories. duce a service in one location and consume it in another, The categories for which a significant complementary rela- thus creating new opportunities for trade in services. tionship have been found—banking; insurance; business, As a result, exports of services have increased for all professional, technical, and legal services (Lennon 2009)— country income groups and regions (figure I.1). Developing account for the bulk of services exports. These services are countries’ share of world services exports increased from also the most dynamic in terms of export growth. 18 percent in 1990 to 30 percent in 2012.1 India’s exports Services provision also requires complementary inputs of software and business process services accounted for (financial services, for instance, require professional and about 33 percent of its total services exports. In Latin telecommunication services). In order to export, it is thus America, Brazil, Costa Rica, and Uruguay export profes- not sufficient to have an advantage in a particular service sional and information technology–related services; Mexico if other required inputs are not in place. Examples of this exports communication and distribution services; and interdependency abound: logistics services depend on the Chile exports distribution and transportation services. interaction among transportation, freight-forwarding ser- Exports of professional services to Europe by Kenya, South vices, warehousing, and cargo handling (Kunaka, Antoci, Africa, and Tunisia are growing. The Philippines, and Sáez 2013). The success of tourism depends on the Singapore, and Thailand are exporting medical services, provision of other services and physical infrastructure, and Malaysia is increasing its exports of education, Islamic such as airports and roads. banking, and telecommunication services. Many services require the proximity of the consumer The increasing tradability of services is one the most and the producer—what is known as jointness in produc- important changes in trade patterns of the last quar- tion. Customers often provide information to producers ter of the 20th century. The tradability of services is regarding their needs that is critical for effective delivery of driven by changes in technology and regulatory reforms. the service (Hoekman and Kostecki 2009). Improvements in information and communication tech- Jointness has important implications for policy design. nologies (ICT) have reduced—and for a range of services For example, regulatory regimes are important determinants eliminated—the need for proximity between consumer of the feasibility of trade in services. Cross-border trade and producer. They have done so by increasing storabil- restrictions may diminish FDI incentives, and restrictions ity and allowing greater use of outsourcing and offshoring on the operations of foreign firms may limit cross-border in the context of global and regional production chains. trade in services. Cross-border service exports are thus Breakthrough reforms that increased the tradability of linked to inward FDI flows. FDI policies are key for promot- services include the elimination of monopoly provid- ing exports of services (Francois and Hoekman 2010). ers and restrictions that affected foreign investors. These Service industries are usually characterized by mar- changes have allowed the private sector and foreign inves- ket failures, such as imperfect and asymmetric informa- tors to provide services traditionally provided by the state tion, lack of competition, and natural barriers to entry, (Francois and Hoekman 2010). The Role of Services in Trade and Competitiveness: Key Issues and Debates 3 Figure I.1. Services Trade as Share of GDP, by Region comparative advantage in services trade and the capacity to deal with the complex nature of trade liberalization, a. Services trade as share of GDP in developed deregulation, and the required policy reforms. Since the and developing countries, 1980–2012 14 mid-1990s, this perception has slowly changed. Policy makers now ask for guidance on identifying areas of 13 Services trade as percent of GDP potential comparative advantage over peer countries in 12 the services sector. 11 10 Services as a Source of Competitiveness to the 9 Whole Economy 8 Services play a strategic role in countries’ competitiveness agenda not only because they are a source of export diver- 7 sification but also because they are used in the production 6 of many downstream competitive products. The competi- 5 tiveness of most goods exports depends not only on access 80 82 84 86 88 90 92 94 96 98 00 02 04 06 08 10 12 to raw material inputs but also on critical services inputs, 19 19 19 19 19 19 19 19 19 19 20 20 20 20 20 20 20 including efficient, competitively priced utilities, financial Developed countries Developing economies services, and other commercial services. Because of the b. Services trade as share of GDP by region, 2011 importance of services to the economy as a whole, more 20 and more governments want to improve their understand- 18 ing of the size, scope, and potential of services produc- tion and exports and to identify obstacles that need to be Services trade as percent of GDP 16 removed to unlock the competitiveness of their services 14 sectors. 12 Increasing the competitiveness of the services sec- 10 tor is important, particularly given internationally frag- mented models of production. Trade in the 20th century 8 was mainly about selling goods to customers in another 6 country. In the 21st century, trade is increasingly about 4 flows of the goods, services, ideas, investment, training, know-how, and intellectual property needed to produce 2 goods and services in multiple locations (Baldwin 2011, 0 2012; Jones 2000; Grossman and Rossi-Hansberg 2008; ica a ia cifi d be nd As ral ric As Feenstra 2010; Helpman 2011). Pa an er nt c Af ia rib a a an Am h ia Ce an ut Ca ic As Global value chains internationalize the nexus e er So d ar rth st an th Am h Ea No Sa of flows, giving rise to the trade-investment-service- pe tin b- ro Su La intellectual property nexus. Global value chains can Eu be thought of as factories that cross international bor- Source: United Nations Conference on Trade and Development (UNCTAD) ders. Traditional factories clustered most stages of and World Development Indicators. production into a single building or industrial district. Coordinating complex production processes involves continuous, two-way flows between goods, people, As the number of champions of services exports from ideas, investment, training, know-how, and other middle- and lower-income country groups has grown, inputs. Factories traditionally minimized the cost of policy makers increasingly recognize the unexploited such flows by keeping all stages of production as close to potential of services trade. Developing countries were one another as possible. initially reluctant to embrace the services trade agenda. The ICT revolution of the 1990s reduced the cost of During negotiation of the Uruguay Round of the General coordinating complexity over long distances, allowing Agreement on Tariffs and Trade (GATT), the conven- production stages that previously had to be within walk- tional wisdom was that developing countries lacked a ing distance to be dispersed internationally. Some stages 4 Valuing Services in Trade of production were offshored to lower-wage countries, the organization of the domestic segment of the value giving rise to global value chains, transforming the nature chain is as important as the international one. of international commerce. The “servicification” of manufacturing is apparent from Global value chains did not end the need for the intra- figure I.2, which shows that about 40 types of services are stage flows of goods, people, ideas, investment, and know- involved when a manufacturing firm internationalizes its how; they internationalized them. Flows that used to take production. This example indicates that even in the manu- place only within factories in high-income countries are facturing sector, a country cannot be competitive and join now a key part of international commerce. global value chains if it does not have efficient domestic ser- Services trade coordinates the dispersed production. vices or is closed to the imports of such services. Managing Thanks to the ICT revolution, production stages can be the complexity of the chain and preserving the production coordinated at a distance, allowing production stages to standards throughout requires coordination of efficient be dispersed geographically and to benefit from price, cost, services and the movement of key personnel across borders. and wage differentials. Recent data suggest that services represent more than Easy access to efficient services and infrastructure in 30 percent of the value added of trade in manufactures the host country—inexpensive and reliable power, reli- (figure I.3). able telecommunications and transport, finance and trade The central role of services becomes more appar- support—is critical to domestic and particularly foreign ent when one looks at trade through the lens of the new investors. Among firms surveyed, 52 percent cited access trade-in-tasks paradigm proposed by Grossman and to finance and 39 percent cited transport infrastruc- Rossi-Hansberg (2008). As figure I.4 shows, firms’ location ture and services as the most serious supply-side con- decisions are task specific. Country strategies should move straint among global value chain suppliers in developing away from paradigms in which development means evolv- countries (OECD–WTO 2013). For all types of services, ing in terms of sectors, focusing instead on the tasks they Figure I.2. Services Involved in the Internationalization of Production Source: Swedish National Board of Trade (Kommerskollegium) 2010. Note: Figure displays production by Sandvik Tooling, a Swedish producer of construction equipment and tools. The Role of Regulation 5 Figure I.3. Value Added of Services as Share of Value Added of Manufacturing, by Sector, 2009 40 Value added of services as percent of 35 value added of manufacturing 30 25 20 15 10 5 0 Mining Machinery, Textiles Food Transport Chemicals equipment products equipment Distribution and repair Transport and storage Finance Business services Other services Source: OECD-WTO 2013. Note: Distribution and repair does not include distribution of final goods. Figure I.4. From Sector-Based to Task-Based Development Strategies a. Old paradigm: From low- to high- b. New paradigm: From low- to high- value added sectors value added activities Services High-value added support activities Sales (e.g. design, (e.g. preproduction Manufacturing commercialization) research and development, Commodities Primary activities (e.g. postsale services, operations, assembly) technological development, after sale specialized services) Source: Adapted from Cattaneo and Miroudot 2013. can perform within sectors, which may serve as inputs to and development); technological development; and activi- multiple sectors. ties aimed at organizing the firm’s infrastructure, human The new paradigm brings to the forefront the fact that the resources management, and procurement. range of activities necessary to 21st century manufacturing is very broad and includes many services. The classical literature The Role of Regulation on value chains (Porter 1985) distinguishes between primary, support, and sales activities. Primary activities range from Where the domestic services sector is weak, access to manufacturing of inputs, outputs, and assembly operations imports of services (through all modes of production) to inbound and outbound logistics, marketing, sales, and a must be liberalized in order to increase access to efficient, range of other service activities. Support activities include world-class services. Liberalization allows an economy to the production of other inputs (machinery and equip- increase both the services intensity and the value addition ment as well as many services activities, including research of the goods and services it produces and exports. 6 Valuing Services in Trade The empirical literature finds a positive relationship exemptions; and other forms of trade-promotion activ- between productivity in services and productivity in the ities, such as trade fairs and information dissemination. overall economy. Within the Organisation for Economic • Domestic enabling factors include a country’s endow- Co-operation and Development (OECD), productivity ments of factors, especially of human capital, including growth in services has been a key driver of growth, and skills and entrepreneurial ability; natural resources and general productivity growth can be traced to market and cultural endowments that attract tourists; infrastructure, business services (Hoekman and Kostecki 2009; Inklaar, especially telecommunications networks, which facili- Timmer, and Ark 2007, 2008; Triplett and Bosworth 2004). tate the delivery of services; and institutional quality, Services imports serve as a transmission channel for the especially the regulatory environment for services. transfer of new technologies. These imports both improve the performance of services exports in skill-intensive Regulatory Policy and Trade Barriers industries and increase the value added of manufacturing exports (Francois and Woerz 2008; for the positive link Regulation, including competition laws and independent between trade liberalization of services sectors and manu- regulatory bodies, is necessary in sectors in which mar- facturing productivity, see Arnold, Mattoo, and Narciso ket failures are prevalent. Such regulation is often imple- 2008; Arnold and others 2010; and Arnold, Javorcik, and mented to deal with limited competition, asymmetric Mattoo 2011). Services trade plays a key role in increasing information, and differences between private and social the productivity of services sectors and has a strong impact utility (externalities and public goods). Putting in place reg- on overall productivity and economic growth over time ulations that do not unnecessarily restrict trade is difficult. (van der Marel 2012). A wide set of limitations affect imports and exports of For these reasons, services matters to a country’s com- services explicitly. These measures can be discriminatory petitiveness, and investment barriers to services matter to or nondiscriminatory in intent and direct or indirect. They trade. Lack of clarity and predictability in domestic laws or are likely to pertain to market access, national treatment, deficiencies in domestic services may reduce productivity domestic regulations, and other regulations (table I.2). and the attractiveness of a country to foreign investors. Following Goswami, Mattoo, and Sáez (2012), one can Domestic Enabling Factors distinguish three broad types of policies affecting services competitiveness: Good regulatory policies stand to gain from other measures a government can implement. These policies are related to • Regulatory policy affecting trade, investment, and labor the domestic economic and institutional situation inside mobility in services includes all policies affecting cross- the country (fundamentals) rather than the explicit poli- border trade and the domestic provision of services.2 cies in place. Horizontal measures complement specific These policies do not explicitly target trade in services trade and domestic policies. but nevertheless affect it. Domestic enabling factors are broad in scope. They • Trade barriers that target trade, investment, and labor include institutions, governance, business environment, mobility in services include policies designed to pro- labor skills, management and entrepreneurial skills, and mote exports and investment. They include the creation trade-related infrastructure (table I.3). of special economic zones and privileged access to land, A shared language, cultural history, or colonial his- infrastructure, and imported inputs; fiscal incentives for tory is likely to increase services exports, because exporters and investors, in the form of subsidies or tax such factors make it easier to establish business ties. Table I.2. Types and Examples of Trade Barriers Type of barrier Example Nature of effect Market access Measures, usually quantitative, that make it difficult for foreign firms to enter the Direct domestic market National treatment Legal (de jure) restrictions against foreign firms Direct Discrimination through nondiscriminatory barriers (de facto) Indirect Domestic regulation Regulations that do not explicitly target foreign firms or restrict trade but may constitute de Indirect facto trade barriers if they are more burdensome than necessary to achieve a legitimate policy objective. These measures include both nondiscriminatory entry regulations and regulations affecting ongoing operations. The Role of Regulation 7 Table I.3. Domestic Enabling Factors Policy area Objective Labor skills Increase capacity to produce sophisticated services exports; climb up ladder of comparative advantage in services (or from goods to services), especially in professional, computer and computer-related, and business services. Management and Enhance adoption and use of modern technologies that are essential for producing a service or good. entrepreneurial skills Trade-related infrastructure Reduce costs related to delivery of services (transportation, telecommunications, export, transactions, and search costs). Institutions Establish governance arrangements to foster relationships between private parties rather than between private parties and the government (Acemoglu and Johnson 2005). Governance Increase ability of governments to formulate and implement sound policies and regulations that allow and promote private sector development. Business environment Attract foreign direct investment and multinational corporations in order to expand exports and increase domestic competitiveness. These characteristics cannot be changed. By contrast, governance of service regulations. A Regulatory Assessment institutions, governance, the business environment, labor of Services Trade and Investment (RASTI) can help policy skills, management and entrepreneurial skills, and trade- makers assess regulations consistently, streamline the regu- related infrastructure can be enhanced, albeit only in the latory framework, and set up a process for introducing new medium to long term. regulations. A RASTI provides the information required Institutions reduce transaction costs, disproportionately to finalize the competitive diagnostic and provide rele- benefitting sectors that engage in a complex web of transac- vant policy recommendations regarding policies affecting tions with the rest of the economy (Amin and Mattoo 2006). services trade. Certain institutions play a particularly significant role in the Assessing the business environment for FDI and the development of services sectors, for three reasons: ability to attract and protect foreign investors is a critical part of the diagnostic. Empirical research finds a posi- • Asymmetric information is greater in services markets tive correlation between services openness and growth than in goods markets, because it is harder for consum- (Mattoo and Rathindran 2006; Francois and Hoekman ers to evaluate the quality of services before they purchase 2010). Because one of the most relevant modes of supply them. Regulatory institutions therefore matter more. for services trade is commercial presence (which includes • Services sectors that require specialized distribution net- investment decisions of foreign and domestic providers), works (such as roads and rails for land transport, cables the analysis should focus on the following: and satellites for communications, and pipes for energy distribution) are likely to be natural monopolies or oli- • openness and degree of foreign participation gopolies. Regulations are needed to ensure that monopo- • government policy toward foreign capital (risk of listic suppliers do not exploit consumers or undermine expropriation, investment protection) market access (Mattoo, Stern, and Zannini 2007). • degree to which private property rights are guaranteed • For customized services, both consumers and suppliers and protected must make relationship-specific investments, raising • procedures required to set up a new business and free- the cost of switching from one supplier or consumer dom of existing businesses to compete to another. Contract-enforcing institutions therefore • legal environment (contract enforcement, piracy rates, assume greater importance in services than in goods. intellectual property protection and laws governing A firm’s willingness to outsource medical transcription ICT) and perceptions of corruption. or confidential financial information, for example, is much greater if it has confidence in the privacy and data Because many services activities are skill intensive, protection laws of the host country and therefore of the assessing the skill level of the country’s workforce, the firm responsible for processing the information (Amin availability of specific skills, and overall skill intensity, and Mattoo 2006). including managerial skills, of services sectors is critical. Many sectors (business services, banking, telecommunica- Molinuevo and Sáez (2014) provide a methodology tions) are significantly more skill intensive than most goods for assessing both the regulatory environment and the sectors. Cross-country evidence suggests that endowments 8 Valuing Services in Trade of human capital are an important determinant of output infrastructure and regulatory constraints, such as weak supply and growth (figure I.5). Sustaining growth in South Asia of electricity, telecommunications/Internet broadband, and may require addressing the serious systemic weaknesses real estate services as well as burdensome regulations that in higher education. In the Philippines, the success of affect the local business environment, some developing business process outsourcing (BPO) exports may soon be country governments have created special regimes, such as hampered by the dearth of skilled workers. Moving from IT parks. Export-oriented firms in such facilities enjoy dedi- BPO to knowledge process outsourcing (KPO) activities cated infrastructure and streamlined administration. requires higher levels of skills to attract investors. Telecommunications growth is the most powerful sym- Managerial skills also play a key role in the adoption bol of vitality of the services sector. It is also critical to the of modern technologies, which are critical for producing development of other services. ICT has reduced the cost high-quality services for export. Experience working in a of delivering many cross-border services to virtually zero. multinational firm has been found to develop such skills. Development of the Indian software industry is attributed largely to entrepreneurs of Indian origin who had world- Organization of the Toolkit class training in management skills or experience work- A diagnostics analysis of services trade assesses the ing abroad in multinational firms (Gregory, Nollen, and enabling factors and obstacles to services competitiveness. Tenev 2009). Conducting the assessment requires a number of prepara- Trade-related infrastructure is also important (figure I.5). tory steps, including defining the objectives of the assess- Transport and tourism are intensive in physical infrastructure ment, evaluating the main actors, assessing how it relates and sensitive to institutions such as customs and border man- to a country’s overall pro-competitiveness strategies, and agement. These characteristics make these subsectors defining the scope of the assessment. more similar to goods exports than to other services exports. This toolkit provides general guidance, based on the In contrast, commercial services that are usually traded across methodology developed by Reis and Farole (2012) for borders rely on telecommunication infrastructure and are goods. It provides guidance on finding answers to at least sensitive to contract enforcement. The diagnostic should six broad questions: identify the state of the infrastructure sectors required to develop specific services; assess their efficiency; and review the • How can one assess the trade performance of a country’s role of proactive policies, such as the creation of information services sector? technology (IT) parks, in addressing infrastructure needs. • How can one assess the role of services as inputs in the Some governments promote specific services sectors traded sectors of an economy? by providing trade-related infrastructure. To address key • What factors determine services trade performance? Figure I.5. Cross-Country Regression of Services Exports versus Gross Tertiary Enrollment and Internet Penetration a. Gross tertiary enrollment b. Internet penetration 12 12 11 11 Log of services exports Log of services exports 10 10 9 9 8 8 7 7 0 10 20 30 40 50 60 70 80 90 100 110 0 50 100 Gross tertiary enrollment (percent) Internet penetration (percent) Source: World Development Indicators and Goswami, Mattoo, and Sáez 2012. Notes 9 • What is the relative importance of these determinants economist with an understanding of trade policy and com- for different services? petitiveness issues and technical skills in analyzing services • How do policies affect the efficiency of these determi- trade data. The team leader should have in-depth country nants and of services exports? knowledge and experience and, ideally, some trade expe- • What are the main policy constraints to the growth and rience. It should not be necessary to involve technical development of trade in services? experts for each component of the diagnostic. However, if certain topics are considered critical at the outset, it may It does so by providing a methodology and instruments be useful to bring in specialized technical expertise to lead for (a) assessing countries’ performance and potential in those components. services, both as exports and as inputs into other traded Input from a wide variety of stakeholders in the country, activities, and (b) identifying key constraints that affect ser- including government officials and the private sector, is vices trade. Its objective is to help policy makers improve critical to the success of an STCD. Consultation should services trade and facilitate countries’ integration into take place through individual and focus group interviews. global markets by increasing their competitiveness. To In some cases, it may also be useful for key stakeholders achieve this objective, the toolkit provides a framework from government, business, and labor to participate in the for analyzing and assessing the performance of the services assessment. industry and identifying specific enabling factors or obsta- cles that affect the efficiency of services. It complements the analytical framework of trade in goods provided by the Notes Trade Competitiveness Diagnostic Toolkit (Reis and Farole 1. See http://unctadstat.unctad.org/TableViewer/tableView.aspx. 2012), allowing policy makers in developing countries to 2. Trade policies in services are all government measures that limit the international transaction of services by foreign providers. The General better integrate services into their overall trade strategies. Agreement on Trade in Services (GATS) of the World Trade Organiza- The toolkit consists of four modules. Module 1 pro- tion (WTO) applies to measures affecting trade in services. The focus of poses measures of outcomes in trade in services. Module 2 negotiations is on market access limitations and discrimination between assesses the existence and importance of a domestic ser- domestic and foreign services and services providers (national treatment). The GATS defines market access limitations as measures that affect the vices sector. Module 3 assesses countries’ potential for number of services suppliers, transactions, operations, number of peo- expanding trade in services. Module 4 examines broad ple that may be employed in a particular sector, the types of legal entity policy areas and options for addressing the constraints through which a supplier may provide a service, and the participation of foreign capital. Measures that limit national treatment are domestic regu- identified in Modules 1, 2, and 3. It distinguishes between lations that imply de jure or de facto discrimination of foreign services sectoral policies that can be implemented in the short to and providers. medium term (mostly technical regulations and barri- ers) and horizontal measures and strategies related to the domestic economy and institutions, which can be imple- References mented only over a longer time horizon. Acemoglu, Daron, and Simon Johnson. 2005. “Unbundling Institutions.” Conducting a Services Trade Competitiveness Journal of Political Economy 113 (5): 949–95. Amin, Mohammad, and Aaditya Mattoo. 2006. “Do Institutions Matter Diagnostic (STCD) requires six sequential steps: More for Services?” Policy Research Working Paper 4032, World Bank, Washington, DC. • desk-based assessment of trade outcomes Arnold, Jens, Beata Javorcik, Molly Lipscomb, and Aaditya Mattoo. 2010. • initial desk-based research on and formulation of “Services Reform and Manufacturing Performance: Evidence from India,” CEPR Discussion Paper 8011, Centre for Economic Policy hypotheses on failures and policy areas to address in Research, London. detail Arnold, Jens, Beata Javorcik, and Aaditya Mattoo. 2011. “Does • planning of and preparation for fieldwork Services Liberalization Benefit Manufacturing Firms? Evidence from the Czech Republic.” Journal of International Economics • in-country field research 85 (1): 136–46. • follow-up investigation of issues that emerge during Arnold, Jens M., Aaditya Mattoo, and Gaia Narciso. 2008. “Services field analysis Inputs and Firm Productivity in Sub-Saharan Africa: Evidence from Firm-Level Data.” Journal of African Economies 17 (4): 578–99. • preparation of final synthesis STCD report. Balassa, B. 1965. “Trade Liberalisation and Revealed Comparative Advantage.” The Manchester School 33 (2): 99–123. The STCD is designed to be completed within four Baldwin, R. 2011. “Trade and Industrialisation after Globalisation’s 2nd months, including between two to six weeks of field- Unbundling: How Building and Joining a Supply Chain Are Different and Why It Matters.” NBER Working Paper 17716, National Bureau work. It should be conducted by a core team of three or of Economic Research, Cambridge, MA. http://www.nber.org/papers four people. At least one team member should be a trade /w17716. 10 Valuing Services in Trade ———. 2012. “Global Supply Chains: Why They Emerged, Why They Krugman, Paul, and George N. Hatsopoulos. 1987. “The Problem of US Matter, and Where They Are Going.” CEPR Discussion Paper 9103, Competitiveness in Manufacturing.” New England Economic Review Centre for Economic Policy Research, London. (January/February). Buckley, Peter J., Christopher L. Pass, and Kate Prescott. 1988. “Measures Kunaka, Charles, Monica Alina Antoci, and Sebastián Sáez. 2013. “Trade of International Competitiveness: A Critical Survey.” Journal of Dimensions of Logistics Services: A Proposal for Trade Agreements.” Marketing Management 4 (2): 175–200. Journal of World Trade 47(4): 925–50. Cattaneo, O. and S. Miroudot. 2013. “From Global Value Chains to Lennon, Caroline. 2009. “Trade in Services and Trade in Goods: Global Development Chains: An Analysis of Recent Changes in Trade Differences and Complementarities.” WIIW Working Paper 53, Patterns and Development Paradigms.” In 21st Century Trade Policy: Vienna Institute of International Economics Studies (WIIW), Back to the Past? edited by E. Zedillo and B. Hoekman. New Haven, Vienna. CT: Yale University Press. Lipschitz L., and D. McDonald. 1991. “Real Exchange Rates and Copeland, Brian, and Aaditya Mattoo. 2008. “The Basic Economics of Competitiveness: A Clarification of Concepts and Some Measurement Services Trade.” In A Handbook of International Trade in Services, for Europe.” IMF Working Papers, Washington, DC. edited by Aaditya Mattoo, Robert M. Stern, and Gianni Zanini. Magdeleine, Joscelyn, and Andreas Maurer. 2008. “Measuring GATS Oxford: Oxford University Press. Mode 4 Trade Flows.” WTO Staff Working Paper ERSD-2008-05, Durand, M., and C. Giorno. 1987. “Indicators of International World Trade Organization, Geneva. Competitiveness: Conceptual Aspects and Evaluation.” OECD Mandeng, O. 1991. “International Competitiveness and Specialization.” Economic Studies 9: 147–82. CEPAL Review 45: 39–52. Fagerberg, J. 1988. “International Competitiveness.” Economic Journal Markusen, J. 1992. “Productivity, Competitiveness, Trade Performance 391: 355–74. and Real Income: The Nexus among Four Concepts.” Minister of Feenstra, Robert C. 2010. Offshoring in the Global Economy: Supply and Services, Ottawa, Canada. Microeconomic Structure and Macroeconomic Implications. Cambridge: Mattoo, Aaditya, and Randeep Rathindran. 2006. “Measuring Services MIT Press. Trade Liberalization and Its Impact on Economic Growth: An Francois, Joseph, and Bernard Hoekman. 2010. “Services Trade and Illustration.” Journal of Economic Integration 21: 64–98, Center for Policy.” Journal of Economic Literature 48 (3): 642–92. Economic Integration, Sejong University, Seoul. Francois, Joseph, and Julia Woerz. 2008. “Producer Services, Mattoo, Aaditya, Robert M. Stern, and Gianni Zanini, eds. 2007. Manufacturing Linkages, and Trade.” Journal of Industry, Competition A Handbook of International Trade in Services. Oxford: Oxford and Trade 8 (3): 199–229. University Press. Goswami, Arti Grover, Aaditya Mattoo, and Sebastián Sáez. 2012. “Service Molinuevo, Martin, and Sebastián Sáez. 2014. Regulatory Assessment Exports: Are the Drivers Different for Developing Countries?” Toolkit: A Practical Methodology for Assessing Regulation on Trade and In Exporting Services: A Developing Country Perspective, edited by Investment in Services. Washington, DC: World Bank. Arti Grover Goswami, Aaditya Mattoo, and Sebastián Sáez, 25–81. OECD (Organisation for Economic Co-operation and Development) and Washington, DC: World Bank. WTO (World Trade Organization). 2013. “OECD-WTO Database on Gregory, Neil, Stanley Nollen, and Stoyan Tenev. 2009. New Industries Trade in Value Added First Estimates.” January 16, Organisation for from New Places: The Emergence of the Hardware and Software Economic Co-operation and Development, Paris. Industries in China and India. Redwood City, CA: Stanford University Porter, Michael. 1985. Competitive Advantage: Creating and Sustaining Press and Washington, DC: World Bank. Superior Performance. New York: Free Press. Grossman, Gene M., and Esteban Rossi-Hansberg. 2008. “Task Trade ———.1990. The Competitive Advantage of Nations. New York: Free between Similar Countries.” NBER Working Paper 14554, National Press. Bureau of Economic Research, Cambridge, MA. Reis, José Guilherme, and Thomas Farole. 2012. Trade Competitiveness Hatsopoulos, George N., Paul R. Krugman, and Lawrence H. Summers. Diagnostic Toolkit. Washington, DC: World Bank. 1988. “US Competitiveness: Beyond the Trade Deficit.” Science 241: Sharpe, A. 1985. “Can Canada Compete? Part 1.” Canadian Business 200–307. Review 12 (4). Helleiner, G. K. 1989. “Transnational Corporations and Direct Foreign ———. 1986. “Can Canada Compete? Part 2.” Canadian Business Review Investment.” In Handbook of Development Economics, edited by Hollis 13 (1). Chenery and T. N. Srinivasan, 1441–80. Amsterdam: Elsevier. Swedish National Board of Trade (Kommerskollegium). 2010. At Your Helpman, Elhanan. 2011. Understanding Global Trade. Cambridge, MA: Service: The Importance of Services for Manufacturing Companies and Belknap Press of Harvard University Press. Possible Trade Policy Implications. Stockholm. http://www.kommers Hoekman, Bernard, and Michel Kostecki. 2009. The Political Economy of the . se/Documents/dokumentarkiv/publikationer/2010/skriftserien World Trading System, 3rd. edited by Oxford: Oxford University Press. /report-2010-2-at-your-service.pdf. Inklaar, Robert, Marcel Timmer, and Bart van Ark. 2007. “Mind the Gap! Triplett, Jack E., and Barry P. Bosworth. 2004. Productivity in the US International Comparisons of Productivity in Services and Goods Services Sector: New Sources of Economic Growth. Washington, DC: Production.” German Economic Review 8 (5): 281–307. Brookings Institution Press. ———. 2008. “Market Services Productivity across Europe and the US.” van der Marel, E. 2012. “Trade in Services and TFP: The Role of Economic Policy 23: 141–94. Regulation.” The World Economy 35 (11): 1387–429. Jones, Ronald W. 2000. Globalization and the Theory of Input Trade. WEF (World Economic Forum). 2004 “The Global Competitiveness Cambridge, MA: MIT Press. Index.” Global Competitiveness Report. Module 1 Assessing Services Trade and Competitiveness Outcomes Services are important because they are a source of measures include export growth compared with goods export diversification and a source of competitiveness for and peer countries, services exports relative to services the economy as a whole. This module describes instruments value added, measures of revealed comparative advantage that quantify where countries stand in the former respect. (RCA), measures of the evolution of exports and world The methodology for assessing services trade outcomes demand that reveal whether specialization observed focuses on the size and scope of trade in services. It pro- in a country is moving into the most dynamic sec- vides a quantitative and qualitative assessment of historical tors, and firm-level measures of exports and imports performance in trade in services. of services. Both dimensions present important challenges. The indicators in this module are built mainly on the Measuring services trade faces more constraints than basis of balance of payments data, which broadly reflect measuring trade in goods, because detailed industry, cross-border transactions through Modes 1 and 2. Other product, geographical, and time series data are limited. data sources for other modes of supply are discussed Standard statistics for trade in services are wholly unsat- through Modules 2 and 3. Modules 2, 3, and 4 use additional isfactory for describing the competitiveness of the services indicators and other data sources to enrich the analysis. sector. For emerging markets and developing countries, even standard statistics for trade in services are scant. Size of Trade in Services The World Bank’s Trade in Services Database provides better data than were previously available (box 1.1). The toolkit starts with an analysis of services trade out- Because of the difficulty of recording international comes. It provides a quantitative and qualitative assess- transactions of services, however, even this data set does ment of the historical performance of trade in services. The not provide the degree of disaggregation that can be found focus is on determining and comparing the size and scope in goods trade data. of trade in services. Analysis of services trade outcomes is largely a desk- based exercise that involves assessing a series of indicators Exports of Services and analytical tools. The construction of indicators must be complemented by focused, in-depth analysis of specific Measuring the Value of Services Exports research questions and by targeted field analysis, to ensure Several indicators are critical to assessing the value of the correct interpretation of the results and to reveal more services exports. They include the services and goods trade subtle trends and developments that may not be evident balance, the value of trade in services, and the value of exports and imports of services, all as shares of GDP. Data from the data. Firm-level statistics are also a powerful com- on these indicators should be collected for specific years plementary tool. They include census-type information, for all countries. Useful sources of information include the balance sheet data, and records of export and import trans- following: actions. The availability of this information differs from • World Development Indicators (services trade balance, GDP) • IMF (balance of payments statistics) country to country. • UNCTADstat (trade balance in services and trade balance The indicators analyzed in this module focus on in goods). services export performance. The export performance 11 12 Valuing Services in Trade Module 1 Box 1.1. Modes of Services Trade and Data on Trade in Services Services have unique characteristics, such as intangibility and nonstorability, that greatly affect their tradability. They typically also require differentiation and jointness in production—that is, the participation of customers in the production process. In order to capture these aspects and allow for trade in services that also require joint production, the General Agreement on Trade in Services (GATS) of the World Trade Organization (WTO) defines four modes of supply of services: • Mode 1: Cross-border trade (services supplied from the territory of one country into the territory of another) • Mode 2: Consumption abroad (services supplied in the territory of one country to the consumers of another) • Mode 3: Commercial presence (services supplied in the territory of one country by any type of business or professional establish- ment in another [foreign direct investment]) • Mode 4: Presence of natural persons (services supplied by nationals of a country in the territory of another). (For more on these modes, see Module 3.) Only data on cross-border services trade in Modes 1, 2, and 4 are reported in balance of payments statistics. Foreign direct investment (FDI) is not covered. About 60 percent of the global FDI stock is in the service sector, with finance and trade the most important sectors. Services are also traded through cross-border movement of persons (Mode 4). On the consumer side, Mode 2 covers, for example, Germans going to Poland for dental work, as well as tourism. On the producer side (Mode 4), it includes the temporary cross-border movement of skilled labor, such as accountants and software engineers who work across Europe. It also includes Polish construction workers relocating temporarily to work in the Netherlands or France. Trade through affiliates (Mode 3) includes exports that pass through affiliates. Indeed, given the nature of services trade and the role of FDI in the sector, the activities of affiliates include a mix of cross-border and local activities. The quality of trade data in services is far from comparable to the data for goods. Because of the long tradition of tariff revenues, the quality of trade data for goods is high. Because of the intangibility and nonstorability of services, at-the-border-duties cannot be collected. As a result, the data are much less accurate. Particularly with respect to Modes 3 and 4, up-to-date measurement is difficult and incomplete. Ongoing revisions and refinements of the balance of payments classifications are trying to improve the quality of the data. To create a data set covering bilateral services flows (Trade in Services Database), the World Bank aggregated data collected by the Organisation for Economic Co-operation and Development (OECD), Eurostat, the United Nations, and the International Monetary Fund (IMF) (Appendix B describes the data set). IMF data cover only trade with the world as a partner. The OECD, Eurostat, and United Nations provide data on bilateral services trade flows by partners and balance of payments codes. The most comprehensive coverage is by the United Nations, which provides data on 190 reporters. Eurostat and the OECD provide data for a limited number of countries: Eurostat covers 27 members of the European Union plus Croatia, Iceland, Japan, Norway, Turkey, Switzerland, and the United States. Historically, most services sectors were considered non- services exports was linked to the falling costs of trad- tradable. This concept changed dramatically over the last ing, as a result of declines in the costs of transportation two decades, as trade increased in both traditional services and travel; reductions in trade barriers; and improve- sectors (such as transport and postal services) and modern ments in technology. Technological change has been services (such as business services, telecommunications, particularly important for services: increased use of ICT, and information and communications technology [ICT]).1 including the Internet, signifi cantly reduced some of However, compared with goods, services trade remains the costs associated with trade in services (Freund and modest for all income groups, as shown in figure 1.1. Weinhold 2002). Trade in and exports of services are increasing rapidly Services are intrinsically less tradable than most for all income groups, as shown in figure 1.2. The share of goods, because many of them require that buyer and developing countries in world exports of services increased seller meet. But technological advances are enhancing from 18 percent in 1990 to 30 percent in 2012.2 the ability to produce services in one location and con- Figure 1.2 shows the normalized values of exports sume them in another. ICT and other Internet-related of each income group in order to highlight growth dif- technologies, such as cloud computing, create new ferences across income groups during the first decade opportunities for trade in services and expand the scope of the 21st century. Growth was highest for the lower- of services trade. middle-income group, but all groups grew steadily Figure 1.3 shows how the increased tradability of ser- throughout the period. However, in absolute terms, the vices has affected the contributions of different income high-income group remained dominant. The growth of groups to total world exports in services. It reveals that Module 1: Assessing Services Trade and Competitiveness Outcomes 13 Figure 1.1. Shares of Exports of Goods and Services, by Figure 1.3. Contribution to World Services Exports, by Module 1 Country Income Level, 2011 Income Level, 1990–2010 100 Contributions of income groups to service exports 4,000 Exports of services (billions of dollars) 80 Percent of total exports 3,000 60 2,000 40 1,000 20 0 1990 1995 2000 2005 2010 0 Low Lower- Upper- High High income Upper-middle income income middle middle income Lower-middle income Low income Country income level Source: World Development Indicators. Services Goods Note: Income classification is as of July 1, 2012. Source: World Development Indicators. Note: Income classification is as of July 1, 2012. despite their rising participation, low- and lower-middle- income countries still make only a very modest contribu- tion to total trade. Their minor participation in services Figure 1.2. Index of Services Exports, by Country Income trade explains why average world growth in services exports Level, 2000–10 is almost identical to growth in high-income countries. The modest contributions by low- and middle-income coun- 5 tries compared with their rapid economic growth figures can also be explained by the fact that these country groups 4 start from a lower base, reflecting significant potential for Index (2000 = 1) the future development of services trade. The trade balance indicates whether a country is a net 3 importer or net exporter of services. Panel a of figure 1.4 shows that this indicator, taken as a percentage of gross 2 domestic product (GDP), ranges widely across country income groups, with only the high-income group being net 1 exporters. 2000 2002 2004 2006 2008 2010 Panel b of figure 1.4 reports the services trade bal- ance for five middle-income countries: three from Latin year America (Chile, Costa Rica, and Paraguay) and two com- Low income Lower-middle income parators from Asia (India and Malaysia). The indicators Upper-middle income High income vary even more widely than by income groups. Costa World Rica and Paraguay record a consistent and growing sur- Source: World Development Indicators. plus over the period 2001–10, Malaysia’s trade balance Note: Income classification is as of July 1, 2012. increased steadily, and Chile’s trade balance was consis- tently negative. 14 Valuing Services in Trade Figure 1.4. Services Trade Balance as Percent of GDP, by Income Group and in Selected Countries, 2001–10 Module 1 a. By income group b. In selected countries 0.10 0.10 Services trade balance as percent of GDP Services trade balance as percent of GDP 0.05 0.05 0 0 −0.05 −0.05 01 03 01 03 10 08 10 09 09 08 04 02 02 04 05 05 06 06 07 07 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 High Upper middle Costa Rica Paraguay India Low Lower middle Chile Malaysia Source: World Development Indicators. Note: Income classification is as of July 1, 2012. GDP = gross domestic product. Trade Balance for Services and Goods 4–9 percent of GDP in Kazakhstan and 2–12 percent of GDP in Tajikistan. Except in 2010, since 2002, the Kyrgyz It is useful to compare the trade balance for goods and ser- Republic has shown a relatively balanced services trade vices. Figure 1.5 shows that high-income countries have account. Overall, however, the size of the deficits tend to be a positive trade balance in services but a negative trade much lower for services than for total trade in the Kyrgyz balance in goods. This picture is exactly the opposite for Republic and Tajikistan, suggesting that at least for these upper-middle income countries, where the goods balance countries, services are a marginal component of total trade. is positive and the services balance negative. Figure 1.5 also shows that although the services trade balance of the Trade in Services and Development lower-middle-income group is negative, it increased over the last 10 years. This group’s share of services exports rela- The trade balance yields limited information about the tive to total goods was comparable to that of low-income absolute importance of services trade for the economy. countries. Other indicators must be examined. Figure 1.6 shows the trade balance of three countries Figure 1.7 depicts one such indicator: total trade in ser- in Central Asia between 2000 and 2010. With respect to vices (the sum of exports and imports) divided by popu- trade in goods, oil exporter Kazakhstan (and other large lation (per capita) and measured against the economic economies) ran a trade surplus. Tajikistan (and other small development of the country (measured as GDP per capita economies) ran a deficit most of the time. For services, the in purchasing power parities [PPPs]). Trade in services only country with a positive net balance was Uzbekistan, per capita is used to obtain a normalized measure of the which is not shown in the figure, because not all data were importance of services in each country, rather than analyz- available. The services trade balance deficit ranged between ing by country size. Module 1: Assessing Services Trade and Competitiveness Outcomes 15 Figure 1.5. Trade Balance for Goods and Services, by Income Group, 2000–10 Module 1 a. High income b. Low income 6 6 5 5 4 4 3 3 2 2 1 1 0 0 −1 −1 Percent of GDP Percent of GDP −2 −2 −3 −3 −4 −4 −5 −5 −6 −6 −7 −7 –8 –8 –9 –9 −10 −10 –11 –11 –12 –12 –13 –13 00 02 01 03 04 05 06 07 08 09 10 00 02 01 03 04 05 06 07 08 09 10 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 c. Lower-middle income d. Upper-middle income 6 6 5 5 4 4 3 3 2 2 1 1 0 0 −1 −1 Percent of GDP Percent of GDP −2 −2 −3 −3 −4 −4 −5 −5 −6 −6 −7 −7 –8 –8 –9 –9 −10 −10 –11 –11 –12 –12 –13 –13 0 02 01 03 04 05 06 07 08 09 10 0 02 01 03 04 05 06 07 08 09 10 0 0 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 Goods Services Total Sources: Trade data are from UNCTADstat. GDP data are from World Development Indicators (accessed in June 2012). Note: Trade and GDP data are in current U.S. dollars. 16 Valuing Services in Trade Figure 1.6. Trade Balance in Kazakhstan, the Kyrgyz Republic, and Tajikistan, as Percent of GDP, 2000–10 Module 1 a. Kazakhstan b. Kyrgyz Repubilc 25 25 20 20 15 15 10 10 5 5 0 0 Percent of GDP Percent of GDP −5 −5 −10 −10 −15 −15 −20 −20 −25 −25 −30 −30 −35 −35 −40 −40 −45 −45 00 01 02 03 04 05 06 07 08 09 10 00 01 02 03 04 05 06 07 08 09 10 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 c. Tajikistan 25 20 15 10 5 0 Percent of GDP −5 −10 −15 −20 −25 −30 −35 −40 −45 00 01 02 03 04 05 06 07 8 9 0 0 0 1 20 20 20 20 20 20 20 20 20 20 20 Goods Services Total Sources: Trade data are from UNCTADstat. GDP data are from World Development Indicators (accessed in June 2012). Note: Trade and GDP data are in current U.S. dollars. Module 1: Assessing Services Trade and Competitiveness Outcomes 17 Figure 1.7. Per Capita Trade in Services and Per Capita GDP, 2002–04 and 2008–10 Module 1 a. 2002–04 b. 2008–10 12 12 10 Log of per capita trade in services Log of per capita trade in services 10 8 8 CHL CHL 6 6 MEX PERMEX COL PER COL 4 4 2 2 6 8 10 12 6 8 10 12 Log of per capita GDP (PPP) Log of per capita GDP (PPP) Fitted values High-income country Upper-middle-income country Low-income country Lower-middle-income country Source: World Development Indicators. Note: Income classification is as of July 1, 2012. CHL = Chile, COL = Colombia, MEX = Mexico, PER = Peru, PPP = purchasing power parity. The figure shows that the importance of trade in Figure 1.9 shows more specifically the relationship services increases with the level of development. Many between exports and development. Note, however, that both countries—such as Chile, Colombia, Mexico, and Peru, imports and exports are important in trade patterns. and also highly developed ones—actually export less Openness of countries is also explained by larger import than one might expect based on the level of development. shares, which implies greater export potential. This is true Figure 1.7 shows that between 2002–04 and 2008–10, trade particularly when considering the capabilities of develop- openness seemed to increase for Colombia and Peru. Most ing countries to join a global value chain (GVC), because other countries, including some high-income ones, how- a country cannot become a major exporter within GCVs ever, seem to be stuck in their initial situation. without first becoming a major importer: exports within Other structural factors may also help explain these GVCs require imports (box 1.2). results. Figure 1.8 plots the residuals of the regression shown As stated, both imports and exports are of vital impor- in table 1.1, which regresses the log of trade in services per tance; both indicate the extent to which a country partici- capita against log of population and log of per capita income pates in the globalization process. Yet, because the main in PPP. Both population and GDP are significant variables in purpose of this diagnosis is to assess external competitive- table 1.1, but there is substantial variance in the performance ness, the focus is on exports. of countries not explained by these variables, as shown by the large residuals from this regression plotted on the horizontal Share of Services versus Share of Services Exports in GDP axis of figure 1.8. There is also a strong correlation between the log of trade in services per capita and the residuals, sug- The share of services exports in GDP indicates whether the gesting that the more important services are for the economy, importance of the services sector in the domestic economy the higher the residuals (or, the more important other struc- translates into competitiveness externally. If it does not, tural factors are in explaining this performance). the diagnostic should reveal whether the reason is the 18 Valuing Services in Trade Figure 1.8. Per Capita Trade in Services and Ordinary Least Squares Residuals, 2002–04 and 2008–10 Module 1 a. 2002–04 b. 2008–10 12 12 10 Log of per capita trade in services Log of per capita trade in services 10 8 8 CHL CHL 6 6 MEX MEX PER COL PER COL 4 4 2 2 −2 −1 0 1 2 3 −2 −1 0 1 2 3 Ordinary least squares residuals Ordinary least squares residuals Fitted values High-income country Upper-middle-income country Low-income country Lower-middle-income country Source: World Development Indicators. Note: Income classification is as of July 1, 2012. CHL = Chile, COL = Colombia, MEX = Mexico, PER = Peru. Table 1.1. Results of Regression of Trade in Services against Figure 1.10 suggests that countries are becoming more Development and Population competitive on the world market in goods and hence are Item ln(services trade pc) exporting more goods but that such exports add only ln(population) –0.243*** marginally to their GDP. While exports of goods increase, (0.0175) the share of industry in GDP declines. Similarly, although ln(gdp_pc_ppp) 1.280*** (0.0274) the share of exports of agriculture products in GDP Observations 321 increases, the share of agriculture value added decreases. R-squared 0.891 The value added of goods rose significantly in Mauritius; Root mean square deviation 0.610 Hong Kong SAR, China; and Lebanon, but the increase did Note: Standard errors are in parentheses. *** p < 0.01. not reflect increased exports of goods. In contrast, larger shares of services exports in GDP appear to be associated with larger shares of value added in GDP. This indicator nontradability of services produced in the economy or does not, however, provide information on who is actually other causes, such as regulatory obstacles. involved in trade in services. Box 1.3 discusses the charac- Figure 1.10 shows a positive relationship between the teristics of exporters using firm-level data. value added of services in the domestic economy and the share of services exports in GDP. A negative relationship Comparing a Country’s Trade in Services with is evident for industry and agriculture. The figure also Peers and the World shows that in many important economies, the size of the services export sector is smaller than would have been pre- An important starting point in diagnosing the competi- dicted based on the size of services in the domestic econ- tiveness of individual countries is the scale of services omy. (Module 3 further explores the export potential of exports. Figure 1.11 compares Turkey’s commercial ser- services.) vices exports to BRICS countries (Brazil, Russia, India, Module 1: Assessing Services Trade and Competitiveness Outcomes 19 Figure 1.9. Per Capita Exports in Services and Per Capita GDP, 2002–04 and 2008–10 Module 1 a. 2002–04 b. 2008–10 12 10 10 Log of per capita exports of services Log of per capita exports of services 8 CHL 5 MEX CHL PER 6 MEX COL PER COL 4 2 0 6 7 8 9 10 11 6 8 10 12 Log of per capita GDP (PPP) Log of per capita GDP (PPP) Fitted values High-income country Upper-middle-income country Low-income country Lower-middle-income country Source: World Development Indicators. Note: Income classification is as of July 1, 2012. CHL = Chile, COL = Colombia, MEX = Mexico, PER = Peru, PPP = purchasing power parity. Box 1.2. How Can Developing Countries Join and Climb the Ladder of Global Value Chains? In recent years, global value chains (GVCs) have profoundly affected international trade and development. Developing countries now need to figure out how they can join different segments of this process rather than enhance the development of entire industries. Joining a GVC can be facilitated by policies such as low tariffs on intermediate goods, an efficient transport and logistics net- work, a good business and regulatory environment, and the ability to meet product standards. Good policies on these factors can help boost competitiveness, increase connectivity with international markets, and improve the business and investment climate. A World Bank toolkit (Making Global Value Chains Work for Development [Taglioni and Winkler 2014]) focuses on GVCs. Services play a crucial role in GVCs: they represent about 30 percent of the value added in goods trade (OECD and WTO 2013). Efficient domestic services and the availability of imported services are therefore vital. The reduction in supply chain barriers such as transport, logistics, and communication infrastructure and related services have a greater impact on growth of GDP and trade than the elimination of tariffs (WEF 2013). Upgrading within GVCs is also important. Upgrading processes, products, or skills can increase the services content of produc- tion by moving away from traditional production (assembly stage) into pre- and postfabrication stages that are more services intensive and contribute more value added in the production cycle (see Cattaneo and others 2013; Milberg and Winkler 2013). China, and South Africa) plus Poland and Romania. is larger in Turkey than in all of its comparators except Turkey’s exports in services reached $34 billion in 2010, India. However, this measure of services exports divided 28 percent of the value of goods exports and 22 percent by exports in goods may give a distorted picture for India, of the value of all goods and services exports. The scale of whose industrial sector is underdeveloped because of his- Turkey’s commercial services export sector compares well torical policy goals. with BRICS’s and regional peers’. The share of commer- One way to correct this picture is to use a different mea- cial services in the total goods and services export basket sure, such as market size, to normalize services exports. 20 Valuing Services in Trade Figure 1.10. Size of Sector and Exports of Sector as Share of GDP in Selected Economies, 2008–10 Module 1 a. Services b. Industry c. Agriculture 100 100 100 MAC MAC MAC 80 80 80 Share of sector in GDP (percent) Share of sector in GDP (percent) Share of sector in GDP (percent) SYC SYC SYC 60 60 60 SGP SGP SGP LBN LBN HKG HKG LBN ATG ATG 40 IRL 40 IRL HKG ATG 40 IRL MUS MUS 20 20 MUS MYS MYS 20 CRI CRI PRY DEU IND PRYIND MOZ GHA MOZ GHA DEU MYS CHL CHL CRI PNG SLE CHN RUS JPNUSA USA SLE RUS PNG CHN BRA JPN BRA 0 AGO 0 AGO PRY MEX MEX DEU IND MOZ CHL GHA USA RUS JPN CHN PNG SLE BRA AGO 0 MEX 20 40 60 80 100 20 40 60 80 100 20 40 60 80 100 Share of exports of Share of exports of Share of exports of sector in GDP (percent) sector in GDP (percent) sector in GDP (percent) Fitted values High-income country Upper-middle-income country Low-income country Lower-middle-income country Source: World Development Indicators. Note: The vertical axis is expressed in value added terms. The horizontal axis is expressed in gross values. Income classification is as of July 1, 2012. For the International Organization for Standardization (ISO) abbreviations, see appendix B. Based on this metric, India performs worse than Poland manufacturing and services value added in GDP (the hori- and even Romania. As exports per capita show how much zontal axis). The horizontal axis is a rough proxy of eco- a country exports relative to its own market size (as mea- nomic diversification. Earning more services exports (per sured by population), it gives a good indication of the capita) from an economic base that is more diversified is presence in foreign markets, according to Wignaraja and expected to generate a more sustainable growth pattern; Taylor (2003). Poland performs very well in foreign mar- diversifying away from natural resources or agriculture is kets given the size of its domestic market. usually a sign of development (McKinsey 2010). The extent of presence in foreign markets is only one Not all countries have diversified their economies. dimension of external competitiveness. Diversification Countries such as Ethiopia or Cambodia, in the top left is another. Ideally, exports should come from a well- corner, have above-average services export income, but diversified set of sectors rather than a few sectors, such their economies are relatively undiversified. In contrast, as minerals or textiles. A well-diversified economy can China, India, and Malaysia, also in the upper left quadrant, generally expect a more sustainable growth path (see, for have more diversified economies. Figure 1.12 also shows instance, Sachs and Warner 1999). that a substantial number of upper-middle- and lower- Figure 1.12 plots the per capita services export earn- middle-income countries, such as Brunei and Mauritius ings (the vertical axis) against the combined share of (bottom right corner), still have much scope for increasing Module 1: Assessing Services Trade and Competitiveness Outcomes 21 Module 1 Box 1.3. What Kind of Firms Export Services? Analysis of firm- and plant-level data from many countries shows that exporters, including exporters of services, are more produc- tive than nonexporters (Bernard and Jensen 1999; Melitz 2003). Services exporters tend to be larger (as measured by the number of employees) than nonexporters (figure B1.3.1). This finding for exporters of goods is consistent with empirical studies by Iacovone, Mattoo, and Zahler (2013) and Brienlich and Criscuolo (2011) that analyze the characteristics of services exporters in Chile and the United Kingdom, respectively. Figure B1.3.1. Number of Employees of Exporting and Nonexporting Firms, 2009 >200 51–200 21–50 11–20 6–10 1–5 No employees 80 60 40 20 0 20 40 60 80 Percent of firms Nonexporters Exporters of services Source: World Bank Structural Business Survey. Figure 1.11. Commercial Services Exports by Turkey and Selected Comparator Countries, 2010 60 Per capita exports of commercial services (dollar) 900 50 800 as a percent of goods export Commercial services exports 40 600 30 400 20 200 10 0 0 Turkey Brazil Russian India China South Poland Romania Federation Africa Left axis Right axis Source: World Development Indicators. 22 Valuing Services in Trade Figure 1.12. Services Exports and Diversification in Selected Economies, 2008–10 Module 1 18 IND CHN USA 16 DEU Log of real services exports per capita JPN BRA ETH MYS RUS HKG IRL SGP 14 LBN GHA KHM MEX LUX MOZ CHL CRI MAC PRY 12 MUS PNG AGO SLE ATG SLB SYC 10 BRN DMA LBY TON 8 40 60 80 100 Manufacturing and services as percent of GDP Fitted values High-income country Upper-middle-income country Low-income country Lower-middle-income country Source: World Development Indicators. Note: Income classification is as of July 1, 2012. For the International Organization for Standardization (ISO) abbreviations, see appendix B. The dotted lines represent the median value of the log of real exports per capita (y-axis) and the share of manufacturing and services sectors in GDP (x-axis). per capita export earnings in services as predicted by their fastest rate of increase in the world. Exports of commercial levels of diversification. services grew at an average annual rate of 14.6 percent in Dynamics are also important. Scale provides a snapshot Kazakhstan, 10.7 percent in Tajikistan, and 9.8 percent in of the relative importance of countries in the global map Uzbekistan. The average for the world was 9.3 percent. of services exports; dynamic trends give a sense of how the Although the region’s global share of services exports relative positions may have changed over time. has been increasing, it remains very low, accounting for Table 1.2 shows annual growth rates for exports of just 0.16 percent of world exports of commercial services in goods and services and export growth rankings in Central 2010. The small share also reflects the fact that the region’s Asia over the last decade. In all of them, exports of ser- services share in total exports has declined, from 10.0 per- vices expanded more rapidly than world services exports. cent in 2000 to 7.6 percent in 2010, because goods exports Exports of commercial services by the Kyrgyz Republic, have grown even more rapidly (16.2 percent). This trend Kazakhstan, Tajikistan, and Uzbekistan reached almost is different from the trend for the world as a whole, which $6 billion in 2010. The Kyrgyz Republic was the most saw the shares of services in global exports increase from dynamic economy in the region: exports of commercial 19.7 percent to 20.7 percent over the same period. services (total services minus government services) grew at Figure 1.13 shows the dynamics of services exports by an annual rate of 24.7 percent between 2000 and 2010, the India, Turkey, and comparator countries in the region Module 1: Assessing Services Trade and Competitiveness Outcomes 23 Table 1.2. Growth of Exports of Goods and Services in Central Asia, by Country, 2000–10 Module 1 Exports of goods Exports of commercial services Rank among Average annual Rank among Average annual Country 207 countries growth rate Country 165 countries growth rate Kazakhstan 12 18.7 Kyrgyz Republic 1 24.7 Uzbekistan 48 14.1 Kazakhstan 30 14.6 Kyrgyz Republic 69 12.4 Tajikistan 69 10.7 Turkmenistan 103 9.5 Turkmenistan — — Uzbekistan 83 9.8 World n.a. 8.6 World n.a. 9.3 Tajikistan 164 4.3 Central Asia n.a. 16.2 Central Asia n.a. 13.9 Source: UNCTADstat Database online, accessed in June 2012. Note: Annual growth rates were calculated as logarithmic compounded returns. Commercial services are total services minus government services. n.a. = Not applicable. — = Not available. Figure 1.13. Annual Growth of Exports of Services by India, Turkey, and Regional and Income-Group Comparators, 1994–2010 a. India b. Turkey 40 15 30 10 Index (1994 = 100) Index (1994 = 100) 20 5 10 Fitted values 0 0 1995 2000 2005 2010 1995 2000 2005 2010 Total services Nontransport services Total services in income group Nontransport services in income group Total services in income group in region Nontransport services in income group in region Source: Trade in Services Database. Note: Regions are South Asia for India and Europe and Central Asia for Turkey. Income groups are lower-middle for India and upper-middle for Turkey. Regional classifications are from World Development Indicators. 24 Valuing Services in Trade between 1994 and 2010. During the period, the value of Is a country exporting services for which global demand is Module 1 India’s services exports increased by a factor of more growing? To which part of the world is a country exporting— than 17; excluding transport services, the value of services to countries that are growing rapidly or to countries that are exports increased by a factor of 37. India performed far already relatively rich and in which growth is limited? better than other lower-middle-income countries. Figure 1.15 investigates these questions for India. As In Turkey, the average annualized rate of growth was just expected, India’s export share is very high for other com- 6.8 percent over the decade—below the global average, below mercial services, business services, and computer and the average for all peer countries (both other countries at the information services (panel a). Global demand for com- same income level and other countries in the region), and puter and information services is high. Global demand is below Turkey’s own performance in goods exports. Between also high in the rail transport; road transport; advertising 1994 and 2007, the value of total services exports, exclud- and market research; and agricultural, mining, and on- ing transport, increased by just 2.6 percent a year. Over the site processing services sectors; India’s share of exports is same period, upper-middle-income countries from the rest lower in these sectors. Ideally, the two indicators would be of Europe and Central Asia recorded a growth in exports of positively correlated, so that the rest of the world can “pull” services excluding transport of 339 percent. Turkey’s overall India’s service exports. services exports increased by 195 percent—far lower than Panel b in figure 1.15 shows the link between India’s for upper-middle-income countries in the region. export share in services and the bilateral increase in A useful way to look at the evolution of services exports demand by partner countries in India’s services overall. over time is to relate it to world demand, measured as This relationship is negative: India’s services exports go total imports of services by all countries. The solid line mainly to destinations in which demand growth was weak in figure 1.14 shows total world demand (imports). The in 2000–08, such as the Germany, the United Kingdom, separate country lines represent the extent to which each and the United States. Higher growth in services trade is country can meet this demand (exports). The figure shows observable in, for instance, Bulgaria, Latvia, and Romania, that world demand for services is growing too quickly for countries in which India’s stake as a services exporter is most countries to meet. Only in India did exports of ser- modest. Although India’s trading partners for services are vices rise more rapidly than world imports. For the world not the countries that enjoyed the highest rates of services as a whole, demand fell after 2007. However, as India still import growth, they still represent large markets, so that falls above the world line, it is still better able than its com- the negative impact as the result of the decline in growth parators to meet world demand. will likely be limited. Figure 1.14. World Demand for Imported Services and Exports of Services by India, Malaysia, and Mexico, 2002–09 2.5 2.0 Index (2002 = 1) 1.5 1.0 0.5 2002 2004 2006 2008 2010 World India Mexico Malaysia Source: Trade in Services Database. Module 1: Assessing Services Trade and Competitiveness Outcomes 25 Figure 1.15. Services Exports by India, by Sector and Destination, 2000–08 Module 1 a. By sector b. By destination 6.5 7 281 LTU 6.0 ROM Log of country growth in demand for services CYP Log of world growth in demand for services 6 SVK 274 BGR LVA SVN HUN EST IRL 5.5 MLT CZE 249 262 POL FIN NOR ESP GBR 279 HRV 5 LUX DNK 219 278 ITA FRA SWE BEL 253 PRT NLD 223 AUT 5.0 273 DEU 280 206 TUR 268 982 245 214 USA 266 260 4 4.5 210 287 236 JPN 4.0 3 −4 −2 0 2 4 −6 −4 −2 0 2 Log of India’s services exports as percent Log of India’s services exports as percent of total exports of total exports Source: Trade in Services Database. Note: For the IMF balance of payments categories, see page xxi; for the country codes of the International Organization for Standardization (ISO), see appendix B. Structure of Services Exports by Sector sectors? Is the country exporting low value-added services, or is it competitive in sectors requiring high-level skills and Assessing the Sectoral Composition of Services strong domestic infrastructure? Exports Several data sets can provide answers to these questions. Several indicators are critical to assessing the structure of services exports. They include the services and goods trade One is the World Bank’s World Development Indicators balance, the value of trade in services, and the value of (WDI). It provides information, based on balance of pay- exports and imports of services, which can be calculated as ments statistics from the IMF, on the travel, transport, shares of GDP. Data on these indicators should be collected for specific years for all countries. Useful sources of informa- financial and insurance, communication, and computer tion include the following: services sectors. The WDI groups all services other than • World Development Indicators (services trade by sector government, travel, and transport services in the “other and GDP) commercial services” category. • IMF (balance of payments statistics) • UNCTADstat (trade balance in services) As figure 1.16 shows, all income groups are net import- • World Bank’s Trade in Services Database (see appendix B). ers of transport services. Only high-income countries are net importers of travel services. Only high-income economies are net exporters of financial services. The positive balance in trade in communications services has been driven largely After assessing aggregate services exports, the next stage by low- and lower-middle-income countries in recent years. is to look at their sectoral composition. Which services Figure 1.17 shows the structure of services exports are exported? Are all exports concentrated in one or a few by selected lower-middle-income countries. India is an 26 Valuing Services in Trade Figure 1.16. Balance of Trade in Services, by Country Income Level and Sector, 2000–10 Module 1 a. Communications b. Finance Balance as percent of GDP Balance as percent of GDP 1.5 1.0 1.0 0.5 0.5 0 0 −0.5 −0.5 2000 2002 2004 2006 2008 2010 2000 2002 2004 2006 2008 2010 Year Year c. Transport d. Travel Balance as percent of GDP Balance as percent of GDP 0 1.5 −0.5 1.0 −1.0 0.5 −1.5 0 −2.0 2000 2002 2004 2006 2008 2010 2000 2002 2004 2006 2008 2010 Year Year High income Upper-middle income Lower-middle income Low income Source: World Development Indicators. Note: Income classification is as of July 1, 2012. important net exporter of communication services, an region were high for communications in Azerbaijan importer of most transport services, and neither a net (940 percent), finance in Georgia (490 percent), and travel exporter nor a net importer of travel services. Ukraine in Armenia (980 percent) and Azerbaijan (940 percent). increased it exports of communication services but saw a Figure 1.18 puts these figures into perspective by show- decline in exports of transport services. (This pattern is also ing the composition of services exports for each of the evident in other emerging countries within this income selected countries in Central Asia. It shows that Tajikistan group, such as the Philippines and, to a lesser extent, drastically changed its export specialization, from trans- Indonesia.) Other lower-middle-income economies, such port to communications, which accounted for more than as Bolivia, Nicaragua, and Senegal, show few differences 70 percent of services exports by 2010. Kazakhstan and across sectors. Some low-income countries have been able Mongolia reduced their dependence on travel services and to increase services exports through travel/tourism services increased their exports of transport services. Each of the rather than finance or transportation. Central Asian economies specializes in a different services Table 1.3 shows the export growth rate of communica- sector. Kazakhstan specializes in exports of transport ser- tions, finance, transport, and travel services between 2000 vices, the Kyrgyz Republic in travel services, and Tajikistan and 2010 for selected countries in Central and South Asia in communications services. and the Caucasus as well as the averages for the income As countries become richer, the role of services in the groups to which these countries belong. What immediately economy increases, in terms of both value added and stands out is the impressive expansion of services exports exports. However, not all services sectors grow in importance in the Kyrgyz Republic, propelled by travel services, which as a country develops, as shown in figure 1.19. The share of increased by a factor of almost 18 between 2000 and 2010. In transport and travel services in total exports is only margin- Mongolia, exports of financial services increased by a factor ally associated with the level of development. For both sec- of 53. In Tajikistan, exports of communication services rose tors, the fitted value line is almost horizontal, meaning that by a factor of more than 12. Growth rates in the Caucasus the share of transport and travel services in a country’s total Module 1: Assessing Services Trade and Competitiveness Outcomes 27 Figure 1.17. Balance of Trade in Services in Selected Lower-Middle-Income Countries, by Sector, 2001–11 Module 1 a. Bolivia b. India Trade balance in services as Trade balance in services as 15 5 share of GDP 10 share of GDP 0 5 −5 0 −5 −10 2001 2003 2005 2007 2009 2011 2001 2003 2005 2007 2009 2011 Year Year c. Senegal d. Ukraine Trade balance in services as Trade balance in services as 10 10 share of GDP share of GDP 0 5 −10 0 −20 −5 −30 −10 2001 2003 2005 2007 2009 2011 2001 2003 2005 2007 2009 2011 Year Year Communications Financial Transport Travel Goods Source: World Development Indicators. Note: Income classification is as of July 1, 2012. Table 1.3. Growth Rate of Services Exports by Selected Countries in Central Asia, South Asia, and the Caucasus Region, by Sector, 2000–10 (rate) Region/country Communications Finance Transport Travel Central and South Asia Kazakhstan 2.9 6.0 3.9 1.8 Kyrgyz Republic 7.8 3.8 8.0 17.6 Mongolia 4.0 53.4 4.8 5.8 Pakistan 9.8 5.3 0.7 2.8 Tajikistan 12.1 6.9 0.1 1.8 Average 8.1 5.5 3.2 6.0 Caucasus Armenia 4.7 3.6 1.5 9.8 Azerbaijan 9.4 1.6 4.4 9.4 Georgia 2.4 4.9 3.5 3.7 Turkey –0.6 2.0 2.2 1.7 Average 4.0 3.0 2.9 6.2 Income group Low 1.7 4.7 2.5 2.0 Lower-middle 3.0 2.7 2.2 1.8 Source: UNCTADstat. Note: Income classification is as of July 1, 2012. 28 Valuing Services in Trade Figure 1.18. Sectoral Shares of Services Exports by Selected Countries in Central Asia, 2002–10 Module 1 a. Kazakhstan b. Kyrgyz Republic Percent of total services exports 100 90 80 70 60 50 40 30 20 10 0 2002 2003 2004 2005 2006 2007 2008 2009 2010 2002 2003 2004 2005 2006 2007 2008 2009 2010 c. Mongolia d. Tajikistan Percent of total services exports 100 90 80 70 60 50 40 30 20 10 0 2002 2003 2004 2005 2006 2007 2008 2009 2010 2002 2003 2004 2005 2006 2007 2008 2009 2010 Communication Finance Transport Travel Source: UNCTADstat. Note: Income classification is as of July 1, 2012. services export basket is not correlated with the level of devel- balance in transport, construction, insurance, financial, opment. Both communications and construction services and business services. One reason why these countries may seem to decline in importance as countries become richer. be less likely to run trade surpluses in these services is that In contrast, insurance, finance, computer, distribution high-income countries probably are net exporters of them. (which includes leasing), and business services all show a Higher per capita GDP for upper-middle-income countries positive correlation between the level of development and the is associated with larger positive trade balances in travel, share of total exports. (In the insurance, finance, and com- computer, and distribution services. puter sectors, some outliers—such as Luxembourg for finance Middle-income countries at all levels of income have and India for computer services—were taken out because been successful at exporting particular services. For exam- their high values would have given a distorted picture. Others ple, Bolivia, Mongolia, and Ukraine are situated low on outliers, such as the United Kingdom in financial services, the ladder of development but have large trade surpluses were included. In outlier countries, the skewed specialization for business services. This high ratio could be biased by pattern often reflects very specific policies or historical rea- the fact that least-developed countries often lack com- sons.) Had all outliers been omitted, the positive correlation plete data coverage, so that not all data flows are recorded. would have been even stronger. Some outliers are visible in Services are difficult to tax at the border. If least-developed each sector, possibly indicating successful policy choices. countries were included, the correlation between the ratio Figure 1.20 depicts the relationship between upper- and of exports to imports and per capita GDP would probably lower-middle-income countries’ trade balance and their have been positive. level of development. Only some sectors reveal a clear asso- The importance of sectoral exports can also be pre- ciation between the two variables. Upper-middle-income sented by showing the breakdown of services for each countries are likely to have a balanced or negative trade country and looking at how the composition changes Module 1: Assessing Services Trade and Competitiveness Outcomes 29 Figure 1.19. Sectoral Shares of Services Exports and GDP Per Capita in Selected Countries, 2009 Module 1 a. Business b. Communications c. Computers Share of exports in total Share of exports in total Share of exports in total 15 8 20 exports (percent) exports (percent) exports (percent) 6 15 10 4 10 5 2 5 0 0 0 4 6 8 10 12 4 6 8 10 12 4 6 8 10 12 Log of per capita GDP Log of per capita GDP Log of per capita GDP d. Construction e. Distribution f. Finance Share of exports in total Share of exports in total Share of exports in total 4 exports (percent) exports (percent) exports (percent) 6 8 3 6 4 2 4 1 2 2 0 0 0 4 6 8 10 12 4 6 8 10 12 4 6 8 10 12 Log of per capita GDP Log of per capita GDP Log of per capita GDP g. Insurance h. Transport i. Travel Share of exports in total Share of exports in total Share of exports in total 5 80 exports (percent) exports (percent) exports (percent) 40 4 30 60 3 20 40 2 1 10 20 0 0 0 4 6 8 10 12 4 6 8 10 12 4 6 8 10 12 Log of per capita GDP Log of per capita GDP Log of per capita GDP Source: Trade in Services Database. over time. The sectoral breakdown for a country can then Turkey’s services exports are also concentrated in a sin- be compared with that of its peers. Figure 1.21 shows the gle sector: travel services. The share of this sector peaked changes in the composition of services exports for India in 2004, at 66 percent of total services exports; in 2009, its and Turkey between 2001 and 2009. share was 63 percent. The share of tourism is much larger The bulk of India’s exports is concentrated in a single than in comparators: worldwide, travel services accounted sector—computer and information services—which for only 23 percent of total services exports in 2009. accounted for about 53 percent of its total services exports “Other business services” accounted for a relatively small in 2009. This share is much larger than in any of India’s share of Turkey’s services exports. In the prerecession year comparators. For lower-middle-income countries as a of 2007, its share of total services exports was 7 percent, whole, for example, computer and information services much lower than the 22 percent for upper-middle-income accounted for just 2.5 percent of total service exports. countries in the Europe and Central Asia region. For the world as a whole, computer and information ser- More detailed analyses provide a more nuanced picture. vices accounted for 4.7 percent of total services trade. Turkey, for example, has some small but very dynamic The second-largest services export sector in India is sectors. Exports of insurance, computer and information “other business services.” Traditional services sectors, such services, and post and telecommunication services experi- as transport and travel, play only a small role in India com- enced double-digit annual growth between 1994 and 2007. pared with other lower-middle-income countries or the Table 1.4 separates growth in the precrisis and crisis peri- world as a whole. ods, given the trade collapse of 2008–09, which affected 30 Valuing Services in Trade Figure 1.20. Trade Balance and Level of Development in Selected Upper- and Lower-Middle-Income Countries, 2009 Module 1 a. Business b. Communications c. Computers 10 8 Ratio of exports to Ratio of exports to Ratio of exports to 8 8 6 imports imports 6 imports 6 4 4 4 2 2 2 0 0 0 6 7 8 9 6 7 8 9 6 7 8 9 10 Log of per capita GDP Log of per capita GDP Log of per capita GDP d. Construction e. Distribution f. Finance 10 6 Ratio of exports to Ratio of exports to Ratio of exports to 6 8 imports imports imports 4 4 6 4 2 2 2 0 0 0 6 7 8 9 6 7 8 9 6 7 8 9 Log of per capita GDP Log of per capita GDP Log of per capita GDP g. Insurance h. Transport i. Travel 6 Ratio of exports to 10 8 Ratio of exports to Ratio of exports to 8 6 imports 4 imports imports 6 4 2 4 2 2 0 0 0 6 7 8 9 6 7 8 9 10 6 7 8 9 10 Log of per capita GDP Log of per capita GDP Log of per capita GDP Source: Trade in Services Database. Note: Dotted line indicates that ratio of exports to imports is 1. Solid line shows fitted values. Trade values of less than $1 million and ratios above 10 were excluded. primarily goods and only marginally services (Borchert the collapse, increasing more than 12 percent in 2010. and Mattoo 2009). It is possible, however, that some ser- Financial services also held strong, shrinking only margin- vices categories were affected more than others, given the ally, from 24 percent of total services export in 2009 to differences in the linkages between services and down- about 21 percent in 2010. Travel and transport also fared stream goods sectors. For this reason, the growth compu- relatively well in the United Kingdom during the great tations are separated. Notable in the 2008–09 period is the trade collapse. exceptional growth not only of certain sectors but also of royalties and license fees, which represent an increasingly sought after segment of services exports. Export Concentration Figure 1.22 explores this dynamic aspect of services. It compares an upper-middle-income country (Malaysia) with The Herfindahl-Hirschman and Theil Entropy Indexes a high-income country (the United Kingdom). Malaysia’s Several indicators are critical to assessing export concen- tration. They include measures of concentration (such main services exports are travel and transport services. as the Herfindahl-Hirschman Index) and dispersion (such as Before 2009, Malaysia saw a steady expansion of its business the Theil Entropy Index). Data on these indicators should services sector. During the great trade collapse of 2009, this be collected for specific years for all countries. Useful sources of information include the following: sector shrank by more than 73 percent. In contrast, both travel and transport services performed relatively well. • World Development Indicators (GDP) • IMF (balance of payments statistics) The United Kingdom specializes in business and • World Bank’s Trade in Services Database (see appendix B). financial services. Business services proved resilient to Module 1: Assessing Services Trade and Competitiveness Outcomes 31 Figure 1.21. Composition of Services Exports by India and Turkey, 2001–09 Module 1 a. India and comparators b. Turkey and comparators 100 100 Share of total services exports (percent) Share of total services exports (percent) 80 80 60 60 40 40 20 20 0 0 01 04 07 09 01 04 07 09 01 04 07 09 01 04 07 09 01 04 07 09 01 04 07 09 01 04 07 09 01 04 07 09 India LMC LMC SA World Turkey UMC UMC ECA World Transport Travel Communications Construction Insurance Financial Computer Royalties Business Personal Source: Trade in Services Database. Note: 01 = 2001; 04 = 2004; 07 = 2007; 09 = 2009. LMC SA = lower-middle-income country in South Asia; UMC ECA = upper-middle-income country in Europe and Central Asia. Table 1.4. Average Annual Growth of Services Exports, 1994–2007 and 2008–09 (percent) Upper-middle income Turkey Upper-middle income Europe and Central Asia World Sector 1994–2007 2008–09 1994–2007 2008–09 1994–2007 2008–09 1994–2007 2008–09 Transport 13.7 7.6 11.8 –6.7 13.0 –2.1 9.0 –7.3 Travel 11.7 7.2 10.5 2.3 13.7 2.2 7.4 –1.5 Post and telecommunications 12.0 11.9 6.5 –2.0 17.5 1.2 13.3 2.3 Construction –3.1 22.2 17.3 9.9 9.8 5.3 10.4 7.3 Insurance 38.1 1.4 6.2 –3.3 29.3 8.8 10.7 –6.0 Financial 9.3 8.4 22.0 –9.9 18.9 –10.1 16.4 –11.3 Computer and information 15.3 39.4 44.1 11.2 42.1 12.0 27.4 5.7 Royalties and license fees 1.6 25.6 36.9 –41.7 26.5 27.5 12.0 5.8 Other business services –0.3 –42.3 13.2 0.2 12.9 –1.3 12.1 –4.1 Total 8.3 4.3 11.9 0.5 12.9 0.8 10.1 –2.7 Source: Trade in Services Database. 32 Valuing Services in Trade Figure 1.22. Value of Services Exports in Malaysia and the United Kingdom, by Sector, 2000–10 Module 1 a. Malaysia b. United Kingdom 30,000 30,000 Exports (current dollars) Exports (current dollars) 20,000 20,000 10,000 10,000 0 0 00 01 02 03 04 05 06 07 08 09 10 00 01 02 03 04 05 06 07 08 09 10 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 Transport Travel Communications Construction Insurance Financial Computer Royalties Business Personal Source: Trade in Services Database. Diversification is important because it prevents econo- N representing the number of sectors in which the econ- mies from being too dependent on one sector. Despite the omy exports. The higher the value of the index, the more importance of diversifying, however, countries tend to spe- concentrated are the country’s exports. cialize. Software and business process services represent For easier interpretation, it is useful to compute the about a third of total exports by India. Brazil, Costa Rica, normalized version of this index. This index ranges from and Uruguay specialize in professional and information 0 to 1, with higher values indicating higher concentration. technology–related exports, Mexico in communication It is calculated as follows: and distribution services, and Chile in distribution and transportation services. In Morocco, Tunisia, Kenya, and H i* = (H i − 1 / N ) /(1 − 1 / N ). South Africa, exports of professional services to Europe are A second measure of concentration of exports is the growing. In Asia, the Philippines and Thailand are exporting Theil Entropy Index. This index is calculated as follows: health services, and Malaysia is increasing its export of edu- cation, Islamic banking, and telecommunication services. N One standard measure of concentration is the share of Ei = − ∑S k =1 ik log(Sik ) total exports by the most important sector. This index—the Herfindahl-Hirschman Index—is calculated as follows: where i stands for country, k for industry, and S for the N share in total exports. A larger value of this index (which Hi = ∑S k =1 2 ik ranges from 0 to infinity) indicates greater export diver- sification. A country that exports only one service would where Sik refers to the share of industry k in the total have a Theil Entropy Index of zero. If all services have exports of country i. The index ranges from 1/N to 1, with an equal share (n) in the export basket, the index would Module 1: Assessing Services Trade and Competitiveness Outcomes 33 have a maximum value that is −log(n). Figure 1.23 shows To answer these questions, figure 1.24 plots the Theil Module 1 the Herfindahl-Hirschman Index and the Theil Entropy Entropy Index against the log of GDP per capita PPP, Index for nine countries from different income groups. a proxy for development. No clear pattern is evident, It shows some changes between 2002 and 2010. For although export portfolios tend to be more diversified among instance, Cambodia’s services sector became relatively less higher-income countries. Higher-income countries also concentrated (more diversified) and Jordan’s export sector export more, as shown by the size of the circles in figure 1.24. became much less diversified (more concentrated). Low-income countries seem to have a lower Theil Entropy In the lowest-income countries, the number of services Index overall, but the index varies widely. The relationship exports and their values tends to be very low; adding just between development and more diversified exports becomes one sector can therefore change the index dramatically. weaker when developing countries are taken into account. In such cases, it is useful to look at simple shares, the num- ber of sectors involved, and the absolute levels of exports. Revealed Comparative Advantage Firm-level data can help determine concentration ratios at a more micro-level. Box 1.4 provides an illustration, using Assessing Revealed Comparative Advantage data on Romania. Several indicators are critical to assessing revealed com- Are services exports in developing countries concen- parative advantage. They include the value, shares, and trated in one or two sectors? Is Malaysia, where one sector growth rates of exports by sector. Data on these indicators should be collected for specific years for all countries. Useful dominates (see figure 1.22), typical of developing countries? sources of information include the following: Or is Malaysia a special case because travel is the dominant • World Bank’s Trade in Services Database (see appendix B) sector? Do all developed countries have an export range that • IMF (balance of payments statistics). looks like that of the United Kingdom (see figure 1.22)? Figure 1.23. Concentration and Diversification Indexes for Selected Countries, 2002 and 2010 a. 2002 b. 2010 Czech Canada Republic Czech Canada Republic Israel Israel Argentina Argentina Malaysia Bulgaria Bulgaria Cambodia Jordan Malaysia Cambodia Jordan Maldives Maldives 0 0.2 0.4 0.6 0.8 1.0 0 0.2 0.4 0.6 0.8 1.0 Herfindahl-Hirschman Index Theil Entropy Index Source: Trade in Services Database. 34 Valuing Services in Trade Module 1 Box 1.4. Using Firm-Level Data to Estimate Concentration Ratios: Evidence from Romania Concentration ratios can be calculated from firm-level data. Figure B1.4.1 shows a normalized Herfindahl-Hirschman Index for Romania. Romania’s exports are concentrated in several sectors, including recreation, computer and modern communication, and com- puter repair. Other important sectors include education, food, and administrative and support services. Another way to show concentration is to show the share of production of the top five firms in each sector (figure B1.4.2). By this measure, some services, such as computer and modern communications, are extremely concentrated. The top five firms in Romania represent almost 50 percent of output production in computer and modern communication. Figure B1.4.1. Herfindahl-Hirschman Index for Romania, by Sector, 2008 Recreation Computer and modern communications Computer repair Education Food services Administrative and support services Electricity, water, waste Real estate Health and social work Accommodation Transport and traditional communications Goods Other services Wholesale and retail trade Professionals Construction 0 0.02 0.04 0.06 0.08 Normalized Herfindahl-Hirschman Index Source: World Bank Structural Business Survey. Figure B1.4.2. Share of Output Produced by Top Five Firms in Romania, by Sector, 2008 Computer and modern communications Recreation Computer repairs Education Administrative and support services Food services Electricity, water, waste Health and social work Real estate Accommodation Transport and traditional communications Goods Other services Wholesale and retail trade Professional services Construction 0 10 20 30 40 50 Share of total output produced by top five firms (percent) Source: World Bank Structural Business Survey. Module 1: Assessing Services Trade and Competitiveness Outcomes 35 Figure 1.24. Relationship between Theil Entropy Index and Level of Development, 2010 Module 1 1.0 0.8 Rescaled Theil Entropy Index 0.6 0.4 0.2 0 7 8 9 10 11 Log of per capita GDP (PPP) Source: Trade in Services Database. Note: The size of the circles reflects the dollar value of exports. PPP = purchasing power parity. Sectoral data enable users to identify sectors in which a It shows that between 2002 and 2010, Argentina increased country has a competitive edge, or revealed comparative its RCA Index in travel, computer and information, advantage (RCA). The RCA Index compares the share of business, and personal services and reduced its RCA a sector’s exports in a country’s total exports with the aver- in communications. The share of transport services in age share of exports of all countries in the sector in total Argentina’s exports declined. world exports. The higher the ratio, which can range from As evident from table 1.5, comparative advantage is a zero to infinity, the more competitive the country is in the dynamic process. For business services, Argentina had an sector. RCA Index below 1 in 2002 but above 1 in 2010. While the The RCA Index is calculated follows: country increased its potential in this sector, it also main- xik tained an RCA Index above 1 in travel, communications, computer, and personal, cultural, and recreational services RCAik = Xi xwk Xw sectors. However, there is huge variation by services sector over time, as figure 1.25 shows for Costa Rica. where xik are country i’s exports of sector k, Xi are total Costa Rica had a strong export presence in travel ser- exports of country i, xwk are world exports of sector k, and vices in 2000. The RCA Index increased slightly before Xw are total world exports. An RCA Index above 1 indi- decreasing in 2009. Business services and computer ser- cates that a country’s share of services exports in a sec- vices became stronger over time, pointing to Costa Rica’s tor is larger than the global share of exports in that sector. strong export potential in these sectors. Other services, A high value of this indicator should not be confused with such as finance, still lag behind. net exports of a sector. A country could have an RCA The determinants that make or shape comparative Index below 1 but still be a net exporter of a service. The advantage are economywide variables on which firms in RCA Index helps identify sectors in which it is less costly specific services sectors can capitalize (see Module 3). Policy for a country to specialize based on “opportunity costs” variables such as regulation can also affect comparative (the cost associated with the next-best alternative). advantage. Other policies, such as undervalued exchange Table 1.5 gives the RCA Indexes for various services rates or subsidies, have less of an effect on comparative sectors, the real values of those exports, and the share advantage (subsidies aim to increase export volumes rather of those exports in total services exports in Argentina. than exploit productivity differences) (Siggel 2006). 36 Valuing Services in Trade Table 1.5. Value, Export Share, and Revealed Comparative Advantage Index of Services Exports by Argentina, 2002 and 2010 Module 1 2002 2010 Value Share Value Share Average annual (U.S. dollars) (percent) RCA (U.S. dollars) (percent) RCA growth (2002–10) Transport 1,538.0 21.3 0.9 4,042.9 15.4 0.7 12.8 Travel 3,070.2 42.5 1.5 9,883.8 37.7 1.8 15.7 Communications 300.0 4.2 1.8 670.9 2.6 1.0 10.6 Construction 46.7 0.7 0.3 126.6 0.5 0.2 13.3 Insurance 86.9 1.2 0.3 150.6 0.6 0.2 7.1 Financial 25.6 0.4 0.1 30.6 0.1 0.0 2.2 Computer 254.5 3.5 1.1 2,496.9 9.5 1.7 33.0 Royalties and licenses 276.9 3.8 0.6 270.0 1.0 0.1 0.3 Business 1,430.0 19.8 0.8 7,819.9 29.8 1.1 23.7 Personal, cultural, and recreational 190.7 2.6 1.8 711.6 2.7 2.7 17.9 Source: Trade in Services Database. Note: RCA = revealed comparative advantage. Figure 1.25. Revealed Comparative Advantage Index for Controlling for overall economic size, products exported Costa Rica, 2000, 2005, and 2009 by higher-income countries are ranked higher than prod- ucts exported by lower-income countries. These product- 3.0 specific calculations are then aggregated to construct the countrywide indexes of export sophistication. Advantage (RCA) Index Revealed Comparative 2.5 Specifically, let countries be indexed by i and products 2.0 by k. Let p be an export category (goods, manufacturing, 1.5 services). Total exports of category p by country i equal 1.0 X ip = ∑ x . Let Y denote the per capita GDP of country i. k p ik i Then the income level associated with product k in category 0.5 p equals the weighted average of per capita GDP, where 0 the weights represent the RCA of each country exporting 2000 2005 2009 that product. PRODY (the revealed income content of the Transport Travel Communications product) thus represents the income level associated with Finance Computer Business that product. PRODY is a weighted average of the per capita Source: Trade in Services Database. GDP of countries exporting that good. Weights in PRODY are based on the RCA. PRODY values of all products that a country exports are then weighted by each product’s share Export Sophistication in the country’s total export basket and summed to derive a country’s level of GDP per capita as inferred from the sophistication of its export basket (EXPY). Assessing Export Sophistication Several indicators are used to assess export sophistication. ∑ ∑ (x /x ) ∗Y . p Data on these indicators should be collected for specific (xik /xip ) years for all countries. Useful sources of information include PRODYkp = i i p p the following: ik i i • World Bank’s Trade in Services Database (see appendix B) p • World Development Indicators (GDP) The numerator of the weight, xik /xip, is the value share of • IMF (balance of payment statistics). the product in the country’s category p export basket. The denominator of the weight, ∑i (xik p /xip ), aggregates the value shares across all countries exporting that product in that The sophistication of exports, or EXPY (the revealed category. The PRODYs are used to compute the income income content of the export basket), is calculated for a level associated with each product k in category p (all goods, country’s export basket by ranking each service exported manufactured goods, or services). Specifically, EXPYi p is by the income levels of the countries that export the service. the average income level associated with all products in Module 1: Assessing Services Trade and Competitiveness Outcomes 37 a category exported by a country. It is computed as the An increase in EXPY indicates a country’s shift from Module 1 weighted average of all relevant PRODYs, where the weights low-PRODY to high-PRODY products. It means that the represent the share of the relevant product in the country’s share of high-PRODY goods, manufactures, and services export basket of category k: in the export basket has increased. Box 1.5 identifies some shortcomings of this measure. ⎛ xik ⎞ ∑ p EXPYi p = ⎜ p p ⎟ * PRODYk . There is a positive relationship between GDP per capita ⎝ xi ⎠ k and the sophistication of services exports (figure 1.26). Box 1.5. Services Sophistication: Handle with Care Calculating export sophistication (EXPY) is a two-stage process. The first stage is to measure the income level associated with each service (or product) (PRODY). The PRODY of a particular service is the gross domestic product (GDP) per capita of the typical coun- try that exports that service. Typical GDP is calculated by weighting the GDP per capita of all countries exporting the service. The weight given to each country is based on its revealed comparative advantage (RCA). The PRODY for a single service is calculated by weighting the GDP per capita of all countries exporting that service. Therefore, a service that typically makes up a large percentage of a poor country’s export basket will be more heavily weighted by poor countries’ GDP per capita. The second stage is to measure the income associated with a country’s export basket as a whole (EXPY). The EXPY is calculated by weighting the PRODYs of each service by the share each service contributes to total exports. For example, if tourism makes up 15 percent of a country’s exports, its PRODY is given a weight of 0.15. Countries whose export baskets are made up of “rich-country goods” will have higher EXPYs; export baskets made up of “poor-country goods” will have lower EXPYs. The concepts of PRODY and EXPY are not free of criticism (see Lederman and Maloney 2012). To begin with, classification of services trade is not well developed. In the balance of payments statistics, for example, a professional services category may include activities that are entirely different across countries. Furthermore, categories can include a wide range of services. An activity such as accounting may refer to a highly commoditized activity (such as preparing simple tax returns) or a highly sophisticated service (such as providing tax advice to a multinational cor- poration); there is no way to differentiate. If these problems are not taken into account, EXPY may overestimate the importance of sophisticated products from low-income countries. The sophistication of a particular country’s export basket may be increasing over time even though the country does not show an RCA in high-value-added services exports, such as information and communication technology (ICT) or business-related services, as illustrated in the table and figure in this box. Chile shows an RCA Index greater than 1 only for transportation services as shown in table B1.5.1. Yet figure B1.5.1 shows that Chile’s services exports sophistication is above what would be expected given its level of development. In India, the high RCA in ICT exports and to a lesser extent “other business services” accounts for the sophistication of its services exports. Table B1.5.1. Revealed Comparative Advantage Index of Selected Countries for Goods and Services Exports, 2009–11 Brazil Chile Egypt, Arab Rep. Hungary India Malaysia Philippines South Africa Ukraine Goods 1.1 1.2 0.7 1.1 0.7 1.1 0.9 1.2 1.0 Transportation 0.5 2.0 3.9 0.9 0.9 0.6 0.5 0.4 2.9 Travel 0.5 0.5 5.2 1.0 0.7 1.7 0.9 2.0 1.2 Other commercial services 0.8 0.3 0.7 0.8 2.1 0.4 1.4 0.3 0.7 Communication 0.3 0.4 3.2 0.7 0.7 5.6 1.0 0.4 1.3 Information and communication technologies 0.1 0.1 0.3 0.9 11.7 0.6 2.6 0.3 0.6 Construction 0.0 — 2.6 0.7 0.3 0.8 0.2 0.1 0.7 Financial 0.6 0.0 0.3 0.1 0.9 0.0 0.1 0.6 0.4 Government (not indicated elsewhere) 1.8 0.4 1.1 0.3 0.3 0.1 — 1.1 2.3 Insurance 0.4 0.8 0.4 0.1 1.0 0.4 0.2 0.6 0.1 Other business 1.4 0.5 0.6 1.0 1.2 0.4 2.2 0.2 0.8 Personal, cultural, and recreation 0.3 0.8 1.4 6.8 0.8 2.4 0.4 0.5 0.9 Royalties and license fees 0.2 0.1 — 0.7 0.0 0.1 0.0 0.0 0.1 Source: IMF various years. Note: — Not available. (continued on next page) 38 Valuing Services in Trade Module 1 Box 1.5. (continued) Figure B1.5.1. Export Sophistication and GDP in Selected Countries, 2009–11 16 Sophistication of services exports IND 15 CHL HUN PHL UKR ZAF (log of EXPY) 14 EGY 13 MYS 12 11 6 8 10 12 Log of per capita GDP (PPP) Source: IMF 2009. Note: CHL = Chile, EGY = Arab Republic of Egypt, EXPY = (revealed) income content of export basket, HUN = Hungary, IND = India, MYS = Malaysia, PHL = Philippines, PPP = purchasing power parity, UKR = Ukraine, ZAF = South Africa. The composition of services exports moves in the direction that countries above the fitted line will see higher economic of modern activities, such as ICT, professional services, growth rates in the future. and other higher value added services that are normally exported by high-income countries. Trade models based on comparative advantage are agnostic about whether wel- Revealed Factor Intensities fare gains accrue from producing a particular product or service. Using the RCA Index therefore reveals little about Physical, human, and ICT-related capital are relevant for services. Useful sources of information on them include the which sector most growth would come from. following: In general, higher value added goods and services have • World Bank’s Trade in Services Database (see appendix B) higher PRODYs. Figure 1.26 shows that many lower- • IMF (balance of payments statistics) income countries exhibited little change in the type of • Barro and Lee (2012) data set on educational attainment • Conference Board data on physical and ICT capital. goods they exported. In contrast, the PRODYs of Latin American countries rose steadily between 2001 and 2007 (this trend ended with the onset of the global financial The Revealed Human Capital Index (RHCI) is an crisis) (figure 1.27). adjusted version of the RCA Index multiplied by a The apparent decline in the sophistication of services country’s factor endowments. It can range from zero to exports between 2007 and 2009 was not necessarily driven infinity. Human capital is the average years of education. by a change in the composition of services exports of these Physical and ICT-related capital is the capital stock per countries; it may instead reflect the construction of the capita for a country corrected using the perpetual inven- sophistication measure (see box 1.5). Moreover, figures tory method. The index for human capital is calculated as 1.26 and 1.27 say nothing about the direction of causation follows where k denotes product, i denotes country, and between export sophistication and per capita GDP. H denotes average years of education: Yet, one can note that India, for instance, exports the type of services that would normally appear in the export ⎛ xik ⎞ portfolio of higher-income countries (see figure 1.26). ⎜ ⎝ xi ⎟⎠ Based on a large pool of countries and exported goods, RHCI k = ∑i * Hi . xik ∑i x Hausmann, Hwang, and Rodrik (2007) find that it is likely i Module 1: Assessing Services Trade and Competitiveness Outcomes 39 Figure 1.26. Sophistication of Services Exports and Per Capita GDP in Selected Countries, 2001 and 2009 Module 1 a. 2001 b. 2009 10.0 10.0 LUX LUX Sophistication of services exports (log of EXPY) Sophistication of services exports (log of EXPY) SLB IRL IND IRL 9.5 IND USA JPN JPN SGP ROM PRY USA 9.5 BRA DEU PNG SGP ROM CMR BRA DEU PRY MOZ AGO COG CHL MUS PNG RUS MYS CHNCRI RUS CHN CRI LBN CHL SLE GHAPAK SYC CMR MUS LBY ATG HND MEX SLE MYS 9.0 SLB SYC MOZ KEN ATG GHA MEX RWA LBY YEM 9.0 MAC HND AGO PAK KEN YEM 8.5 RWA DJI BGD BGD 8.5 DJI 8.0 LBR 6 8 10 12 6 8 10 12 Log of per capita GDP (PPP) Log of per capita GDP (PPP) Fitted values High-income country Upper-middle-income country Low-income country Lower-middle-income country Source: Trade in Services Database. Note: Income classification is as of July 1, 2012. For the International Organization for Standardization (ISO) abbreviations, see appendix B. PPP = purchasing power parity, PRODY = (revealed) income content of product basket. Figure 1.27. Sophistication of Services Exports by Six Countries in Latin America, 2001–09 9.6 Sophistication of services exports 9.5 (log of EXPY) 9.4 9.3 9.2 9.1 2001 2002 2003 2004 2005 2006 2007 2008 2009 Argentina Brazil Mexico Colombia Peru Chile Source: Trade in Services Database. Note: EXPY = (revealed) income content of export basket. 40 Valuing Services in Trade The indexes for the revealed physical capital (RPCI) and Figure 1.28 uses the revealed factor intensities calculated Module 1 the revealed information and communications technology– using the equation for RHCIk. It shows a positive relation- related capital (RICTI) are calculated as follows, where K is ship between human capital and the export value of ser- the physical capital stock, ICT is the ICT-related capital stock, vices. In both South Africa and Turkey, services exports are and L is population: associated with a higher index of human capital. The size of ⎛ xik ⎞ each sector (the circle) indicates the extent to which a sector ⎝ xi ⎟ ⎜ ⎠ Ki is intensive in ICT capital, showing a strong correlation with RPCI k = ∑i * the level of ICT capital and human capital within sectors. xik Li ∑i xi Table 1.6. Factor Intensities of Selected Services Exports ⎛ xik ⎞ ICT High-skill labor ⎝ Xi ⎟ ⎜ ⎠ ICTi Service intensity Service intensity RICTI k = ∑i xik * Li . Travel 0.30 Travel 0.17 ∑i Transportation 0.30 Transportation 0.20 Xi Personal services 0.30 Personal services 0.19 Other business 0.43 Other business 0.24 Another more direct measure of factor intensities for Insurance 0.28 Insurance 0.21 services can be taken from the EU KLEMS (O’Mahony and Information and Information and Timmer 2009), which uses a methodology based on growth communications communications accounting. This database of growth and productivity covers technology 0.25 technology 0.20 mostly developed countries. Table 1.6 shows factor intensi- Finance 0.39 Finance 0.25 ties, which are calculated for each sector directly rather than Construction 0.27 Construction 0.16 computed based on revealed factor intensities. The sectors Communication 0.23 Communication 0.25 are ranked from high to low, based on factor calculations, Source: EU KLEMS (O’Mahony and Timmer 2009). using the methodology of Shepherd and van der Marel (2013). Note: ICT = information and communication technology. Figure 1.28. Revealed Factor Intensities for Human and ICT-Related Capital in Selected Services Sectors in South Africa and Turkey, 2009 a. South Africa b. Turkey 15 15 Revealed factor intensity for human capital Revealed factor intensity for human capital 10 10 5 5 0 0 0 2 4 6 8 10 0 5 10 Log of export value of services Log of export value of services Source: Trade in Services Database. Note: The size of each circle indicates the revealed information and communication technology (ICT) capital intensity of each sector, which tends to be higher the more intensive human capital services become. Module 1: Assessing Services Trade and Competitiveness Outcomes 41 Table 1.7 shows that the indexes for both human capi- Export Markets Module 1 tal (RHCI) and ICT-related capital (RICTI) have signifi- The ability to diversify should be judged not only in cant effects on future growth, which is used as a dependent terms of the sectoral distribution of services exports but variable. Countries specializing in services that are inten- also in terms of the markets importing these services. sive in factor inputs that are usually abundant in developed Exporting to a range of markets not only helps mitigate countries thus have better growth prospects. the effects of asymmetric shocks that might hit some but Table 1.7. Regression Results of Future Growth between not all importers, it also helps exporters improve the 1999 and 2008 on Revealed Human Capital Index (RHCI) and competitiveness of their products. Revealed ICT-Related Capital Index (RICTI) Success can be evaluated along two dimensions. The Item (1) (2) first, the intensive margin, looks at how volumes of log (RHCI) 1.283*** exports to existing partners developed over time. It evalu- –0.266 ates the depth of the relationship with each importer. log (RICTI) 0.701*** The second, the extensive margin, looks at the num- –0.222 ber of existing relationships and markets and examines log GDP pc “99 –7.003*** –9.696*** their dynamics over time. It assesses whether exports are –0.366 –0.471 growing because a particular country is exporting to new Observations 2,790 1,983 markets or new products. R-squared 0.154 0.337 Figure 1.29 depicts one way of analyzing the intensive Note: log GDP pc “99 = log of per capita GDP in 1999. Cross-country margin. It shows that between 2002 and 2010, Germany regression used robust clustered standard errors. *** p < 0.01. maintained its position as the top importer of Turkey’s Figure 1.29. Intensive Margin of Turkey, 2002 and 2010 a. 2002 b. 2010 0.25 0.25 0.20 0.20 Share in total export Share in total export 0.15 0.15 0.10 0.10 0.05 0.05 0 0 Fr m he ium de m Au es es n Bu ain Bu ain ly ly Au ds m y ng n ce Un en ia d rk ia Ki ny e s D aria k ria nd ite ar an nc tio tio D tr ar Ita Ita do o at at ite ma iu an n st Ru ted rma Sp Sp Un nm n ngd s N Fra lg lg rla m rla N elg St ra St lg ra er Be de he d e e B Ki G Un G Fe Fe et et d ite n ia ia i Un ss ss Ru Source: Trade in Services Database. 42 Valuing Services in Trade services, despite some decline in its imports of Turkish could be the strong transition dynamics experienced there, Module 1 goods. The United Kingdom’s share of services exports possibly in combination with the expected accession to the from Turkey did not change much. In 2010, other coun- European Union. tries, such as Italy and the Netherlands, were also impor- Services exports to unknown markets increased in tant export destinations for Turkish services. Bulgaria also 2002, however, suggesting that the observed dynamics may increased in importance. These findings suggest that reflect data rather than market issues. Data problems are Turkey may have expanded to new markets or that coun- particularly problematic in analyzing the extensive mar- tries with previously low levels of services imports from gin for export markets (that is, services sectors rather than Turkey became more important export destinations. trading partners). The bilateral data included in appendix B provide These problems notwithstanding, the data can help insights on the dynamics of the extensive margin. They understand the dynamics and relative position of pairs allow researchers to assess the total number of relationships of trading partners. Researchers should keep an eye on as well the number by sector. For trade in goods, detailed the residual category, however, and probably focus only disaggregated data are available for most trading pairs. For on the most recent years, for which bilateral coverage is services, the quality of bilateral data varies, for the reasons strongest. described earlier. The Trade in Services Database offers the Figure 1.31 plots the intensive margin against the exten- best coverage of services trade. sive margin for Mexico, the Slovak Republic, and Turkey. Figure 1.30 looks at the number of trading relationships It shows that over time, Mexico relied much more on its for Mexico and the Slovak Republic. Mexico added part- intensive margin than its extensive margin. One explana- ners rather steadily until the 2008 crisis. The changes in the tion for this result could be Mexico’s strong ties with the Slovak Republic were more erratic. Part of the explanation United States through the North American Free Trade Figure 1.30. Extensive Margins of Mexico and the Slovak Republic, 1994–2010 a. Mexico b. Slovak Republic 40 100 80 30 Number of trading partners Number of trading partners 60 20 40 10 20 0 0 20 1 20 1 20 9 00 20 8 20 9 10 19 8 20 9 20 0 20 3 20 3 20 8 20 9 10 19 4 02 04 19 4 19 5 19 6 19 8 95 20 5 02 04 19 6 20 5 20 6 06 97 07 97 20 7 0 0 9 9 9 0 0 0 0 0 0 0 9 9 9 9 9 0 9 0 0 0 20 20 20 19 20 19 19 19 20 19 20 20 Retained trading partner New trading partner Source: Trade in Services Database. Module 1: Assessing Services Trade and Competitiveness Outcomes 43 Figure 1.31. Extensive and Intensive Margins of Mexico, the Slovak Republic, and Turkey, 2010 Module 1 2001 2002 1999 1.0 2000 Intensive margin (percent of world exports) 2003 2004 2005 2003 2005 2004 2009 2006 2006 2009 2002 2007 2008 2008 0.5 2000 2001 2007 1999 2005 2006 2002 2004 2008 2007 2009 2001 2003 2000 1999 0 20 40 60 80 100 Extensive margin (number of trading partners) Mexico Slovak Republic Turkey Source: Trade in Services Database. Agreement (NAFTA), which may reduce its incentive to Computing the Trade Complementarities Index (TCI) explore new markets. Over time, the extensive margins can help identify markets with export potential. It iden- in the Slovak Republic and to some extent in Turkey had tifies whether a potential importer buys services that an broader reach than their intensive margins. Generally, economy exports abroad. The TCI complements the RCA years of trade experience translates into more trading Index. Ideally, a country would export those services in partners, although this relationship is bumpy. For most which it has a comparative advantage and import services countries, there is a positive relationship between the two in which it has a comparative disadvantage, as outlined in margins over time. Cadot, Carrère, and Strauss-Kahn (2011). The TCI is calculated by summing the (absolute) differ- ence between a country’s share of imports of a service in Trade Intensity, Trade Complementarities, and Gravity its total imports and the exporter’s share of a service in its Models total exports: ⎡ i − xkj ⎤ ∑ Indicators Used to Measure Trade Opportunities mk TCI ij = 100 ⎢1 − ⎥ Several indicators are critical to assessing trade opportu- 2 nities. They include the Trade Complementarities Index, ⎢ ⎣ k ⎥ ⎦ the Trade Intensity Index, and the difference between predicted and actual services exports (determined using a where k stands for services sectors, i for countries that theory-grounded gravity model). Data on these indicators should be collected for specific years for all countries. Useful import the service, and j for countries that export the ser- sources of information include the following: vice. In this index, which ranges from 0 to 100, a higher • World Bank’s Trade in Services Database (see appendix B) score indicates that two countries are a better “fit” as • Centre d’Etudes Prospectives et d’Informations Internationales trading partners and that their trade baskets are comple- (CEPII) (gravity model–related variables) mentary. A lower score indicates that two countries are • World Development Indicators • Stata. similar in their trade structure and therefore lack scope to develop additional bilateral exports. 44 Valuing Services in Trade Figure 1.32. Trade Complementarities Index between Figure 1.33. Trade Intensity Index for Turkey and Selected Module 1 Malaysia and Selected Trade Partners, 2004–10 Trade Partners, 2008 Malaysia 1.5 TCI (Trade Complementarities Trade Intensity Index (scale x-z) 80 70 1.0 Index) 60 50 0.5 40 2004 2006 2008 2010 United States Germany Japan 0 Thailand Turkey Russian Germany France Australia United Japan Federation States Source: Trade in Services Database. Source: Trade in Services Database. Figure 1.32 displays the TCI for Malaysia and five of its States, despite the fact that these countries are in a different trading partners. It shows that Malaysia shares higher trade income group and specialize in different sectors. Turkey’s complementarities with Germany than with Thailand, a exports to Germany represent a much larger share of country in the same region and income group as Malaysia. its total exports than the share of the rest of the world’s Thailand seems to have an export structure that is simi- exports to Germany. lar to Malaysia’s. The TCI can help policy makers evalu- Japan’s TII is much lower, even though Japan’s import ate whether trade barriers are preventing a country from structure should make it an ideal trading partner for reaching the full potential provided by trade opportunities Turkey. Why is Turkey’s trade with Japan not greater? Is with a trading partner or whether there is a scope to nego- the fact that Turkey is geographically closer to Germany tiate a bilateral trade agreement to reduce barriers. It can than it is to Japan relevant? also help in the analysis of potential gains from a bilateral A gravity model (box 1.6) can be used to assess or regional trade agreement. whether a country is actually exporting what it “should” A second index for measuring trade opportunities is be exporting to its partners. Malaysia exports more to the Trade Intensity Index (TII). It focuses on total trade Australia because Australia is geographically closer than rather than trade by sector. most of Malaysia’s other partner countries. In contrast, The TII is measured as country i’s exports to country Germany imports a large share of Turkey’s exports j relative to its total exports divided by world exports to because Germany has a very large market. These factors country j relative to total world exports. It is calculated as can be taken into account by performing a regression follows: analysis using a gravity model. The gravity model does a good job of predicting xij Malaysia’s trade relationships with partner countries TII ij = Xi xwj Xw (figure 1.34). Market size and distance are good deter- minants of Malaysia’s strong export relationships with where xij are country i’s exports to country j; Xi are total Australia and Japan. Malaysia trades less than would be exports of country i; xwj are world exports to country j, expected with Poland, Portugal, and Turkey. It trades and Xw are total world exports. The TII thus indicates the more than predicted with France. extent to which a country is present in the export market of The reasons for such deviations could lie outside a trading partner. standard gravity explanatory variables, which this time Figure 1.33 displays the TII for Turkey. It shows high are not incorporated in the regression but could be values for Germany, France, and the Russian Federation. included. Policy barriers that are not captured by the dis- It shows low values for Australia, Japan, and the United tance proxy—such as specific tariffs or domestic logistics Module 1: Assessing Services Trade and Competitiveness Outcomes 45 Module 1 Box 1.6. What Is a Gravity Model? The gravity equation—introduced by Nobel laureate Jan Tinbergen, in 1962—is one of the most successful empirical tools in international trade. It posits that bilateral trade is explained largely by (a) distance between any pair of countries (which proxies for transport costs, policy barriers, behind-the-border barriers, and informational asymmetries) and (b) the market size or mass of the exporting and importing countries (proxied by the countries’ GDPs). The gravity equation often includes additional control variables, such as common language, border, currency, and legal systems and whether the countries have signed a free trade agree- ment. Based on these factors, it predicts the value of exports between countries and compares it with the actual value of trade. Recent advances in the empirical trade literature have refined the gravity model in various ways. First, the volume of trade between partners depends not only on direct (or absolute) trade costs between partners but also on the (indirect) relative trade costs with respect to third countries. These “multilateral resistance terms,” developed by Anderson and van Wincoop (2003), should be incorporated into the model. Second, several new techniques allow analysts to take stock of zero-trade flows. One way of doing so is to include the Heckman sample selection correction method, which takes the probability of being included in the sample as an explanatory variable into account. When performing ordinary least squares regressions, researchers usually drop zero observations by taking their logs. Such a procedure can create sample selection bias, as the probability of having nonzero trade flows could be correlated with GDP or distance. Another way of correcting for zero trade flows is by using the Poisson pseudo-maximum likelihood (PPML) estimator, introduced by Santos Silva and Tenreyro (2006). Figure 1.34. Gravity Model for Malaysia and Selected Trade Figure 1.35. Trading Partners and Development, Partners, 2008 2008 10 11.0 log of predicted services exports Weighted log of per capita GDP (PPP) of importer 8 10.5 USA 6 GBR JPN DEU ITA 10.0 BEL FRA NLD 4 ESP DNK SWE IRL AUT POL FIN CYP LUX 9.5 2 PRT ROU SVK CZE TUR HUN LVA EST HRV 0 SVN 9.0 0 2 4 6 8 10 8 9 10 11 12 log of actual services exports Log of per capita GDP (PPP) of exporter Source: Trade in Services Database. Source: Trade in Services Database. Note: For the International Organization for Standardization (ISO) abbrevia- Note: PPP = purchasing power parity. tions, see appendix B. performances (Arvis and others 2010)—may explain why is no more likely to trade with a low-income or high- countries “over” or “under” trade with Malaysia. income country. The previous section established that the type of ser- A limitation of figure 1.35 is that the Trade in Services vice a country exports matters. Does the destination also Database does not record all trade relationships for ser- matter? Higher-income countries export more sophisti- vices. Part of the reason why the data are spotty is that cated services than do lower-income countries. It may services are intangible and therefore difficult to tax at therefore make sense for developing countries to try to the border, particularly in developing countries. For this specialize in these services. A similar argument could be reason, trade in services, particularly between developing made for export markets. Figure 1.35 shows that there is countries, is probably underestimated, possibly explaining no clear relationship between the levels of development why there are more dots on the right side of figure 1.35 and trading partners—that is, a high-income country than on the left. 46 Valuing Services in Trade Notes Iacovone, Leonardo, Aaditya Mattoo, and Andrés Zahler. 2013. “Trade Module 1 and Innovation in Services: Evidence from a Developing Economy.” 1. For the purpose of this toolkit, modern services are services that Policy Research Working Paper 6520, World Bank, Washington, DC. can be traded across borders without the buyer and seller being in the IMF (International Monetary Fund). 2009. Balance of Payments and same place. Delivery of these services is less dependent on physical International Investment Position Manual, 6th ed. Washington, DC: IMF. infrastructure and more dependent on telecommunications and ———. Various years. Balance of Payments Statistics. electric supply. Examples of such services include communication, Washington, DC. http://elibrary-data.imf.org/finddatareports banking, insurance, business, and remote access services; transcrip- .aspx?d=33061&e=170784. tion of medical records; call centers; and education. These services Lederman, D., and W. Maloney. 2012. “Does What You Export Matter? differ markedly from traditional services, which demand face-to-face In Search of Empirical Guidance for Industrial Policies.” World Bank, interaction. Washington, DC. 2. Trade figures do not reflect the true importance of services trade. McKinsey. 2010. Lions on the Move: The Progress and Potential of African When trade is measured in terms of value added, the share of services may Economies. McKinsey Global Institute, Washington, DC. increase by as much as 50 percent by some estimates (Francois, Manchin, Melitz, Marc. 2003. “The Impact of Trade on Intra-Industry Reallocations and Tomberger 2013). and Aggregate Industry Productivity.” Econometrica 71 (6): 1695–725. Milberg, William, and Deborah Winkler. 2013. Outsourcing Economics: Global Value Chains in Capitalist Development . Cambridge: References Cambridge University Press. OECD (Organisation for Economic Co-operation and Development) and Anderson, J. E., and E. van Wincoop. 2003. “Gravity with Gravitas: A WTO (World Trade Organization). 2013. “OECD-WTO Database on Solution to the Border Puzzle.” American Economic Review 93 (1): Trade in Value Added First Estimates.” January 16, Organisation for 170–92. Economic Co-operation and Development, Paris. Arvis, Jean-François, Monica Alina Mustra, Lauri Ojala, Ben Shepherd, O’Mahony, Mary, and Marcel P. Timmer. 2009. “Output, Input and Daniel Saslavsky. 2010. Connecting to Compete 2010: Trade and Productivity Measure at the Industry Level: The EU KLEMS Logistics in the Global Economy. The Logistics Performance Index and Its Database.” Economic Journal 119 (538): 374–403. Indicators. Washington, DC: World Bank. Sachs, Jeffrey D., and Andrew M. Warner. 1999. “The Big Push, Natural Barro, Robert J., and Jong-Wha Lee. 2012. “A New Dataset of Educational Resource Booms and Growth.” Journal of Development Economics Attainment in the World, 1950–2010.” NBER Working Paper 15902, 59 (1): 43–76. National Bureau of Economic Research, Cambridge, MA. Santos Silva, J. M. C., and S. Tenreyro. 2006. “The Log of Gravity.” Review Bernard, A., and J. B. Jensen. 1999. “Exceptional Exporter Performance: of Economics and Statistics 4 (88): 641–58. Cause, Effect, or Both?” Journal of International Economics 47 (1): Shepherd, Ben, and Erik van der Marel. 2013. “International Tradability 1–25. of Services.” Policy Research Working Paper 6712, World Bank, Borchert, Ingo, and Aaditya Mattoo. 2009. “The Crisis-Resilience of Washington, DC. Services Trade.” Policy Research Working Paper 4917, World Bank, Siggel, Eckhard. 2006 “International Competitiveness and Comparative Washington, DC. Advantage: A Survey and a Proposal for Measurement.” Journal of Brienlich, Holger, and Chiara Criscuolo. 2011. “International Trade Industry, Competition and Trade 6 (2): 137–59. in Services: A Portrait of Importers and Exporters.” Journal of Taglioni, Daria, and Deborah Winkler. 2014. Making Global Value Chains International Economics 84: 188–206. Work for Development. Washington, DC: World Bank. Cadot, Olivier, Céline Carrère, and Vanessa Strauss-Kahn. 2011. “OECD Tinbergen, Jan. 1962. Shaping the World Economy: Suggestions for an Imports Diversification of Suppliers and Quality Search.” Policy International Economics Policy . New York: Twentieth Century Research Working Paper 5627, World Bank, Washington, DC. Fund. Cattaneo, Olivier, Gary Gereffi, Sébastien Miroudot, and Daria Taglioni. Trade in Services Database. World Bank, Washington, DC. 2013. “Joining, Upgrading and Being Competitive in Global UNCTADstat Database. http://unctadstat.unctad.org/ReportFolders Value Chains: A Strategic Framework for Developing Countries.” /reportFolders.aspx. World Bank Policy Research Paper 6406, Washington, DC. WEF (World Economic Forum). 2013. Enabling Trade: Valuing Growth Francois Joseph, Miriam Manchin, and Patrick Tomberger. 2013. Opportunities. Geneva: World Economic Forum. “Services Linkages and the Value Added Content of Trade.” Policy Wignaraja, Ganeshan, and Ashely Taylor. 2003. “Benchmarking Research Working Paper 6432, World Bank, Washington, DC. Competitiveness: A First Look at the MECI.” In Competitiveness Freund, Caroline, and Diana Weinhold, 2002. “The Internet and in Developing Countries: A Manual for Policy Analysis, edited by International Trade in Services.” American Economic Review 92 (2): Ganeshan Wignaraja. London: Routledge. 236–40. World Development Indicators (database). World Bank, Washington, Hausmann, Ricardo, Jason Hwang, and Dani Rodrik. 2007. “What You DC. http://data.worldbank.org/data-catalog/world-development Export Matters.” Journal of Economic Growth 12 (1): 1–25. -indicators. Module 2 Services as a Source of Competitiveness in the Whole Economy Before a country can export services, it needs to develop a purchashing power parity (PPP) for two periods, 2000–02 competitive domestic services sector. Cross-country com- and 2008–10. The fitted line reflects the stylized fact that parisons of basic indicators, such as the share of a country’s services tend to grow in importance for the economy as the value added from services activity, provide a first assess- level of a country’s development rises. ment of the existence and importance of a domestic sector. Figure 2.1 indicates whether the services sector is larger or smaller than predicted based on a country’s level of development. It shows that most developed countries Services in the Domestic Economy are highly dependent on services, even controlling for per capita income. In contrast, per capita value added Assessing the Role of Services in the Domestic of services are much smaller in countries such as Peru, Economy which falls below the fitted line. Several indicators are critical to assessing the role of services in the domestic economy. They include the value added of Value added also reveals substantial differences in the the services, industry, and agriculture sectors as a percent size of the services sector at all levels of development. of gross domestic product (GDP) versus the log of GDP per Although the share of services in GDP generally increases capita; employment by sector; and the openness of services trade. Data on these indicators should be collected for spe- with economic development, it varies across economies cific years for all countries. World Development Indicators at the same level of development. The share of services is provides the following information: about 70 percent in Brazil, for example, but just 40 percent • value added by sector as a percent of GDP in Malaysia, even though per capita income is similar in • GDP and GDP per capita • employment per sector the two countries. • total services trade. Comparison of the two panels of figure 2.1 shows that there has been little change in the relative positions of most countries, although all countries, including low-income Table 2.1 shows a simple cross-country regression countries, have become somewhat richer. between the share of services value added in GDP in 1985 Other structural factors may help explain these results. and the share of services trade in GDP in 2010. It shows that Figure 2.2 plots the residuals of the regression shown in the size of the domestic services sector is a robust precon- table 2.2, which regresses the log of per capita value added dition for specialization in services exports in later years. of services against the log of population and the log of per However, this relationship weakens when more recent data capita income in PPP. The population variable does not are used for the independent variable, suggesting that over show a significant correlation with value added per cap- time, other forces, such as policy, play roles. ita, but GDP per capita does. These variables do not fully explain the wide variance in the performance of coun- tries; figure 2.2 illustrates the large residuals. Within each Value Added of Services income group there is a correlation between the log of per Figure 2.1 shows scatter plots of per capita value added capita services value added and the residuals, suggesting of services (in logs) against the log of GDP per capita that the more important services are for the economy, the 47 48 Valuing Services in Trade Table 2.1. Cross-Country Regression of Size of Services countries at the second stage in manufacturing, and coun- Sector against Specialization in Services Export tries at the most advanced stage of development in ser- Item ln(STR perc GDP 2010) vices. This pattern is captured by the slopes of the curves in log (SVA perc GDP “85) 0.837*** figure 2.3, which plots log of per capita value added of each (0.280) of these sectors against level of development. Observations 111 R-squared 0.085 In 2012, the share of services value added represented 75 percent of total value added in high-income econo- Note: STR perc GDP 2010 = services trade as percent of GDP in 2010. SVA perc GDP “85 = services value added as percent of GDP in 1985. Robust mies. It accounted for 49 percent of value added in low- clustered standard errors are in parentheses. *** p < 0.01. income countries and 56 percent in upper-middle-income countries. In many low- and lower-middle-income coun- tries in Africa, the figure is just 20 percent. The countries larger are the residuals. This result is somewhat weaker in with the smallest shares of services tend to be countries more recent years. Nonetheless, the correlations seem to with a history of conflict, such as Angola, Iraq, and Libya. imply that other structural factors explain the growth of The only country classified as high income that has a services as countries develop. very small share of services in value added is Equatorial Module 2 This pattern of specialization across stages of develop- Guinea. ment was first observed in the late 1930s. Fisher (1939, Since 1990, the share of services in value added has 1952) and Clark (1940) observed that countries at the increased steadily across all income groups (figure 2.4). first stage of development tend to specialize in agriculture, Even low-income economies, which have not yet attained Figure 2.1. Per Capita Services Value Added and Per Capita GDP in Selected Countries, 2000–02 and 2008–10 a. 2000–02 b. 2008–10 12 12 Log of per capita value added of services Log of per capita value added of services 10 10 CHL MEX MEX COL 8 CHL 8 PER COL PER 6 6 4 4 2 2 4 6 8 10 12 4 6 8 10 12 Log of per capita GDP (PPP) Log of per capita GDP (PPP) Fitted values High-income OECD and non-OECD Low income Upper-middle income Lower-middle income Source: World Development Indicators. Note: Income classification is as of July 1, 2012. In the WDI database, services correspond to International Standard Industrial Classification (ISIC) divisions 50–99. The sector is derived as a residual (GDP less agriculture and industry) and may not reflect the sum of services output. For some countries, values include product taxes (minus subsidies) and may reflect statistical discrepancies. CHL = Chile, COL = Colombia, MEX = Mexico, OECD = Organisation for Economic Co-operation and Development, PER = Peru, PPP = purchasing power parity. Module 2: Services as a Source of Competitiveness in the Whole Economy 49 Figure 2.2. Services Value Added and Structural Factors in Selected Countries, 2000–02 and 2008–10 a. 2000–02 b. 2008–10 12 12 Log of per capita value added of services Log of per capita value added of services 10 10 CHL MEX MEX 8 8 COL CHL PER COL PER 6 6 Module 2 4 4 2 2 −3 −2 −1 0 1 −3 −2 −1 0 1 Ordinary least squares residuals Ordinary least squares residuals Fitted values High-income OECD and non-OECD Low income Upper-middle income Lower-middle income Source: World Development Indicators. Note: CHL = Chile, COL = Colombia, MEX = Mexico, OECD = Organisation for Economic Co-operation and Development, PER = Peru. their full production potential in industry, have seen the Structural characteristics, such as the level of regulation, services sector grow (box 2.1). the availability of specific skills, and urbanization, also These observations suggest that it is difficult to say matter. In India the services that were most underdevel- whether a country’s share of services in the economy is oped were also the most regulated and grew fastest after small or large without putting other characteristics in deregulation, according to Eichengreen and Gupta (2011, context. Income is only a partial explanation of why the 2012). Part of the reason the services sector is less devel- services sector thrives in some countries but is relatively oped in China than in India is that India is more urban- underdeveloped in others. ized, according to Liping and Evenett (2010). The tradability of a service also matters. When trading is not burdened by regulatory obstacles, tradable services tend to grow faster than their nontradable counterparts. Table 2.2. Cross-Country Regression of Per Capita Services Amin and Mattoo (2006, 2008) show that services output Value Added against Level of Economic Development is greater in Indian states with larger bases of skilled work- Item ln(services VA pc) ers and that countries with good institutions have larger ln(population) –0.0236* and more dynamic services sectors (see Module 4). (0.0141) ln(gdp_pc_ppp) 1.330*** Figure 2.5 shows the sectoral composition of value (0.0222) added expressed in GDP in several Central Asian econo- Observations 329 mies in 1990, 2000, and 2010. Over this period, the ser- R-squared 0.919 vices share increased in some of these economies. By 2010, Root mean square deviation 0.510 it represented the main economic activity in three out of Note: Standard errors are in parentheses. * p < 0.1, *** p < 0.01. the five economies (Tajikistan, the Kyrgyz Republic, and 50 Valuing Services in Trade Figure 2.3. Per Capita Value Added of Agriculture, Industry, and Services versus Per Capita GDP, 2008–10 a. Agriculture b. Industry 12 10 Log of per capita Log of per capita CHL value added value added 10 CHL 8 MEX COL COL 8 PER PERMEX 6 6 4 4 2 6 8 10 12 6 8 10 12 Log of per capita GDP (PPP) Log of per capita GDP (PPP) c. Services 10 Log of per capita value added 8 Module 2 CHL 6 COL MEX PER 4 2 6 8 10 12 Log of per capita GDP (PPP) Fitted values High-income OECD and non-OECD Low income Upper-middle-income Lower-middle income Source: World Development Indicators. Note: Income classification is as of July 1, 2012. Linear regression is based on a quadratic GDP term. Each panel shows a fitted line of the scatter plot. The lines above and below it are the 95 percent confidence intervals. CHL = Chile, COL = Colombia, MEX = Mexico, OECD = Organisation for Economic Co-operation and Development, PER = Peru, PPP = purchasing power parity. Uzbekistan). In Azerbaijan and Turkmenistan, industry agriculture was accompanied by a decline in the share of is the main economic sector, and the share of services in services and an increase in the share of industry. In Turkey the economy declined over the past two decades. This pat- and Tunisia, the value added from services increased sig- tern probably reflects the fact that these economies are nificantly. Their economic structure is similar to that of largely dependent on the gas and petroleum sector, which other upper-middle-income countries. prevents them from climbing the ladder of comparative The services sector is still the main economic activity in advantage. Other factors, such as infrastructure or regula- almost all countries in the region for which data are avail- tory problems, could also play roles. able. The exceptions to this trend are Iraq and Libya in 2000, In the lower-middle-income countries of the Kyrgyz where the industrial sector represented the main economic Republic and Uzbekistan, the relative importance of agri- activity. Lebanon has the most services-intensive economy culture shrank. However, in Uzbekistan the growth of in the region, with a services share of 71 percent in 2010. services was accompanied by the simultaneous growth of industry, which was not the case in the Kyrgyz Republic Employment or Tajikistan. The continuously growing services sector in Uzbekistan may be directly related to its industry. The services sector is also an important source of employ- The steady increase in the share of services over the past ment. Unfortunately, employment data are very limited; the two decades was not experienced in many countries in the latest observations from World Development Indicators Middle East and North Africa, including the Arab Republic for a large number of countries are from 2005. Nevertheless, of Egypt, the Islamic Republic of Iran, and the Syrian Arab figure 2.7 shows that services are the dominant employer in Republic (figure 2.6). In all countries in the region, the developed countries and that their importance has been siz- relative importance of agriculture shrank considerably. able in middle-income countries. As much as 71 percent of Except in Lebanon, Tunisia, and Turkey, the decline of the workforce in high-income countries was employed in Module 2: Services as a Source of Competitiveness in the Whole Economy 51 Figure 2.4. Sectoral Share of Value Added, by Income Group, 1990, 2000, and 2010 100 90 Share of value added (percent) 80 70 60 50 40 30 Module 2 20 10 0 LIC LMC UMC HIC LIC LMC UMC HIC LIC LMC UMC HIC 1990 2000 2010 Agriculture Industry Services Source: World Development Indicators. Note: LIC = low-income country; LMC = lower-middle-income country; UMC = upper-middle-income country; HIC = high-income country. Data for 2009 instead of 2010 were used for high-income countries. Box 2.1. Does India Defy the General Pattern of Ladders of Comparative Advantage? In general, countries first develop a strong domestic services sector before they begin to export services. Not all countries follow this path, however. India, for example, did not have a large domestic services sector before it became a major provider of international call services. Its cost and skill advantage led straight to a strong external sector. India also has a relatively large pool of people in the agricultural sector who, because of policies, cannot move into industry. In the meantime, India’s services sector has expanded. services in 2005 (up from 65 percent in 1994). In 2005, the For example, in Senegal, Vietnam, and to a lesser extent share of employment generated by services was 31 percent Ghana, which are lower-middle-income countries, their sec- in lower-middle-income and 40 percent in upper-middle- toral shares of employment are close to the average for their income countries (up from 27 percent in lower-income and income group. In contrast, in Cambodia, Mozambique, 29 percent in upper-middle-income countries in 1994). and Tanzania, which are low-income countries, the agri- Data are not available for low-income countries. cultural sectors account for a large share of employment, Employment by a sector can be assessed by comparing especially relative to industry. In Kenya, the services sector it with employment in other countries at the same level is in line with the average low-income country. of income. In Senegal, a lower-middle-income country, Assessing the employment absorption capacity of each for example, the services sector employs 43 percent of the sector and comparing them to the value added contri- labor force. In most other peer countries, agriculture is the bution can also be insightful. Figure 2.3 showed that the main employer (figure 2.8). The services sector accounts richer a country becomes, the more value added in ser- for just 16 percent of employment in Mozambique and 19 vices it produces. Figure 2.9 suggests that employment in percent in Tanzania. services rises with value added. Figure 2.8 reveals that the sectoral employment shares Figure 2.9 shows that the services sector gener- of a few countries are in line with their development. ates less employment in Turkey than in countries with 52 Valuing Services in Trade Figure 2.5. Value Added of Agriculture, Industry, and Services as Share of GDP in Central Asian Countries, 1990, 2000, and 2010 Value added of sector as percent of GDP 100 90 80 70 60 50 40 30 20 Module 2 10 0 AZE KGZ TJK TKM UZB AZE KGZ TJK TKM UZB AZE KGZ TJK TKM UZB 1990 2000 2010 Agriculture Industry Services Source: World Development Indicators. Note: AZE = Azerbaijan, KGZ = Kyrgyz Republic, TJK = Tajikistan, TKM = Turkmenistan, UZB = Uzbekistan. For the Kyrgyz Republic, data for 1992 instead of 1990 were used. Figure 2.6. Value Added of Agriculture, Industry, and Services as Share of GDP in Selected Countries in the Middle East and North Africa, 1990, 2000, and 2010 100 Value added of sector as percent of GDP 90 80 70 60 50 40 30 20 10 0 1990 2000 2010 1990 2000 2010 1990 2000 2010 1990 2000 2010 1990 2000 2010 1990 2000 2010 1990 2000 2010 EGY IRN JOR LBN SYR TUN TUR Agriculture Industry Services Source: World Development Indicators. Note: EGY = Arab Republic of Egypt, IRN = Islamic Republic of Iran, JOR = Jordan, LBN = Lebanon, SYR = Syrian Arab Republic, TUN = Tunisia, TUR = Turkey. Data were not available for Iraq in 1990 and 2010 or Libya in 1990. Module 2: Services as a Source of Competitiveness in the Whole Economy 53 Figure 2.7. Employment in Agriculture, Industry, and Services, by Income Group, 1994, 2000, and 2005 100 90 Share of total employment (percent) 80 70 60 50 40 30 20 Module 2 10 0 LMC UMC HIC LMC UMC HIC LMC UMC HIC 2005 2000 1994 Agriculture Industry Services Source: World Development Indicators, accessed in February 2013. Note: LMC = lower-middle-income country; UMC = upper-middle-income country; HIC = high-income country. Figure 2.8. Employment in Agriculture, Industry, and Services in Selected Low- and Lower-Middle-Income Countries, 2006 100 90 Share of total employment (percent) 80 70 60 50 40 30 20 10 0 Ghana Kenya Cambodia Mozambique Senegal Tanzania Vietnam Agriculture Industry Services Source: World Development Indicators, accessed in February 2013. Note: Data for Kenya are for 2005; data for Mozambique are for 2003. 54 Valuing Services in Trade Figure 2.9. Share of Value Added and Share of Employment in Selected Countries, by Sector, 2008–10 a. Agriculture b. Industry Percent of total employment 80 40 Percent of total employment generated by sector generated by sector 60 30 SYR TUR EGY 40 20 JOR EGY 20 TUR SYR 10 0 JOR 0 0 20 40 60 0 20 40 60 80 Percent of total value added generated by sector Percent of total value added generated by sector Module 2 c. Services Percent of total employment 100 generated by sector 80 JOR 60 SYR TUR 40 EGY 20 20 40 60 80 100 Percent of total value added generated by sector Source: World Development Indicators, accessed in February 2013. Note: Each panel shows a fitted line of the scatter plot. The lines above and below it are the 95 percent confidence intervals. EGY = Arab Republic of Egypt, JOR = Jordan, SYR= Syrian Arab Republic, TUR = Turkey. similar levels of value added generated by the services in total exports) is linked with countries’ economic sector. Jordan employs an above-average share (nearly complexity. Figure 2.10 plots the Economic Complexity 80 percent) of its workforce in services, with another Index (ECI) constructed by Hausmann and others (2014) 18 percent employed in industry. In Syria and Egypt, against the share of services value added exports in total employment in the services sector is about average for the value added exports. The intuition behind the ECI is that size of their services sector. making a product requires a particular type and mix of Firm-level data are another source with which to knowledge. Countries that possess more knowledge can assess the importance of services in the domestic economy produce a more diverse set of products. In other words, (box 2.2). the amount of embedded knowledge of a country is rep- resented in the number of distinct products that it makes. In addition, products that are knowledge-intensive are Domestic Services Sectors in Relation to International only produced in countries where the required knowl- Trade edge is available (Hausmann and others 2014). Although A final measure for assessing the importance of the ECI refers essentially to goods trade characteristics, the domestic services sector is its relation to international figure 2.10 shows that complexity is also positively corre- trade. As shown in table 2.1, there is a positive relation- lated with modern services exports contribution to total ship between the size of the domestic services sector exports. Generally, the ECI is positively correlated with and its competitiveness in trade. Building on this fact, development (income per capita), and so the contribu- figure 2.10 shows the degree to which value added of ser- tion of services exports in total trade either measured in vices exports (expressed as a percentage of value added terms of backward or forward linkages (defined below) Module 2: Services as a Source of Competitiveness in the Whole Economy 55 Box 2.2. What Is the Structure of the Services Sector? Evidence from Romania Firm-level data provide insights into the structure of the services sector. Figure B2.2.1 shows the shares of employment, value added, output, and exports of goods, traditional services, modern services, and professional services in Romania. Figure B2.2.1. Concentration of Employment, Value Added, Output, and Exports in Romania, by Sector, 2008 a. Goods sector b. Traditional services 100 100 Share of total Share of total 80 80 (percent) (percent) 60 60 40 40 20 20 0 0 Module 2 0 20 40 60 80 100 0 20 40 60 80 100 Percent of all firms Percent of all firms c. Modern services d. Professional services 100 100 Share of total Share of total 80 80 (percent) (percent) 60 60 40 40 20 20 0 0 0 20 40 60 80 100 0 20 40 60 80 100 Percent of all firms Percent of all firms Employment Output Value added Total exports Source: World Bank Structural Business Survey. The structure of the sectors varies substantially. For goods, most of the indicators are distributed in a uniform manner: all four lines are close to the 45-degree line. To some extent, this is also the case for the traditional services sector, although a small number of firms accounts for a substantial share of employment (almost 40 percent). In the modern services sectors, however, less than 20 percent of firms account for almost 80 percent of output, and about 50 percent of firms account for almost 80 percent of exports. For professional services exports, about 5 percent of firms account for almost 50 percent of exports. will be higher. In other words, the ability to successfully participate in services exports is linked to the existing Assessment of the services content of goods is based on methodologies proposed by Francois, Manchin, and type and mix of knowledge available in a country. Tomberger (2013) and Francois and Woerz (2008). They include the value added versus direct share, the share of a country’s value added in world value added, Linkages to the Rest of the Economy and forward and backward linkages of a region or country. Data on these indicators should be collected for specific years for all countries. Useful sources of information include Assessing the Value Added of Services the following: Several indicators are critical to assessing the value added of services and their linakges to other economic activi- ties. Assessing the importance of the services sector for • OECD’s Database for Structural Analysis (input-output the overall economy is based on the value added content tables) that is generated by services and goes into downstream • OECD-WTO’s Trade in Value Added (TiVA) or Global exported sectors. This methodology helps appraise the rel- Trade Analysis Project (GTAP) databases evance of specific services inputs for the overall competi- • World Bank’s Export of Value Added Database (see tiveness of a country and for particular types of production. appendix A). 56 Valuing Services in Trade Figure 2.10. Knowledge and Services Exports Contribution to Total Exports, 2007 a. Backward linkages b. Forward linkages 0.35 0.45 Modern services value added export share Modern services value added export share 0.40 0.30 0.35 0.25 0.30 0.20 0.25 0.15 0.20 0.15 0.10 0.10 Module 2 0.05 0.05 0 0 –3 –2 –1 0 1 2 3 –3 –2 –1 0 1 2 3 Economic Complexity Index Economic Complexity Index Sources: Hausmann and others 2014; Export of Value Added Database. Services can be exported in two ways: as a final service only for intermittent years. They are also subject to restric- or as an intermediate input in another good or service by tive assumptions that may significantly bias the quantifi- any downstream industry. Most trade analysis examines cation of the technology coefficients assumed for different gross values. However, policy makers are more and more types of production. interested in what a country produces and exports in net Disentangling value added from gross figures is terms, because goods and services cross borders many important because it provides insights into the impor- times. Net or value added analysis therefore avoids double tance of the domestic content of trade (appendix A counting. explains the methodology employed to calculate trade Methodologies assessing the relevance of services as in value-added terms). Figure 2.11 breaks down gross inputs to downstream exports also present challenges. and net exports of machinery and business services by These methodologies are based on value added and input- income groups. (Business services were selected because output data. Combined with trade data, they provide this sector has been the most dynamic services subsector an indication of the reliance on domestic versus foreign in recent years.) inputs of production destined for domestic and foreign Across income groups, the share of value added in markets. exports is much larger than the gross share for services The OECD-WTO’s Trade in Value Added (TiVA) compared to goods. For both machinery and business Database provides net and gross value added figures for services, high-income countries export the most value exports of goods and services. Most countries in the data- added, although lower-middle-income countries are not base are Organisation for Economic Co-operation and far behind. For both machinery and business services, Development (OECD) countries; most developing coun- lower-middle-income countries export more value added tries are left out. than upper-middle-income countries. This section uses the World Bank’s Export of Value Added Database of gross and net production and export Gross versus Net Value-Added Share figures based on input-output tables provided by the Global Trade Analysis Project (GTAP) Database. GTAP tables A cleaner output measure is therefore the net value pro- provide a more aggregated picture than other sources, but duced in a sector. It shows the real importance of a good or they include many developing countries and are available service for the economy. A useful indicator that can reveal Module 2: Services as a Source of Competitiveness in the Whole Economy 57 Figure 2.11. Gross Exports and Value Added of Exports of Machinery and Business Services, by Income Level, 2007 a. Machinery b. Business services 30 30 20 20 Share Share 10 10 Module 2 0 0 High-income Low-income Lower-middle Upper-middle High-income Low-income Lower-middle Upper-middle country country -income -income country country -income -income country country country country Gross exports Value added exports Source: Export of Value Added Database. this importance is the ratio of the net share to the gross form of intermediate inputs; it is not brought forward by share of value added: these sectors themselves as they produce relatively little value added. In contrast, higher-income countries seem VAki to export more value added produced by their own goods sector. Net-gross ratio = VAi GRki This relationship disappears in the services sector GRi (panel b). The fitted value line is flat, indicating that on average, developing countries produce and bring forward where VAki is the exported value added in a sector k in as much value added in services as developed countries. country i; VAi is the total exported value added of country The ratio of the shares of value added to gross value is high i; GRki is the gross value of exports in sector k in country for most countries, indicating that evaluating services at i; and GRi is the total gross value exported of country i. gross values is a misplaced way of looking at the impor- A ratio higher than 1 means that looking at the net value tance of the sector to the economy as a whole. added term is the right way of judging a sector’s contri- bution to the national economy. The indicator also shows Value Added of Production for Domestic and how much the sector is underestimated in gross terms. Export Markets Figure 2.12 plots this ratio for each country against the level of development for a goods and a services sector. How can exports be calculated in value added terms? How In panel a, this ratio is less than 1 for many countries, can one estimate how much value added a country carries indicating that the manufacturing sector tends to have a over from other sectors? How can one estimate how much much higher component of gross exports than value added. value added a country brings forward into other sectors? Panel a also shows a positive relationship between the net- This section uses input-output information from the gross ratio and the level of development. This relationship Export of Value Added Database to calculate forward implies that most of the value added in goods exported and backward linkages of value added. Forward link- by developing countries comes from other sectors, in the ages indicate how much value added is carried by other 58 Valuing Services in Trade Figure 2.12. Ratio of Value Added to Gross Value and Development for Machinery and Business Services in Selected Economies, 2007 a. Machinery b. Business services 2.0 2.5 CHN 2.0 1.5 Value added share/gross share Value added share/gross share MEX CRI MYS BRA IND JPN DEU HKG MOZ RUS 1.5 CHL DEU HKG CHL USA 1.0 JPN MEX Module 2 RUS USA BRA IRL CHN CRI IRL PRY MYS GHA 1.0 0.5 IND MOZ GHA 0.5 PRY 0 7 8 9 10 11 7 8 9 10 11 Log of per capita GDP (PPP) Log of per capita GDP (PPP) Fitted values High-income country Upper-middle-income country Low-income country Lower-middle-income country Source: Export of Value Added Database. Note: BRA = Brazil, CHL = Chile, CHN = China, CRI = Costa Rica, DEU = Germany, GHA = Ghana, HKG = Hong Kong SAR, China, IRL = Ireland, JPN = Japan, MEX = Mexico, MOZ = Mozambique, MYS = Malaysia, PPP = purchasing power parity, PRY = Paraguay, RUS = Russian Federation, USA = United States. sectors—that is, how important each (services or goods) the domestic production of other goods and services. sector is as an intermediate input for other sectors. Business services represent about 18 percent of value Backward linkages show how much value added a sector added production. The figures for backward linkages are carries from other sectors—that is, how much the sector similar in the two sectors. For all sectors except dwell- embodies inputs that it will further process. ings, transport and trade services are the most important These indicators are usually available as an aggregate for source of inputs (after inputs from the sector itself). This the world or country (the Export of Value Added Database pattern is also evident in the value added that is inte- includes about 120 economies). Value added can be calcu- grated into Albania’s exports. Most of this value added lated for the domestic economic activity and for a country’s stems from or uses transport and distribution services. exports. The value added of exports indicates how much The pattern of value added for Albania is not charac- value added a country, income group, or region carries for- teristic of all countries. In high-income countries, business ward through its exports to other countries (forward link- services and government services (such as health and edu- ages) or how important a given sector is as a channel of cation) are the most used and carried forward in domes- “indirect exports” for its domestic inputs (backward link- tic production. Value added integrated in exports that ages) (see appendix A). are further processed or consumed abroad is done mostly Table 2.3 shows these indicators for Albania, where through business services (26 percent); heavy industry almost 32 percent of value added production comes goods, such as machinery (20 percent); and machinery from transport and trade services, which are used in equipment (19 percent). Table 2.3. Value Added of Domestic Production and Exports in Albania, by Sector, 2007 (percent of total) Electricity, gas, Construc- Transport Business Government Dwel- Primary Food and Machinery Other Forward Sector and water tion and trade services services lings products beverages Textiles equipment machinery linkages Domestic value added Electricity, gas, and water 0.99 0.19 0.18 0.31 0.02 0.02 0.04 0.07 0.06 0.00 0.15 2.04 Construction 0.00 7.77 0.09 0.06 0.07 0.09 0.01 0.01 0.00 0.00 0.01 8.12 Transport and trade 0.05 1.98 25.83 0.96 0.23 0.09 0.59 0.80 0.37 0.08 0.84 31.79 Business services 0.07 1.10 3.34 11.76 0.16 0.27 0.39 0.33 0.21 0.05 0.51 18.21 Government services 0.00 0.01 0.04 0.14 8.25 0.00 0.01 0.00 0.00 0.00 0.01 8.47 Dwellings 0.00 0.00 0.00 0.00 0.00 3.94 0.00 0.00 0.00 0.00 0.00 3.94 Primary products 0.04 0.66 2.03 0.82 0.07 0.03 9.33 3.22 0.21 0.01 0.79 17.21 Food and beverages 0.00 0.02 0.16 0.14 0.01 0.00 0.08 2.12 0.01 0.00 0.07 2.62 Textiles 0.00 0.01 0.01 0.01 0.00 0.00 0.00 0.00 1.14 0.00 0.01 1.19 Machinery equipment 0.00 0.02 0.02 0.01 0.00 0.00 0.00 0.00 0.00 0.43 0.00 0.48 Other machinery 0.00 1.88 0.33 0.15 0.05 0.06 0.03 0.05 0.03 0.02 3.31 5.91 Backward linkages 1.17 13.63 32.04 14.36 8.85 4.51 10.49 6.60 2.05 0.59 5.71 100.00 Export value added Electricity, gas, and water 0.04 0.01 0.25 0.45 0.01 0.00 0.12 0.02 0.20 0.01 0.45 1.56 Construction 0.00 0.23 0.13 0.09 0.02 0.00 0.02 0.00 0.02 0.00 0.03 0.54 Transport and trade 0.00 0.06 34.75 1.41 0.06 0.00 0.43 0.23 1.27 0.20 2.25 40.66 Business services 0.00 0.03 4.49 17.33 0.04 0.00 0.31 0.11 0.70 0.13 1.35 24.49 Government services 0.00 0.00 0.06 0.21 2.20 0.00 0.02 0.00 0.01 0.00 0.02 2.52 Dwellings 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 Primary products 0.00 0.02 2.74 1.20 0.02 0.00 6.64 0.95 0.72 0.03 1.89 14.21 Food and beverages 0.00 0.00 0.22 0.20 0.00 0.00 0.03 0.65 0.03 0.00 0.27 1.40 Textiles 0.00 0.00 0.02 0.01 0.00 0.00 0.00 0.00 3.88 0.00 0.04 3.95 Machinery equipment 0.00 0.00 0.02 0.01 0.00 0.00 0.00 0.00 0.00 1.16 0.01 1.21 Other machinery 0.00 0.06 0.44 0.23 0.01 0.00 0.03 0.02 0.12 0.05 8.52 9.46 Backward linkages 0.04 0.40 43.11 21.15 2.36 0.00 7.61 1.98 6.94 1.59 14.81 100.00 Source: Export of Value Added Database. 59 Module 2 60 Valuing Services in Trade Box 2.3. Where Do Services Inputs Go? Evidence from Ghana How important are services as inputs in the domestic economy and as exports? Who are the downstream users of services inputs? Social accounting matrixes (input-output data) from the Export of Value Added Database developed by the World Bank as a cross-country dataset spanning intermittent years from 1992 to 2007 shed light on these questions. This database contains two tables—a domestic value added table and an export value added table—that identify the value added contribution of particular inputs to sectors that either sell the final good to the domestic market or export it (table B2.3.1). Policy analysts can find use in such tables at a more disaggregated level. In Ghana, for instance, four services sectors (utilities, construction, transport and trade, and business services) together con- tributed more than a third to domestic value added and almost 36 percent to export value added. Trade and transport services accounted for a larger share of export value added (23 percent) than domestic value added (14 percent). The same pattern is evident for business services (which contributed 12 percent to export value added and 8 percent to domestic value added) but not manufacturing (which contributed 17 percent to export value added and 28 percent to domestic value added). Construction and agriculture are much more domestically oriented in terms of the value added they produce. Table B2.3.1. Value Added of Exports and Domestic Production in Ghana, by Sector, 2007 (percent of total value added) Module 2 Manufacturing Services Agriculture Sector Domestic Exports Domestic Exports Domestic Exports Agriculture 17.8 11.3 0.0 0.0 85.9 85.8 Manufacturing 59.3 63.6 3.1 0.5 0.6 0.4 Services Utilities 1.6 1.8 2.2 1.8 0.3 0.4 Construction 0.4 0.2 46.5 0.9 0.0 0.0 Transport and trade 18.6 20.5 17.5 46.4 10.3 10.0 Business 1.6 1.6 27.6 46.4 1.1 1.8 Other (for example, public services) 0.8 0.7 1.3 3.6 0.6 0.7 Total services 22.1 24.2 93.9 95.5 11.8 12.1 Source: Export of Value Added Database. Table B2.3.1 shows that most services inputs in Ghana are used for production in other services sectors (93.9 percent for domestic production and 95.5 percent for exports). Within services, construction services are used mostly as inputs in domestic production, whereas transport and trade and business services are used mainly in exports. The downstream manufacturing sector is the second-most important user of services inputs (mainly transport and trade). This pattern may change as Ghana climbs up the supply chain. A second important sector providing inputs to Ghana’s manufacturing industry is agriculture. This pattern reflects Ghana’s specialization in the manufacture of products closely connected to agriculture, such as processed cocoa beans. This structure is visible in figure 2.13, which identifies but the actual value added “produced” by a specific sector is the forward and backward linkages for machinery and much lower, as indicated by the lower bars for forward link- business services for each income group. Middle- and high- ages of machinery in figure 2.13. For services, the short bars income countries carry a substantial share of inputs back- for backward linkages in business services reflect the fact that ward in machinery. For services, this pattern is reversed, most services are produced domestically and then exported so that across all income groups, forward linkages prevail as inputs for further production elsewhere. The difference over backward linkages. Box 2.3 illustrates the different between exports of high-income and lower-income country roles that services play as inputs in the case of Ghana. groups is much greater in services than in goods, indicat- Gross values and backward linkages are more important ing that high-income countries have a dominant position in for goods, whereas net values and forward linkages are more services, which they use as inputs for production. relevant for services. Goods, especially machinery, include Figure 2.14 plots selected countries’ shares of forward inputs from many other sectors, from either the domestic and backward linkages in a goods sector (machinery) economy or abroad. When the value of a good is recorded at and a services sector (business services). It shows that face value, it eventually constitutes an amount of trade mea- services are different from goods for most countries. For sured in gross numbers, amounting to a larger share in total goods (panel a), most countries fall above the 45-degree trade each time it crosses the border. The value added to a line, indicating that value added is brought backward product or service by each sector is also taken into account, mainly through goods. In contrast, for services (panel b), Module 2: Services as a Source of Competitiveness in the Whole Economy 61 Figure 2.13. Forward and Backward Linkages for Machinery and Business Services, by Income Group, 2007 Share of value added in total exports (percent) a. Machinery b. Business services Share of value added in total exports (percent) 0.25 0.25 0.20 0.20 0.15 0.15 0.10 0.10 0.05 0.05 Module 2 0 0 High Upper- Lower- Low High Upper- Lower- Low income middle middle income income middle middle income income income income income Income group Income group Forward linkages Backward linkages Source: Export of Value Added Database. Figure 2.14. Backward and Forward Linkages for Goods and Services in Selected Countries, 2007 a. Goods (machinery) b. Services (business services) 0.6 0.5 0.4 0.4 Backward linkages Backward linkages MYS MEX 0.3 CHN CRI THA 0.2 0.2 ROM BLR TUN TUR BGR 0.1 ZAF LTU BRA LVA MUS MDV COL ARG RUS NAM ECU URY PER BWA CHL KAZ 0 IRN VEN PAN AZE 0 0 0.1 0.2 0.3 0.4 0 0.1 0.2 0.3 0.4 0.5 Forward linkages Forward linkages 45° line High-income country Upper-middle-income Lower-middle-income country Low-income country country Source: Export of Value Added Database. Note: For the International Organization for Standardization (ISO) abbreviations, see appendix B. 62 Valuing Services in Trade most countries fall below the 45-degree line, indicat- a–c show each sector’s direct value added share plus its ing that value added is carried forward mainly through forward linkages divided by the gross share of exports services. of that same sector. Panels d–f show the same ratio for The role of forward and backward linkages can be backward linkages. explored for other services sectors separately. Not all ser- Figure 2.15 shows considerable variation across sectors. vices have the same importance for value added; some For forward linkages, both business and construction ser- services could have a more important direct or gross export vices show a flat relationship with per capita GDP, mean- component. Not all services have the same role as inputs, ing that the value added component is as important as the either backward or forward. These values may also differ gross share of exports across the level of development. For across countries or income groups. transport and trade services, the relationship is negative, To shed some light on the differences among services, meaning that the gross share of exports is more important the direct value added shares of a sector plus its forward the richer a country becomes. Put differently, value added and backward linkages is separately divided by the gross in this sector is carried forward in larger part by lower- share of trade of the relevant sector. Figure 2.15 shows income countries. For backward linkages, figure 2.15 the outcome of such an exercise for business, construc- shows a slight positive relationship for both business Module 2 tion, and transport and trade services. This measure is services and construction and a negative relationship related to the level of development of a country. Panels for transport and trade. Thus, apart from transport and Figure 2.15. Direct Value Added plus Linkages/Gross Share and Per Capita GDP in Selected Economies, 2007 a. Forward linkages of b. Forward linkages c. Forward linkages of business services of construction transport and trade 15 15 6 Value added share + forward Value added share + forward Value added share + forward linkages/gross share linkages/gross share linkages/gross share 10 10 4 JPN USA 5 5 2 IND DEU DEU USA IRL IND JPN HKG IND USA DEU HKG HKG 0 IRL 0 JPN 0 7 8 9 10 11 7 8 9 10 11 7 8 9 10 11 Log of per capita GDP (PPP) Log of per capita GDP (PPP) Log of per capita GDP (PPP) d. Backward linkages of e. Backward linkages f. Backward linkages of business services of construction transport and trade 2.0 2.0 1.6 Value added share + backward Value added share + backward Value added share + backward 1.4 linkages/gross share linkages/gross share linkages/gross share 1.5 1.5 DEU 1.2 IND JPN DEU USA IND JPN JPN HKG 1.0 HKG USA IND DEU IRL 1.0 IRL 1.0 USA IRL 0.8 HKG 0.6 0.5 0.5 7 8 9 10 11 7 8 9 10 11 7 8 9 10 11 Log of per capita GDP (PPP) Log of per capita GDP (PPP) Log of per capita GDP (PPP) Fitted values High-income country Upper-middle-income country Lower-middle-income country Low-income country Source: Export of Value Added Database. Note: DEU = Germany; HKG = Hong Kong SAR, China; IND = India; IRL = Ireland; PPP = purchasing power parity; USA = United States. Module 2: Services as a Source of Competitiveness in the Whole Economy 63 trade, there is no clear-cut relationship between value uses the direct value added share as the denominator. Panels added and the gross value component and a country’s a–c show the ratio of forward linkages to value added. Panels level of development. Figure 2.16 shows the importance d–f show the ratio of backward linkages to value added. of value added versus gross shares for the transport and No clear pattern emerges for either business services or trade sector for several economies over time. Panel a construction (see figure 2.17, panels a, b, d, and e). These shows that for value added plus forward linkages, most results seem to suggest that forward or backward linkages developing countries except India (after 1995) show an are as important for high-income countries as they are for upward trend. India and the United Kingdom reduced developing countries. their presence as forward carriers for inputs of transport In the transport and trade sector, forward linkages are and trade services. This pattern is in line with figure 2.15, somewhat less important the more developed a coun- which shows that on the whole, higher-income countries try becomes, although the relationship is not very strong. show a similar pattern. An upward trend is also evident Backward linkages are more important for higher-income for Argentina, Thailand, and Turkey in panel b, meaning countries. These results suggest that value added brought that these countries bring forward value added from the forward in the transport and trade sector is contributed transport and trade sector. largely by lower-income countries. Module 2 A ratio can be calculated that shows whether a product The lack of a clear pattern in other sectors may reflect is used predominantly as an intermediate input or carries the fact that developing countries export value added in value added from other sectors. This ratio is calculated by services as easily as developed countries (an example is the taking the direct value added share plus each of the linkages well-developed computer sector in India). Another rea- and dividing it by the direct value added share without link- son could be the high regulatory burdens for both busi- ages. Figure 2.17 replicates the exercise in Figure 2.15 but ness and construction services, which inhibits tradability. Figure 2.16. Direct Value Added plus Linkages/Gross Share for Transport and Trade Services in Selected Economies, 1992–2007 a. Forward linkages b. Backward linkages 3.5 1.5 Value added + backward linkages/gross share of services Value added + forward linkages/gross share of services 3.0 1.4 2.5 1.3 2.0 1.2 1.5 1.1 1.0 1.0 1990 1995 2000 2005 2010 1990 1995 2000 2005 2010 Thailand Turkey United Kingdom Argentina India Source: Export of Value Added Database. 64 Valuing Services in Trade Figure 2.17. Relationship between Value-Added Linkages and Development in Selected Economies, 2007 a. Forward linkages b. Forward linkages c. Forward linkages in business services in construction in transport and trade linkages/direct value added share linkages/direct value added share linkages/direct value added share 15 8 Value added share + forward Value added share + forward Value added share + forward 8 6 6 10 4 4 JPN 5 DEU 2 2 IND USA IRL USA USA IND DEU IRL HKG HKG JPN IND DEU HKG JPN 0 0 0 7 8 9 10 11 7 8 9 10 11 7 8 9 10 11 Log of per capita GDP (PPP) Log of per capita GDP (PPP) Log of per capita GDP (PPP) Module 2 d. Backward linkages e. Backward linkages f. Backward linkages in business services in construction in transport and trade linkages/direct value added share linkages/direct value added share linkages/direct value added share 1.5 3.0 2.5 Value added share + backward Value added share + backward Value added share + backward 2.5 2.0 2.0 1.0 1.5 IRL 1.5 USA DEU IRL HKG IND HKG IND HKG 1.0 IRL JPN 1.0 DEU USA JPN IND DEU USA JPN 0.5 0.5 0.5 7 8 9 10 11 7 8 9 10 11 7 8 9 10 11 Log of per capita GDP (PPP) Log of per capita GDP (PPP) Log of per capita GDP (PPP) Fitted values High-income country Upper-middle-income country Lower-middle income country Low-income country Source: Export of Value Added Database. Note: DEU = Germany; HKG = Hong Kong SAR, China; IND = India; IRL = Ireland; JPN = Japan; PPP = purchasing power parity; USA = United States. Domestic or Imported Services Content of Exported The diagnostic requires a deep understanding of the Goods and Services country context, based on discussions with government authorities, experts, private sector representatives, and Several indicators are critical to assessing the domes- other stakeholders. Table 2.4 indicates some of the ques- tic services content of exported goods and ser- tions that need to be addressed. vices. They include forward and backward linkages; Services are used as inputs in the production of many total exports through Modes 1 and 2; the Revealed Comparative Advantage (RCA) Index; and firm-level cus- goods. They are an essential part of the production pro- toms, fi rm-level census data, and balance sheet data. cess because they increase the productivity of the goods Data on these indicators should be collected for spe- firms (Arnold and others 2012; Arnold, Javorcik, and cific years for all countries. Useful sources of information include the following: Mattoo 2011). These services can be produced and pro- • OECD-WTO’s TiVA or GTAP databases vided by domestic firms or imported. The presence of for- • World Bank’s Trade in Services Database (see appendix B) eign firms providing services inputs tends to increase the • World Bank’s Export of Value Added Database (see productivity of domestic goods firms to an even greater appendix A) degree than the presence of domestic firms (Arnold and Module 2: Services as a Source of Competitiveness in the Whole Economy 65 Table 2.4. Domestic Services Content of Exported Goods and Services: Issues for Discussion Interviewee Issue Senior policy makers at • In which goods sectors is the country specialized? the ministry of trade or • In which goods sectors does the country have a comparative advantage? foreign affairs • What services do these goods sectors require? • What sectors in which the country has a comparative advantage carry forward and backward value added? • Which services sectors are necessary to capitalize on that value added? • Is the country well connected to global supply chains (that is, does it trade parts and components of goods as opposed to services)? • Are services domestically produced and exported, or does the country import many services as well? Officials at export • Have firms encountered any difficulties in obtaining services for their production and exports? If so, what promotion agencies goods sectors do these firms provide? • Are firms forced to specialize in lower-value-added products because they encounter problems obtaining services inputs? Major goods exporters • Which services are required for optimal production of the goods you produce? • Which services are essential for exporting the goods you produce? • Are these services easily obtainable? Do you purchase these services from domestic or foreign suppliers? Module 2 • Are there substantial barriers to importing services that are essential for production and export? • Is your firm inhibited from climbing the value chain by highly regulated services markets? • Would easier access to imported inputs allow your firm to specialize in products with higher value added? Table 2.5. Impact of Broad Categories of Foreign Services Inputs on Manufacturing Exports Services inputs Nature of manufacturing sector Business Communication Finance Insurance FDI inflow Technology intensive Positive*** Labor intensive Negative** Resource intensive Source: Francois and Woerz 2008. Note: ** p < 0.05; *** p < 0.01. Blank cells indicate that the relationship is not statistically significant. FDI = foreign direct investment. others 2012). It is therefore essential that firms have good individual manufacturing industries, because they are qual- access to services markets. Such access is especially impor- itatively different; looking at the aggregate picture could tant when firms are part of global supply chains, because mask essential features of the effects of foreign services services increase specialization and concentration of the input. At the same time, an aggregate manufacturing clas- manufacturing firms providing goods further down- sification can be thought of based on factor inputs. Francois stream in the chain. and Woerz (2008) estimate the effects of a larger import Different industries require different types of services share on exports of technology-, labor-, and resource-inten- inputs. Sophisticated information technology (IT) services, sive industries. for example, may be more important in automobiles than Table 2.5 summarizes their aggregate findings. It shows in food processing. Industries that are closely tied to global that imported business services have a clear and strong supply chains are more likely to be dependent on a well- positive effect on the exports of manufacturing sectors functioning domestic transport and logistics services sector that are technology intensive. In contrast, a higher level than firms that are not. The downstream effects of services of imported business services is negatively associated with supply thus depend on the “depth” of the intermediate link- labor-intensive industries. Other services do not seem to ages of each goods sector. have a statistically significant association with a particular This section focuses on the knock-on effect of higher lev- type of industry. els of services imports on the tradability of the downstream This division of sectors may be too broad. Table 2.6 manufacturing or services sector. Francois and Woerz therefore displays a finer degree of disaggregation based (2008) note that it is important to distinguish between on the factor intensities of various manufacturing sectors. 66 Valuing Services in Trade Table 2.6. Impact of Disaggregated Foreign Services Inputs on Manufacturing Exports Services inputs Sector Business Communication Finance Insurance FDI inflow Technology intensive Chemicals Positive*** Positive** Electrical equipment Positive** Positive** Negative** Machinery Positive** Negative** Motor vehicles Positive** Negative** Labor intensive Textiles Negative*** Positive*** Negative** Clothing Negative*** Positive*** Leather Negative*** Positive*** Negative** Food Other transport equipment Positive*** Resource intensive Coke Positive*** Negative** Negative** Module 2 Minerals Metals Paper Wood Source: Francois and Woerz 2008. Note: *** p < 0.01, ** p < 0.05. Blank cells indicate that the relationship is not statistically significant. FDI = foreign direct investment. It shows that most linkages between services imports and foreign services suppliers, policy makers need to assess manufacturing exports take place in sectors that are inten- which industries a country specializes in and where the sive in technology or labor, with varying effects. Imported value added of particular exports lies. business services are positively correlated with most sectors that are technology (and skill) intensive (exceptions are Services Indirect Linkages motor vehicles and food). Insurance services have a nega- This section examines more closely indirect linkages tive relationship with technology-intensive manufacturing. through exports. Such linkages provide valuable informa- Financial services appear to have a positive effect on expan- tion on how much value added a country, income group, sion of some technology-intensive manufacturing exports, or geographical region carries forward as inputs in exports. including electrical equipment and motor vehicles. By analyzing value added at the individual-country Business services are negatively associated with expan- level, one can investigate the validity for specific cases of sion of trade in light manufacturing, a labor-intensive the relationship identified by Francois and Woerz (2008) sector. Communication services appear to be a lever for and summarized in figures 2.18 and 2.19. This information expansion of light manufacturing: whereas exports in can shed light on two important questions: the textiles, clothing, and leather manufacturing indus- tries suffer from having more imported business services • To what extent does services value added carried over in inputs, light manufacturing is boosted by imports of com- a sector increase the value of the sector’s exports? munication services. As light manufacturing is a sector in • To what extent is a services sector a component of which many developing countries specialize, exploring this exports in downstream goods sectors? relationship through field analysis and country-specific evidence is recommended. Figure 2.18 shows the extent to which business services This methodology of tradability provides several insights. are a component of value-added exports in other sectors in It suggests that openness to business services from abroad Malaysia, Romania, and the United States. is highly correlated with the competitiveness of skilled- The business services sector carries forward more value labor- and technology-intensive manufacturing industries. added in the United States than in Malaysia or Romania. However, before making any decisions on opening up to The importance of services value added embodied in other Module 2: Services as a Source of Competitiveness in the Whole Economy 67 Figure 2.18. Ratio of Exported Forward to Backward Linkages of Business Services in Other Sectors in Romania, Malaysia, and the United States, 2001 and 2007 a. 2001 b. 2007 aff aff wap b_t wap b_t trn cns trn cns tex crp tex crp t_t egw t_t egw ppp egy ppp egy pfd fmp pfd fmp Module 2 osg i_s osg i_s omn lea omn lea omf .25 lum omf lum .25 nmm mac nmm mac nfm nfm Romania Malaysia United States Source: Export of Value Added Database. Note: Center indicates no value added. The larger the circle, the greater the role of forward linkages relative to backward linkages. aff = agriculture, forestry and fisheries; b_t = beverages and tobacco products; cns = construction; crp = chemical, rubber, and plastic products; egw = electricity, gas, and water; egy = energy extraction; fmp = metal products; i_s = ferrous metals; lea = leather products; lum = wood products; mac = machinery and equipment not elsewhere classified; nfm = metals not elsewhere classified; nmm = mineral products; omf = manufactures not elsewhere classified; omn = minerals not elsewhere classified; osg = public services; pfd = processed foods; ppp = paper products and printing; t_t = transport and trade; tex = textiles; trn = transport equipment; wap = wearing apparel. sectors increased in both Malaysia and Romania (even in way to expand exports in technology- and skill-intensive basic sectors such as agriculture, forestry, and fisheries) goods sectors. The analytical framework outlined earlier between 2001 and 2007, reducing the gap relative to the also shows a correlation between performance in these sec- United States. Both Malaysia and Romania also expanded tors and imported business services inputs, underlying the the use of imported services (data are available but not importance of facilitating the sourcing of the best services, reported in figure 2.18). An alternative way to identify a whether they are produced domestically or abroad. country’s position regarding the value added of services There are also notable differences between Mexico and is to analyze complementarity based on each sector’s for- the United States. Business services, for example, account ward linkages. Figure 2.19 shows how important business for a larger share of inputs in Mexico than in the United services are as an input for other sectors’ exports (that is, States. Policy makers have to determine the reasons for value added carried forward), expressed as a share of busi- such a pattern. Another difference between the two coun- ness services total exported value added by all other sectors. tries is the extent that business services are used in the In both Mexico and the United States, the machin- machinery and equipment sector, which is much higher in ery and equipment sector uses a relatively large share Mexico than in the United States. Examination of differ- of business services. Other manufacturing sectors that ences in RCA Indexes based on gross flows, forward link- account for large shares of business services as inputs in ages, and backward linkages can enrich the analysis. both countries are chemicals, rubber and plastic products, Box 2.4 provides an example of how firm-level data can and transport equipment. This pattern is consistent with be used to conduct the analysis. Data on Romania identify the notion that business services inputs are an important additional trends in specific sectors. 68 Valuing Services in Trade Figure 2.19. Sectors Exporting Business Services Value Added as Share of Total Exported Business Services Value Added based on Forward Linkages in Mexico and the United States, 2007 a. Mexico b. United States mac mac trn t_t t_t crp wap trn egy osg crp pfd tex aff lum ppp fmp omf nmm fmp b_t nmm aff nfm Module 2 omf tex pfd egy nfm lum i_s i_s ppp wap cns b_t osg cns omn lea lea omn egw egw osp osp dwe dwe 0 0.05 0.1 0.15 0.2 0.25 0 0.05 0.1 0.15 0.2 0.25 Source: Export of Value Added Database. Note: aff = agriculture, forestry and fisheries; b_t = beverages and tobacco products; cns = construction; crp = chemical, rubber and plastic products; dwe = dwellings; egw = electricity, gas, and water; egy = energy extraction; fmp = metal products; i_s = ferrous metals; lea = leather products; lum = wood products; mac = machinery and equipment not elsewhere classified; nfm = metals not elsewhere classified; nmm = mineral products; omf = manufactures not elsewhere classified; omn = minerals not elsewhere classified; osg = public services; osp = other private services; pfd = processed foods; ppp = paper products and printing; t_t = transport and trade; tex = textiles; trn = transport equipment; wap = wearing apparel. Box 2.4. Output, Value Added, and the Importance of the Wholesale Sector for Exports of Downstream Goods: Evidence from Romania National input-output tables—and the forward and backward linkages they reveal—provide a good overview of where value added is produced in a country. Firm-level data can help identify trends and developments. Such data can help determine whether the assertion that value added is mostly produced in services but exported through goods holds in specific countries and sectors. In Romania, the goods sector accounts for a larger share of output exported than value added carried abroad (figure B2.4.1). The opposite is true for most services activities. The share of value added is substantially larger than exported output in the com- puter and modern communications and computer repair sectors. However, some variation between output and value added is evident across services. In sectors such as construction, accommo- dation, transport, and traditional communication, exported output is greater than value added. This finding is largely in line with the results in Module 1 on the export sophistication of modern services. Modern services, which have higher value added, are more likely to be exported by developed countries than lower-income countries. Yet Romania exports computer and modern com- munications and professional services. Ideally, policy would enable Romania to take advantage of sectors that export larger shares of value added, helping these sectors expand. Firm-level data also highlight the importance of the wholesaling sector to other sectors. Firms can export either directly or through wholesalers. New exporters often use intermediaries, because it saves them the cost and effort associated with learning about foreign markets and export procedures or finding customers abroad. An increasing share of Romanian exports goes through wholesalers. In 2005, 7 percent of exports from Romania were sold by wholesalers. By 2011, the share had reached 12 percent. Intermediaries play the most prominent role in the export of food and bever- ages, chemicals, and wood. They do not appear to be important in sectors in which a lead firm (buyer or supplier) dominates the value chain production networks, such as apparel; motor vehicles; electrical machinery; and radio, TV, and communications equipment. (continued on next page) Module 2: Services as a Source of Competitiveness in the Whole Economy 69 Box 2.4. (continued) Figure B2.4.1. Exports and Value-Added Shares of Selected Services Subsectors in Romania Goods Computer and modern communications Other services Wholesale and retail trade Professionals Administrative and support services Transport and traditional communications Electricity, water, waste Computer repairs Construction Health and social work Food services Real estate Module 2 Accommodation Recreation Education 0 20 40 60 80 Share of output and value added (percent) Output Value added Table B2.4.1 indicates that export unit values (a proxy for quality) are higher for exporters that use wholesalers. This relationship is driven by the food and beverages, chemicals, plastics, and wood and furniture sectors. More experienced and sophisticated exporters prefer to deal directly with foreign customers. Although initiating such business relationships may be costly, direct interactions with foreign customers allow producers to tailor their products better to the needs of the market. Such relationships also increase margins and establish long-term relationships. Table B2.4.1. Unit Values of Direct Exports versus Exports through Intermediaries Regression coefficients of log unit values on Number of Sector dummy for exports by intermediary exporters observations All 0.0530** 362,822 (–2.16) Food products and beverages 0.199*** 69,672 (–5.08) Chemicals 0.244*** 66,255 (–4.81) Wood 0.173*** 73,622 (–6.91) Coke –0.202 61,631 (–1.35) Leather –0.0870** 82,995 (–2.23) Basic metals 0.261*** 64,541 (–4.41) Furniture 0.0963** 86,979 (–2.5) Radio, TV, and communications equipment –0.106 67,043 (–1.25) Apparel 0.0257 164,327 (–0.42) Machinery and equipment 0.0903** 83,941 (–2.54) (continued on next page) 70 Valuing Services in Trade Box 2.4. (continued) Table B2.4.1. (continued) Regression coefficients of log unit values on Number of Sector dummy for exports by intermediary exporters observations Other transport equipment –0.102 64,010 (–1.45) Rubber and plastic 0.175*** 74,007 (–4.87) Electrical machinery 0.0236 77,728 (–0.6) Motor vehicles –0.233** 77,696 (–2.34) Source: Romanian Structural Business Survey. Note: Only manufacturing and wholesale firms are included. Intermediary exporters are exporters whose primary activity is wholesale (code 51 in NACE rev. 1.1). Regressions are run separately for several two-digit NACE (rev. 1.1) industries. Observations are defined Module 2 on firm-product-destination-year level and include product-destination-year fixed effects. Standard errors are clustered on firm level. t-statistics are in parentheses. ** p < 0.05, *** p < 0.01. NACE = Nomenclature statistique des activités économiques dans la Communauté européenne (Statistical Classification of Economic Activities in the European Community). Note ———. 2012. “Exports of Services: Indian Experience in Perspective.” Working Paper 12/102, National Institute of Public Finance and 1. For example, both Baldwin and Lopez-Gonzalez (2013) and Policy, New Delhi. Feenstra and Jensen (2012) find that the proportionality assumption in Feenstra, Robert C., and J. Bradford Jensen. 2012. “Evaluating Estimates the use of intermediates and by different industries leads to nontrivial of Materials Offshoring from US Manufacturing.” Economics Letters errors in some cases. 117 (1): 170–73. Fisher, Alan G. B. 1939. “Production, Primary, Secondary and Tertiary.” References Economic Record 15 (1): 24–38. ———. 1952. “A Note on Tertiary Production.” Economic Journal 62 Amin, Mohammad, and Aaditya Mattoo. 2006. “Do Institutions Matter (248): 820–34. More for Services?” Policy Research Working Paper 4032, World Francois Joseph, Miriam Manchin, and Patrick Tomberger. 2013. Bank, Washington, DC. “Services Linkages and the Value Added Content of Trade.” Policy ———. 2008. “Human Capital and the Changing Structure of the Indian Research Working Paper 6432, World Bank, Washington, DC. Economy.” Policy Research Working Paper 4576, World Bank, Francois, Joseph, and Julia Woerz. 2008. “Producer Services, Washington, DC. Manufacturing Linkages, and Trade.” Journal of Industry, Competition Arnold, Jens Matthias, Beata Javorcik, Molly Lipscomb, and Aaditya and Trade 8 (3): 199–229. Mattoo. 2012. “Services Reform and Manufacturing Performance: Global Trade Analysis Project (GTAP) Database. Purdue University, Evidence from India.” Policy Research Working Paper Series 5948, Lafayette, IN. https://www.gtap.agecon.purdue.edu/databases. World Bank, Washington, DC. Hausmann, Ricardo, César A. Hidalgo, Sebastián Bustos, Michele Arnold, Jens M., Beata S. Javorcik, and Aaditya Mattoo. 2011. “Does Coscia, Sarah Chung, Juan Jimenez, Alexander Simoes, and Services Liberalization Benefit Manufacturing Firms? Evidence from Muhammed A. Yıldırım. 2014. The Atlas of Economic Complexity: the Czech Republic.” Journal of International Economics 85 (1):136 –46. Mapping Paths to Prosperity (Revised edition). Cambridge, MA: The Baldwin, Richard, and Javier Lopez-Gonzalez. 2013. “Supply-Chain MIT Press. Trade: A Portrait of Global Patterns and Several Testable Hypotheses.” Liping, Zahang, and Simon Evenett. 2010. “The Growth of China’s NBER Working Paper 18957, National Bureau of Economic Research, Services Sector and Associated Trade: Complementarities between Cambridge, MA. Structural Change and Sustainability.” International Institute for Clark, Colin, 1940. The Conditions of Economic Progress. London: Sustainable Development, Winnipeg, Canada. Macmillan and Company. Trade in Services Database. World Bank, Washington, DC. Eichengreen, Barry, and Poonam Gupta. 2011. “The Service Sector as World Development Indicators (database). World Bank, Washington, India’s Road to Economic Growth.” NBER Working Paper 16757, DC. http://data.worldbank.org/data-catalog/world-development- National Bureau of Economic Research, Cambridge MA. indicators. Module 3 Assessing the Potential for Trade in Services This module explores how countries can assess their other stakeholders. Table 3.1 indicates some of the ques- services trade potential and identify untapped opportuni- tions that need to be addressed. ties. It does so by: Analytical Approach • examining alternative modes of supply • identifying the geographical pattern of production and Services are tradable through various modes. Many services demand of services require face-to-face contact (simultaneous consumption • identifying services sectors in which a country has a and production) between producer and consumer. For comparative advantage example, many medical services can be provided only if the • using gravity models to determine trade potential. doctor and patient are in the same place. Other services, such as telecommunication services, are delivered over the These steps can help identify “low-hanging fruit”— Internet or by a foreign affiliate that invests in the domestic underexploited sectors that can easily become sources of market. exports. The General Agreement on Trade in Services (GATS) of the World Trade Organization (WTO) and other trade agreements define four modes of supply (figure 3.1): Assessing Tradability by Modes Several indicators are critical to assessing tradability by • Mode 1: Supply of a service from the territory of one modes. They include the volume of trade and the share of country to another country. Supplier and consumer trade through Modes 1, 2, and 3; the most dominant mode of trade for each sector; and the ratio of Modes 1 and 2 interact across a distance (cross-border trade or trade trade to Mode 3 trade. Data on these indicators should be over the Internet). collected for specific years for all countries. Useful sources of • Mode 2: Consumption of a service by consumers of one information include the following: country who visit or move temporarily to another country, • World Development Indicators • World Bank’s Trade in Services Database (see appendix B) which supplies the service (consumption abroad). • Organisation for Economic Co-operation and • Mode 3: Services provided by a foreign affiliate that has Development (OECD), United Nations Conference on established a commercial presence in another country Trade and Development (UNCTAD), Eurostat, and natio- nal sources (such as the Bureau of Economic Analysis for (commercial presence). statistics on the United States and Foreign Affiliates Trade • Mode 4: Services supplied by a foreign natural person, Statistics [FATS]) employed or self-employed, who at the moment the ser- • UNCTAD (foreign direct investment statistics). • World Trade Organization vice is provided stays temporarily in the territory of the consuming country (presence of natural persons). Table 3.2 provides examples of activities traded through each mode of supply, estimates the importance of each The diagnostic requires a deep understanding of the mode in total trade in services, and describes some of country context, based on discussions with government the data inadequacies that the study of services trade still authorities, experts, private sector representatives, and confronts. 71 72 Valuing Services in Trade Table 3.1. Assessing Tradability by Modes: Issues for Discussion Interviewee Issue Senior policy makers • Do exporters export mainly through Mode 1 and Mode 2 or through Mode 3? at the ministry of • Which mode accounts for a larger share of trade in this sector, Mode 1 or Mode 3? trade or foreign • In which mode is growth highest? affairs • Has there been a slowdown in a particular mode in recent years? • Has policy changed with reference to a mode in recent years? • Are companies allowed to hire foreign nationals? Major exporters • Is your company using both Mode 1 and Mode 3 for exports? • Are Mode 1 and Mode 3 complements or substitutes in your company’s business strategies? • Does the choice of mode depend on the characteristic of the service, or do other factors such as endowments, transportation costs, demand, and the regulatory environment play a role in determining the mode of export? • Which of the following are important in choosing an export mode: regulatory barriers or cost of supplying reliable and quality services? • Has trade in services changed from one mode to another over time? What evolution is likely? • Are payment systems in place that allow your company to export through your preferred mode of supply? Are there different problems depending on the destination? Chamber of • How large are services exporters? Does the size of the company affect the mode through which it exports? Commerce • Which sectors are most likely to export through Mode 1? Mode 3? • Are services exports through Mode 3 concentrated in one sector? Note: For a description and examples of the modes of trade in services, see figure 3.1 and table 3.2. Figure 3.1. The Four Modes of Trade in Services Border Country A Country B Mode 1 Supplier Consumer Service crosses the border Module 3 from A from B Mode 2 Supplier Consumer from A from B Consumer crosses the border Mode 3 Direct investment in Consumer Supplier from country Foreign from B country A affiliate Mode 4 Supplier Consumer Producer crosses the border from A from B Module 3: Assessing the Potential for Trade in Services 73 Table 3.2. Examples of Services Trade by Mode Estimated share of total services trade Mode of supply Examples Challenges (percent) Mode 1: • International phone calls Balance of payments does not distinguish 25–30 Cross-border trade • Telemedicine between cross-border supply and presence of natural persons • Online courses • International transportation services Mode 2: • Tourism Tourism includes goods and is not subdivided into 10–15 Consumption • Education categories of services consumed by tourists. Some abroad transactions are in other balance of payments • Health tourism categories. Mode 3: • Financial services Foreign Affiliates Trade Statistics (FATS) 55–60 Commercial • Distribution services are available for only a limited number of presence countries (largely OECD countries). Foreign • Construction services direct investment (FDI) statistics cover a larger set of countries, not only majority-controlled companies. Statistics do not distinguish between Modes 3 and 4. Mode 4: Presence • Professional services Balance of payments does not distinguish Less than 5 of natural persons • Entertainment services between cross-border supply and presence of natural persons or between modes. Sources: Maurer and others 2008; Magdeleine and Maurer 2008. Note: OECD = Organisation for Economic Co-operation and Development. Services in a sector can be provided through several Table 3.3. Dominant Modes of Services Exports by Sector modes of supply. In the tourism sector, for example, Mode Mode 2 is the main mode of supply, but exports of travel Type of service 1 2 3 4 agency services fall under Mode 1, investment in hotels and Business x x x restaurants under Mode 3, and the movement of managers Communications x x Computer and information x x and chefs under Mode 4. Table 3.3 identifies the dominant Construction x x x Module 3 modes of services exports for various sectors. Distribution x Trade through the four modes is recorded in vari- Education x x x ous sources. International organizations such as the Financial x x Organisation for Economic Co-operation and Development Health x x x (OECD), the United Nations Conference on Trade and Insurance x x Personal, cultural, and recreational x x Development (UNCTAD), and Eurostat, report interna- Transportation x x tionally comparable trade data across countries and sec- Travel x tors through Modes 1 and 2 using the Extended Balance of Payments Services (EBOPS) classification system. Sources for trade flows also include the Trade in Services Database, These data can also be constructed using firm-level bal- World Development Indicators, and the International ance sheet and census data. Doing so is cumbersome and Monetary Fund’s Balance of Payment Statistics. time consuming, however. Firm-level information includes Comparable and comprehensive cross-country data sales; output; employment; value added; exports and imports for Mode 3 can be accessed through the Foreign Affiliates of goods and services; the number of enterprises; and the Trade Statistics (FATS), available from the OECD, source of ownership, imports, and exports. Many countries UNCTAD, and Eurostat. FATS provides a range of indica- collect foreign direct investment (FDI) flows, but few pro- tors on the activities of foreign affiliates, including exports vide geographical and activity breakdowns. Moreover, these and imports, sales, expenditures, profits, value added, data stand as a very rough proxy for sales by foreign affiliates. inter- and intrafirm trade, and employment, with a focus Data for Mode 4 are even scarcer, and there is no cen- on services. Inward FATS cover the operations of foreign- tralized source of data. Although some Mode 4 trade is owned firms in the host economy. Outward FATS report recorded in the balance of payments, it is difficult to distin- statistics on foreign affiliates of domestic companies. guish Mode 4 trade from other modes of supply (box 3.1). 74 Valuing Services in Trade Box 3.1. Remittances are not a measure of Mode 4 For many countries, remittances exceed official aid flows or foreign direct investment (FDI). The two standard items that relate to remittances, as defined in Balance of Payments and International Investment Position Manual, Sixth Edition (BPM6), are compensation of employees and personal transfers. Compensation of employees represents “remuneration in return for the labour input to the production process contributed by an individual in an employer-employee relationship with the enterprise” (BPM6, para. 11.10). It refers to the income of border, seasonal, and other short-term workers who are in an employer-employee relationship in a country where they are not resident and residents in an employer-employee relationship with a nonresident entity. Personal transfers consist of “all current transfers in cash or in kind, made or received, by resident households to or from non- resident households” (BPM6, para. 12.21). Such transfers include all current transfers from resident to nonresident households, independent of (a) the sources of income of the sender (whether he or she receives income from labor, entrepreneurial or property income, social benefits, or any other types of transfers or disposes of assets); (b) the relationship between the households (whether they include related or unrelated members); and (c) the purpose for which the transfer is made (inheritance, alimony, lottery, and so forth). BPM6 recommends recording a supplementary item entitled “workers’ remittances,” which covers current transfers made by employees to residents of another economy. Data on compensation of employees and personal transfers broken down by relevant categories (for example, intracorporate transferees or people directly employed by a foreign affiliate in services) could provide additional information on Mode 4. However, these flows will not reflect the value of the service contract (or sales of services) and cannot therefore be used to measure the inter- national supply of services. Based on the definition of compensation of employees and personal transfers, as well as other remit- tances indicators, a majority of contractual service suppliers do not belong to the set of people to whom these balance of payments items refer. In addition, in the (rare) cases where some Mode 4 persons are covered, double counting would occur, because the value of the services would be included implicitly, either within the value of trade in services statistics (between residents and nonresidents) for Mode 4 or in Foreign Affliliates Trade Statistics (FATS) for Mode 3. In some instances, however, compilers might use this information to extract a Mode 4 estimate for some contractual service suppliers from appropriate balance of payments services items. Although this additional information could not be tracked to specific transactions, it could be used as a check. However, it is often difficult to identify specific compensation of employees/personal transfers for categories of interest for Mode 4 (that is, people who become residents of the host economy), because related transactions often represent a small proportion of the relevant income and transfer transactions. Source: Manual on Statistics of International Trade in Services 2010, Department of Economic and Social Affairs of the United Nations Secretariat, Statistics Division, © (2011) United Nations. Reprinted with permission of the United Nations. Implementation Sales of foreign affiliates include only companies that are majority owned by U.S. companies. The importance of trade in a particular service through With the exceptions of travel and transport services, Module 3 a particular mode can be checked with country-specific tradable U.S. services exports are concentrated in Mode data. It is particularly important to assess the relevance 3. Mode 2 dominates travel, because it includes consump- of Mode 3, which represents a sizable and growing share tion in the United States by foreign tourists. Transport ser- of services trade. Data for Modes 1 and 2 can be verified vices are traded though both Modes 1 and 3, but Mode 1 through the Trade in Services Database or other readily dominates. available cross-country sources. Trade data for Mode 3 are Exports through Mode 1 are nevertheless important for more difficult to locate. a range of services, including construction; business ser- The preferred alternative is to complement cross- vices; personal, cultural, and recreational services; and, to country resources with country-specific information. a lesser extent, communications. This mode accounts for Sources for these data can be accessed with permission from almost 30 percent of exports for business services; per- national authorities and statistical offices. Identification of sonal, cultural, and recreational services; and construction access points to obtain these data should be a key compo- (and about 15 percent of communications services). These nent of a services trade diagnostics exercise. services sectors indicate that although most U.S. services Table 3.4 illustrates the analysis of tradability of services exports are Mode 3 transactions, Mode 1 still accounts in the United States. Data on Mode 3 come from the Bureau for a significant share of exports in these sectors. It forms of Economic Analysis (BEA) of the U.S. Department of a sizable part and therefore remains important for U.S. Commerce (under the heading “Data on Direct Investment competitiveness. and Multinational Company Activities”). Data on Modes 1 Other services sectors in the United States, such as and 2 come from the Trade in Services Database. The table finance, insurance, and to a lesser extent computer shows the volume of exports through Modes 1 and 2 and and information services, have very high absolute compares it to the volume of exports through Mode 3. trade figures for Mode 1. However, the likely mode of tradability nonetheless appears to be Mode 3 where Module 3: Assessing the Potential for Trade in Services 75 Table 3.4. Tradability of U.S. Services through Modes 1, 2, and 3, 2009 Exports (millions of dollars) Ratio Mode Service Mode 1/2 Mode 3 1/2:3 Travel 121,153 8,315 14.57 Transportation 61,841 38,761 1.60 Business services 91,650 232,679 0.39 Personal, cultural, and recreation (including 13,710 35,264 0.39 motion picture and sound-recording industries) Construction 6,792 17,970 0.38 Communications 9,549 51,884 0.18 Financial 55,456 335,099 0.17 Computer and information 13,380 119,665 0.11 Insurance 14,654 145,198 0.10 Education 76 2,940 0.03 Health 25 2,493 0.01 Distribution 2,854 1,375,403 0.00 Sources: Trade in Services Database and U.S. Bureau of Economic Analysis. Figure 3.2. Number of Exporting Firms in Romania, by Sector, 2009 4,000 Number of exporting firms 3,000 2,000 1,000 Module 3 0 Goods Traditional Services Modern Services Source: Romanian Structural Business Survey. exports are even higher, so that the ratio remains low. Assessing Tradability Based on Production Distribution, health, and education services are also traded largely through Mode 3. Although exports Several indicators are critical to assessing tradability through Mode 1 exist in these sectors, trade potential through production. They include employment and GDP data by region, the share of inputs used in final production within this mode is weak. by downstream industries, the share of employment in a Another option is to use firm-level data. Figure 3.2 particular industry in a region, the Tradability Index devel- shows that in Romania in 2009, the number of firms oped by Jensen and Kletzer (2006), and the Gini coefficient. Data on these indicators should be collected for specific exporting goods was more than 4 times the number years for all countries. Useful sources of information include of firms exporting traditional services and more than the following: 16 times the number of firms exporting modern ser- • World Development Indicators (services flows for Modes vices. Traditional services such as trade and transport 1 and 2) • World Bank’s Trade in Services Database (see appendix B) were the dominant subsectors within the services sector. • National sources measuring Mode 3 trade (FATS) Although the number of firms exporting modern ser- • National sources (input-output matrixes, employment vices is small, their importance should not be underes- statistics). timated, as expansion of this sector offers an alternative growth path, as explained in Ghani (2010) and Mishra, The diagnostic requires a deep understanding of the Lundstrom, and Anand (2011). country context, based on discussions with government authorities, experts, private sector representatives, and 76 Valuing Services in Trade Table 3.5. Assessing Tradability Based on Production: Issues for Discussion Interviewee Issue Senior policy makers at • Is production of some services sectors centralized within a region or other geographical unit? the ministry of trade • Which services sectors have a tendency to agglomerate? and other relevant • Do the centralized and agglomerated sectors correlate with the Tradability Index? ministries (infrastructure, transport, health) • Are there strong employment effects in sectors that are considered tradable? • How much growth in employment have tradable sectors experienced in recent years? Major exporters • Are all relevant ministries aware that the concentration of production is a good indicator of tradability? • Is production in your sector largely local, or does it take place in a particular place within the country? • How does your industry compare with others in terms of centralization or agglomeration? • Are economies of scale the main reason for concentration, or are other factors, such as regulation or concentrated demand, at work? • Does the reason for concentrating production in a particular location say anything about the tradability of the services produced in your sector? other stakeholders. Table 3.5 indicates some of the ques- production. Jensen and Kletzer (2006) use employment tions that need to be addressed. data for both measures, so that Analytical Approach Gi = 1 − ∑ (σ Y r i , r −1 + σ Yi , r ) ∗ (σ X i , r −1 − σ X i , r ) Like goods, some services are produced in particular locations. Production may be concentrated for a variety of reasons, including increasing returns to scale, prox- where r stands for region (or any other geographical unit), imity to inputs, the high cost of trade, and regulation. which must be sorted by the region’s share of industry Goods and services that are tradable usually show a employment; i for industry; s Yi, r for the region’s cumula- higher level of geographical concentration of production tive share of industry i’s employment; s Yi, r−1 the cumula- than goods and services that are not. Firm-level data can tive share of each industry’s employment in the region with reveal tradability and the characteristics of the traded the next-lowest share of industry employment; s Xi, r the Module 3 activities. region’s cumulative share of total employment; and s Xi, r−1 The methodology of Jensen and Kletzer (2006) com- the cumulative share of total employment in the region bines insights from the literature on agglomeration and with the next-lowest share of industry employment. In this revealed comparative advantage, which are closely linked. equation, similar employment data are used for both the Implicit in both concepts is the notion that the place of Y’s and X’s, but they need to be adjusted for the fact that production of a service is not related to the place of con- services are often used as inputs for further production. If sumption. When production is concentrated in an area, a services industry is geographically concentrated and has some of it is consumed there; what is not consumed is high demand for the nontraded services that are used as sent to another area of the country, exported for con- inputs, this nontradable service will also be geographically sumption, or incorporated as an input in a final good or concentrated. The sector would be incorrectly identified as service produced abroad. Therefore, if production does tradable. not match demand in a geographical area, the sector is To correct such miscalculation, one needs to deter- probably tradable. Lower transportation costs or greater mine how much of industry i’s output is used in any other production economies induce this concentration of industry, j, for each region. This region-specific measure of production. demand (IDSir) is calculated as follows: Jensen and Kletzer (2006) use a Gini coefficient (G) to ∑ ⎛⎜⎝ Y Yi , j EMPjr ⎞ measure the difference between production and demand IDSir = ∗ . within a country. Their index is different from other pure j i EMPj ⎟⎠ agglomeration indexes, because it is unaffected by the number of firms that produce in the same location. To This measure corrects for the value of services used as measure G, one needs empirical indicators of demand and inputs. The first term is the share of industry i’s output Module 3: Assessing the Potential for Trade in Services 77 used in industry j, multiplied by industry j’s employment Table 3.6. Gini Classes share in region r. IDSir adjusts for the concentration of Gini class Gini coefficient the downstream industry and corrects the X’s. Data on 1 <0.1 how much a particular service or good is used as an 2 0.1–0.3 input in another industry are available from national 3 >0.3 input-output tables. Each country should have a detailed Source: Jensen and Kletzer 2006. record of its output structure to trace back how much industries produce and use inputs from one another. Cross-country input-output tables are also available Kletzer (2006) distinguish three classes of coefficients (from the World Bank or OECD), albeit at a more aggre- (table 3.6). The classes should not always be interpreted gated level. on a strictly ordinal basis: a sector in a higher Gini class The last step in constructing the true Gini coefficient for may not have greater export potential than a sector in a production concentration for each service is to plug in the lower class. Higher Gini coefficients may indicate that region-specific demand measures for the X’s in the previ- forces other than production economies or transporta- ous equation to obtain: tion costs are at stake. Sometimes the nature of a service is such that it cannot be produced everywhere (one example Gi = 1 − ∑ (σ Y r i , r −1 + σ Yi , r ) ∗ (σ IDSi , r −1 − σ IDSi , r ) . is mining). This module follows Jensen and Kletzer (2006) in using 0.1 as the threshold for tradability. Sectors that fall into Gini classes 2 or 3 are thus considered tradable. The Gini coefficient measures the absolute difference Annex table 3A.1 shows the Gini class for various ser- between production and demand in a region for specific vices sectors, as classified by the North American Industry industries in a country. The greater the difference between Classification System (NAICS). Only about half of all ser- these two variables, and hence the higher the index, the vices subsectors are assessed as tradable. Within financial more production of the sector is concentrated in a particu- and insurance services, for example, activities performed lar region. Such concentration is a good indicator of inter- by banks and savings institutions are not tradable, whereas national tradability. activities related to securities and commodities or finan- Data on employment are available from national data- cial investment are highly tradable. About 46 subsectors Module 3 bases. For instance, employment data for the United (half of all subsectors) appear to be tradable (that is, in States are available from the Place of Work Consolidated Gini class 2 or 3). This figure varies widely across sectors Metropolitan Area (POWCMA) field on the Decennial (figure 3.3). Census Public-Use Microdata Samples (PUMS). The next step is to connect the trade figures with the Tradability Indexes. This can be done by adding trade data to subsectors that are tradable. Implementation Table 3.7 assesses the distribution of exports across Jensen and Kletzer (2006) calculate Gini coefficients for the sectors. It is based on benchmarking Modes 1 and 2 data United States. These data serve as rough proxies for other in Malaysia against the United States from the Trade in developed countries and give a sense of which services are Services Database. It shows that Malaysia’s export basket most likely to be tradable if an economy reaches a high is heavily skewed toward tourism, exports of which level of development. represent about 95 percent of total services exports. Analysis of the tradability of services for a country This share is very high, particularly given that only one consists of two steps. The first is to identify the tradabil- subsector of the sector (travel accommodation) is trad- ity of sectors based on the estimates for the United States able. Malaysia also shows export potential in profes- by Jensen and Kletzer (2006). The second is to connect sional, scientific, and technical services and management the Tradability Indexes with trade data from the country services. under analysis. Field interviews and other qualitative infor- The export patterns of services in the United States and mation are also necessary. Malaysia differ greatly. In the United States, tourism repre- Gini coefficients range from 0 to 1, with higher values sents only 34 percent of total services trade; in Malaysia it indicating a greater likelihood of tradability. Jensen and accounts for 95 percent. Other important services exports 78 Valuing Services in Trade Figure 3.3. Tradable versus Nontradable Services Subsectors 15 Number of subsectors 10 5 0 51 52 53 54 55 56 61 62 71 72 81 92 Subsector code Tradable Nontradable Source: Jensen and Kletzer 2006. Note: Subsectors are defined by the North American Industry Classification System (NAICS): 51 = information; 52 = finance and insurance; 53 = real estate and rental; 54 = professional, scientific, and technical services; 55 = management; 56 = administrative support; 61 = education; 62 = health care and social services; 71 = arts, entertainment, and recreational services; 72 = accommodation; 81 = other services; 92 = public administration. Module 3 Table 3.7. Share of Trade and Tradability Index by Services Sector in Malaysia and the United States, 2009 Number of tradable/number Percent of total services trade Sector of nontradable subsectors Malaysia United States Tourism 1/3 94.64 33.86 Professional, scientific, and technical 8/2 2.12 8.95 Management 1/0 1.97 2.03 Finance and insurance 3/2 0.52 15.78 Information 8/3 0.45 7.84 Administrative support 4/4 0.20 25.05 Public administration 8/7 0.06 5.82 Real estate and rental 3/2 0.02 0.57 Arts, entertainment, and recreation 2/2 0.02 0.06 Education 1/3 0.00 0.02 Health care and social services 2/12 0.00 0.01 Other services 5/8 0.00 0.00 Source: Trade in Services Database and Jensen and Kletzer 2006. by the United States are administrative support (25 per- public administration sectors are tradable. These figures cent); finance and insurance (16 percent); professional, suggest that the United States could exploit its trade poten- scientific, and technical (9 percent); and information (8 tial in these sectors even more than it has. Box 3.2 expands percent) services. Geographical concentration ratios show the application of the Jensen and Kletzer (2006) methodol- that many sectors in the information, professional, and ogy using firm-level data for Romania. Module 3: Assessing the Potential for Trade in Services 79 Box 3.2. Quantifying Tradability Potential: Evidence from Romania Panel a of figure B3.2.1 shows the number of firms in the goods sector and in 15 services sectors in Romania. The number of firms in the largest services sector (wholesale and retail trade) is slightly higher than the number of firms in the goods sector. This pattern is observed in most countries. Panel b shows that almost 30 percent of Romanian goods producers export. This figure is much lower for services: even in the sector with the largest percentage of exporters (computer and modern communications), just 10 percent of firms export. Romania’s retail sector has the second-highest share of exporters. This finding is interesting, as the tradability list of Jensen and Kletzer (2006) excludes retail sales (see annex table 3A.1). Exporting is a measure of firm performance. Comparing the rankings of the number of firms and the percentage of exporting firms helps identify the scope for efficiency gains in different sectors. For example, wholesale and retail trade and construction services appear to be populated largely by nonexporters. Policy interventions might be able to increase the export potential of these sectors. Firm-level data also allow the computation of Jensen and Kletzer–like indexes at the country level. U.S. industry data reveal that about 37 percent of industries are classified as Gini class 1, 37 percent as Gini class 2, and 27 percent as Gini class 3. The final classification for all industries by two-digit NAICS is shown below. Table B3.2.1 lists all services sectors for Romainia in which the Tradability Index was high (Gini class = 3 or Gini coefficient > 0.2) or moderate (Gini class = 2, Gini coefficient = 0.1–0.2). Figure B3.2.1. Number of Firms and Percentage of Exporters in Romania, by Sector, 2009 a. Number of firms by sector b. Percentage of firms that export Wholesale and retail trade Goods Goods Computer and modern communications Construction Wholesale and retail trade Other services Transport and traditional communications Professional Professional Transport and traditional communications Electricity, water, waste Computer and modern communications Other services Food service Administrative and support Electricity, water, waste Construction Real estate Real estate Recreation Computer repair Accommodation Food service Administrative and support Accommodation Health and social work Recreation Computer repair Health and social work Education Education 0 5,000 10,000 15,000 0 10 20 30 Module 3 Number of firms Percent of exporters Table B3.2.1. Services Tradability Index for Romania, 2008 Sector Gini class Gini coefficient Wholesale trade, except of motor vehicles and motorcycles 3 0.62 Land transport and transport via pipelines 3 0.36 Air transport 3 0.34 Computer programming, consulting, and related activities 3 0.31 Waste collection, treatment, and disposal activities, materials recovery 2 0.28 Activities of head offices, management consulting 2 0.26 Electricity, gas, steam, and air conditioning supply 2 0.25 Architectural and engineering activities, technical testing and analysis 2 0.22 Water transport 2 0.21 Telecommunications 2 0.21 Repair and installation of machinery and equipment 2 0.20 Real estate 2 0.18 Wholesale and retail trade and repair of motor vehicles and motorcycles 2 0.17 Publishing 2 0.17 Other professional, scientific, and technical activities 2 0.17 Office administrative, office support, and business support activities 2 0.16 Building construction 2 0.15 Warehousing and support activities for transportation 2 0.13 Rental and leasing 2 0.10 All other services sector 1 > 0.1 Note: Two-digit sectors are defined by the North American Industry Classification System (NAICS). For Gini classes, see table 3.6. (continued on next page) 80 Valuing Services in Trade Box 3.2. (continued) The availability of a wide range of firm-level indicators helps analysts identify the relationship between the tradability of a sector and the underlying characteristics of the firm. Table B3.2.2 does so for goods and services, testing the explanatory power of trad- ability for firm wages, turnover, investment in information technology (IT), and investment in research and development (R&D). It shows that firms whose goods are more tradable pay higher wages, have higher turnover, and invest more in IT. In contrast, for services only turnover seems correlated with exporting. Table B3.2.2. Correlation between Sectoral Tradability and Firm Characteristics in Romania, 2009 Goods Services (1) (2) (3) (4) (1) (2) (3) (4) Item ln(wage) ln(turnover) ln(IT) ln(R&D) ln(wage) ln(turnover) ln(IT) ln(R&D) Gini coefficient 5.26*** 6.93** 7.37** 4.38 1.87 3.95** 2.71 1.68 (1.35) (2.97) (3.39) (4.370) (1.67) (1.48) (2.33) (2.06) Observations 23 23 23 23 45 45 41 45 R-squared 0.364 0.255 0.168 0.057 0.038 0.156 0.037 0.012 Root mean square deviation 0.785 1.336 1.850 2.007 1.279 1.238 1.899 2.027 Note: Robust standard errors are in parentheses. ***p < 0.01, **p < 0.05. Box 3.3. The Updated Tradability Index of Gervais and Jensen Gervais and Jensen (2013) define an industry’s geographic concentration index as follows: M G= ∑(s − x ) i =1 i i 2 where si represents the share of industry employment in region i, and xi represents the share of total employment in region i. A high index indicates that production of the good or service in a region greatly exceeds consumption. This mismatch indicates the Module 3 tradability of a good between regions. The authors show that manufacturing has the highest G-score, followed by business services, wholesale and retail services, and personal services. However, there is considerable variation within each sector. The Jensen and Kletzer (2006) methodology provides Assessing Tradability Based on Comparative an intuitive and straightforward way to determine the Advantage tradability of particular services sectors. It has limita- tions, however. The method was developed in order to Several indicators are critical to assessing tradability based circumvent the often poor quality of services trade data; it on comparative advantage. They include standard trade cost proxies (as used in gravity equations, described therefore does not use any trade indicators. For instance, below); factor intensities; factor endowments; Costinot’s according to the Tradability Index, accounting and book- Complexity Index (described below); and measures of keeping services are considered nontradable (Gini class 1). the strength of legal institutions. Data on these indica- tors should be collected for specific years for all countries. However, these services have proven to be highly tradable Useful sources of information include the following: and an important source of services exports for many • World Bank’s Trade in Services Database (see appendix B) countries: in 2009, Malaysia exported about $3.5 million • Panel Study of Income Dynamics (PSID) survey data and the United States about $567 million of these services. (services sectors) • EU KLEMS (factor intensities in services) Gervais and Jensen (2013) have updated the methodol- • The Conference Board ogy of Jensen and Kletzer (2006) (box 3.3). Their meth- • World Development Indicators (number of Internet users odology uses the geographical distribution of production per 100 people) • Barro and Lee (2012) data set (educational attainment) based on data from the U.S. Bureau of Economic Analysis’ • Fraser Institute Labor Market Area. • World Bank governance indicators (rules of law and regu- latory quality). Module 3: Assessing the Potential for Trade in Services 81 Table 3.8. Assessing Tradability Based on Comparative Advantage: Issues for Discussion Interviewee Issue Senior policy makers at the ministries • What is the current state of physical infrastructure? How does it affect the export of services? of trade, tourism, telecommunications, • In which services sectors does your country specialize in terms of exports? transports, and relevant services • Do these services use much high-skilled labor? Do they use much ICT-related capital? (infrastructure, transport, health care) • How available are high-skilled labor and ICT-related capital? • How strong is contract enforcement? Major exporters • How do transportation costs affect the supply of services to foreign markets? • Is good physical infrastructure necessary to export your services? • What is the main endowment factor used to produce your services? • What kind of institutional climate would help your company export? • Is a strong rule of law an important factor for exporting your company’s services? • How much training is required at your company? • Is it difficult to find workers with the right qualifications? • What type of capital is critical to produce and export your company’s services? Is it easy to obtain? The diagnostic requires a deep understanding of the levels of services exports. These factors can affect a coun- country context, based on discussions with government try’s competitiveness in the world market for services, but authorities, experts, private sector representatives, and countrywide characteristics such as the availability of high- other stakeholders. Table 3.8 indicates some of the ques- skilled labor or the Internet and a strong rule of law can- tions that need to be addressed. not alone increase the productivity of a service or allow a company to export. At a sectoral level, what matters are not just relative endowments but their interaction with Analytical Approach sector-specific factor intensities. Chor (2010) proposes a This section introduces a methodology that uses com- relative endowments framework in which specialization by parative advantage to identify prospective export sectors. comparative advantage implies that countries tend to pro- Unlike the Tradability Index of Jensen and Kletzer (2006), duce and trade services that are relatively intensive in the this methodology is applicable to a wide set of countries endowments in which they are relatively abundant. using trade in services data that are commonly available. To export, different services sectors require different Module 3 The methodology consists of two steps. The first is the shares of skilled labor. The tourism sector, for example, identification of the drivers of comparative advantage in uses much less high-skilled labor than the financial ser- services. The second is the computation of an index of vices sector. Some sectors require more information and services tradability, following the methodology in van der communication technology (ICT) equipment than oth- Marel (2011) and van der Marel and Shepherd (2013). ers to facilitate production and export. Factor intensities The literature on the factors influencing tradability thus range widely across sectors. Many service sectors need relates largely to goods. It uses the foundational Ricardo more mid-skilled labor and capital than highly sophisti- model of comparative advantage (which attributes trade to cated factors of production. differences in technology) to assess variations in industry Table 3.9 shows the intensities of high-skilled labor, ICT- productivity across countries. capital, and service complexity for selected services sectors at One set of factors that influences a country’s ability to the most disaggregated level available (three-digit EBOPS). trade in services is its “endowments.” In early trade lit- It ranks sectors by the median value of these factor intensi- erature, it was typical to view endowments as primarily ties. Box 3.4 explains how factor intensities are computed. encompassing different types of labor and capital. Recent Computer and information services rank highest in the literature has developed a broader set of endowments, use of high-skilled labor, followed by business services and which includes such factors as institutional quality and a financial services. Travel and construction rank last. ICT- strong rule of law to enforce contracts. This type of insti- capital intensity is highest for “other business services,” tutional factor is an important source for trade in goods. finance, and transportation. The production and supply These factors probably play an important role for services. of these services appear to rely heavily on investment in Figure 3.4 shows that both the amount of high-skilled ICT equipment. Financial services appear to be the most labor supply and the strength of the domestic legal sys- complex and travel and transportation the least complex tem and property rights are strong predictors of higher sectors. 82 Valuing Services in Trade Figure 3.4. Services Trade, High-Skilled Labor Supply, and the Rule of Law in Selected Countries, 2009 a. High-skilled labor b. Rule of law 15 15 Log of trade in services Log of trade in services 10 10 5 5 0 0 −2 0 2 4 2 4 6 8 10 Log of high-skilled labor share Rule of law Fitted values High-income country Upper-middle-income country Low-income country Lower-middle-income country Sources: Trade in Services Database; Barro and Lee 2012; and World Bank Governance Indicators. Table 3.9. Rank of Factor Intensities for High-Skilled Labor, Information and Communications Technology (ICT) Capital, and Complexity Module 3 High-skilled labor ICT-capital Sector intensity intensity Complexity Computer and information services 1 8 5 Other business services 2 1 7 Financial services 3 2 1 Communications services 4 9 4 Insurance services 5 6 3 Personal, cultural, and recreation services 6 5 6 Transportation 7 3 8 Travel 8 4 9 Construction services 9 7 2 Sources: EU KLEMS and van der Marel and Shepherd 2013. Box 3.4. How Are Factor Intensities Computed? Factor intensity for high-skilled labor is calculated as the ratio of the hours worked by high-skilled labor to the total number of hours worked by all types of workers in a sector, multiplied by the total share of high-skilled labor in value added. Information and communication technology (ICT) capital intensity is measured as the ratio of ICT compensation to total capital compensation, multiplied by the total share of ICT capital in value added for each sector. The same procedure can be performed for non–ICT capital intensity. The production complexity of a service is computed from responses in the Panel Study on Income Dynamics (PSID) survey to the question, “How long does it take for each employee to be fully educated and qualified for the job in each service sector?” This meth- odology, borrowed from Costinot (2009), provides a measure of the magnitude of fixed training costs. It proxies the intensity of sectors in the use of good institutions. Complex sectors usually depend on good domestic institutions in order to produce efficiently. Module 3: Assessing the Potential for Trade in Services 83 Implementation data set. The supply of ICT-capital intensity is measured by the number of Internet users per 100 people. These These sectoral features are to a large extent dependent on data come from the World Development Indicators. A country characteristics. An empirical exercise to assess the proxy for institutional quality is a standard measure of tradability of a service should therefore include both coun- the strength of the domestic legal system and enforcement try and industry dimensions. Such an approach builds on of property rights. These data come from Gwartney and Eaton and Kortum (2002) and Chor (2010), which explains Lawson (2012). the observed pattern of specialization for industries by a Figure 3.5 displays the country supply measures productivity term that contains a systemic component. rescaled from 0 to 10, with increasing values indicating a Empirically, this pattern is driven by interaction effects higher endowment. There is significant variation across between country and industry variables. Sectors vary in countries for all three types of factor inputs. The United the factor and institutional conditions needed for produc- States is well endowed with all three factors. Indonesia has tion, and countries differ in their ability to provide these substantially lower levels of high-skilled labor and Internet requirements. Comparative advantage stems from the supply, but its institutional quality is fairly good. Haiti is interaction between the two components. Therefore, one better endowed in skills than ICT, and its legal systems lag needs to find the country supply factors that best match the far behind most countries. Romania scores well in terms of factor intensities described above. legal institutions and skill supply. The recent empirical literature on comparative The next step is to compute the index of tradability, advantage has developed variables that measure country which can be done by examining the relation between endowments in terms of high-skilled labor, ICT capital the interaction terms of comparative advantage and the supply per worker, and institutional quality. The supply of effect on exports. Van der Marel and Shepherd (2013) high-skilled labor is measured by the proportion of high- perform such an analysis within a gravity model (see skilled (tertiary-educated) employees in the population below) by applying standard trade cost proxies, such over 24. These data come from the Barro and Lee (2012) Figure 3.5. Index of Supply of High-Skilled Labor, Legal Institutions, and Internet in Selected Countries, 2009 Module 3 8 6 Index (0−10) 4 2 0 United States Malaysia Romania Turkey Indonesia Haiti Number of Internet users Strength of domestic legal system Share of high-skilled per 100 people and enforcement of property rights workers in the population Sources: World Development Inicators; Gwartney and Lawson 2012; and Barro and Lee 2012. 84 Valuing Services in Trade Box 3.5. What Does Firm Productivity Say about the Likelihood of Exporting? Evidence from Romania As the basis of the Tradability Index is the productivity “shock” observed in a pattern, it explains the extent to which firms in a typical services sector can capitalize on the endowments in an economy. The match between these endowments and factor intensity for a service explains productivity, as explained by the comparative advantage model (see Chor 2010 and van der Marel and Shepherd 2013). Firm-level data can help determine whether this productivity match is observable. Romania’s position on the index of factor supply is relatively high for legal institutions and high-skilled labor supply (as shown in figure 3.5), suggesting that Romania should specialize, or have a comparative advantage, in sectors that use these factors intensively. One should therefore observe above-average productivity for these sectors. Figure B3.5.1 indicates that productivity is above average for education, computer repair, and other services. Romania’s relatively high-skilled-labor endowments may explain the high productivity of education and computer repair services. The figure also suggests that there is unexploited export potential for construction and communications services, as both types of services are complex and Romania scores relatively high on institutions. Figure B3.5.1. Relative Productivity of Selected Export Sectors in Romania, 2009 Education Computer repairs Other services Wholesale and retail trade Electricity, water, waste Computer and modern communications Transport and traditional communications Food services Health and social work Recreation Goods Administrative and support services Professionals Accommodation Real estate Construction 0 1 2 3 4 Exporter productivity/ average productivity Module 3 Source: Romanian Structural Business Survey. as distance, common language, colony, and a regional Table 3.10. World Tradability Index for Selected Services trade agreement, together with the comparative advan- Sectors Rank Sector Index tage terms for each exporter. They use a probit estimator (of which the results are given in annex table 3A.2) to 1 Other business services 0.725 2 Travel 0.649 propose a Tradability Index for services based on trade 3 Personal, cultural, and recreational services 0.480 through Modes 1 and 2 by country and industry. Box 3.5 4 Communications services 0.442 provides an extension of this analysis by using firm-level 5 Computer and information services 0.421 data for Romania. 6 Insurance services 0.372 This index is displayed in table 3.10 for sectors at the 7 Government services 0.364 8 Financial services 0.348 most aggregated (one-digit) level. Higher values indicate 9 Construction services 0.337 greater trade potential. In principle, all services listed are 10 Transportation 0.187 tradable, as even personal, cultural, and recreational ser- Source: van der Marel and Shepherd 2013. vices have positive indexes. Business services have the Note: Index ranges from 0 to 1. highest index, followed by travel and communication ser- vices. Both insurance and financial services score relatively low, possibly because both sectors are still highly restricted data to verify their country’s relative position. The index in most parts of the world for Mode 1. of tradability can also be connected to trade patterns of This Tradability Index can be extended to a sector clas- countries along the intensive margin, as Jensen and sification that is more disaggregated or includes indexes Kletzer (2006) do. Information on the factor and institu- by exporters or importers. Policy makers can use these tional intensities for each sector yields additional policy Module 3: Assessing the Potential for Trade in Services 85 insights regarding which country characteristics can be used to compute bilateral trade potential. Subsequently, improved to increase tradability. the level of bilateral trade between a pair of countries rela- tive to trade potential is estimated to categorize bilateral exports as “overtraded” or “undertraded.” The regression Assessing Tradability through Gravity Models includes a country’s Services Trade Restrictions Index A gravity model can be used to assess whether a country is from the World Bank Services Trade Restrictions Database currently fully exploiting its trade potential with selected to assess whether these determinants are important in partners or whether country-specific barriers could be explaining the level of bilateral services trade (see Module removed to enhance integration. This section provides 4 for an explanation of the database). Bilateral trade flows an example in which there is scope to increase services are from the Trade in Services Database (described in trade integration between Association of Southeast Asian appendix B), which covers bilateral services flows for about Nations (ASEAN) countries at the aggregate level. The 200 countries across a multitude of sectors.1 analysis is based on an estimated gravity model of trade The average 2008–09 bilateral exports flows for 102 in services that is founded in microeconomic theory. countries are regressed on the following country-specific The gravity model of trade relates countries’ bilateral trade and bilateral characteristics: the log of distance, dummy flows to structural determinants of gross domestic product variables for contiguity, common language, a common (GDP), geographic distance, and other factors. The model colonial power, the Services Trade Restrictiveness Index of has been extensively used in international trade because of exporter and importer, and the log of GDP of exporter and its intuitive empirical and theoretical appeal. (Anderson importer (to proxy economic mass). The results are pre- and van Wincoop 2003, Feenstra 2004, and Baldwin and sented in the middle column of table 3.11. Taglioni 2006, among others, present exhaustive literature The right column of table 3.11 shows the results of reviews on the gravity equation as applied to interna- an alternative specification for the gravity equation in tional trade.) which the economic mass variable is picked up not by The structural determinants for each pair of countries GDP but by importer and exporter fixed effects (referred together with the estimated regression coefficients are to as a dyadic gravity equation).2 It shows the results of Table 3.11. Regression Results of Gravity Model of Trade in Services, 2008–09 Dependent variable: Log(export value) Module 3 Item Coefficient estimate Dyadic coefficient estimate Log(distance) –0.8526*** –0.9059*** (0.037) (0.029) Contiguity 0.3454** 0.4285*** (0.168) (0.110) Common language 0.9000*** 0.4314*** (0.124) (0.082) Common colonial power 0.3089 0.6241*** (0.217) (0.109) Importer Services Trade Restrictiveness Index 0.0012 (0.003) Exporter Services Trade Restrictiveness Index –0.0141*** (0.003) Log(importer GDP) 1.0866*** (0.021) Log(exporter GDP) 1.0808*** (0.021) Observations 2,533 4,925 Adjusted R-squared 0.700 0.813 Sources: World Development Indicators, Trade in Services Database, and Services Trade Restrictions Database (all from the World Bank) and Centre d’Etudes Prospectives et d’Informations Internationales (CEPII). Note: Coefficients on the fixed effects are not shown. Robust standard errors are in parentheses. ***p < 0.01, **p < 0.05. 86 Valuing Services in Trade regressing bilateral exports for 198 countries on the log data set included only data on cross-border trade (Mode 1) of distance and dummy variables for contiguity, com- and consumption abroad (Mode 2). Had FDI (commercial mon language, and common colonial power. Only bilat- presence or Mode 3) or the temporary movement of people eral characteristics can be included in a dyadic model, (Mode 4) been included, the coefficients may have been because the nation dummies prevent the inclusion of significant. country-specific variables such as GDP and the Services Figure 3.6 plots the residuals against the model’s fitted Trade Restrictions Index. Fixed effects (dyadic) estima- values for the dyadic specification as well as a specifica- tions control for a wide variety of country-specific fac- tion that does not correctly control for such barriers. The tors that affect bilateral trade flows, including omitted Services Trade Restrictiveness Index was removed from the variables that are too difficult to measure directly. These specification in order to focus on what happens when these variables include all country-specific (nonbilateral) trade barriers are not accounted for. Fitted values perform better policy barriers that cannot be measured by, for example, once these other obstacles in the specification with fixed the Services Trade Restrictiveness Index. These nonmea- effects are properly controlled for. surable country-specific characteristics fall in the residual Figure 3.7 plots the predicted bilateral trade levels in the specification in which GDP proxies economic mass from the specification with GDP on the x-axis and the (potentially biasing the estimates). predicted levels from the specification with fixed effects The results of the gravity model seem to suggest that on the y-axis. The dots represent Indonesia’s bilateral the Services Trade Restrictiveness Index of the exporting exports with other countries. Observations below the country matters in determining the bilateral exports of 45-degree line show that the predicted levels from the that country. Countries with more restrictive services reg- specification with fixed effects are lower than they are ulatory environments are significantly less likely to export for the specification with GDP. Because the specifica- services. The relationship between the level of restrictions tion with fixed effects properly controls for country- in the importing country and that country’s bilateral ser- specific barriers to trade, this result suggests that barriers vices imports is not statistically significant. However, the to trade at the national level are preventing Indonesia Figure 3.6. Residuals and Fitted Values of Gravity Model, Estimated with GDP and Fixed Effects, 2008–09 Module 3 a. GDP b. Fixed effects 10 10 8 8 6 6 4 4 Residuals Residuals 2 2 0 0 −2 −2 −4 −4 −6 −6 −5 0 5 10 15 −5 0 5 10 15 Fitted values Fitted values Sources: World Development Indicators, Trade in Services Database, and Services Trade Restrictions Database (all from the World Bank) and Centre d’Etudes Prospectives et d’Informations Internationales (CEPII). Module 3: Assessing the Potential for Trade in Services 87 Figure 3.7. Predicted Trade between Indonesia and Other Countries, Estimated with GDP and Fixed Effects, 2008–09 8 Log of predicted services exports SGP 6 based on fixed effects THA MYS 4 VNM PHL 2 BRN KHM 0 −2 0 2 4 6 8 Log of predicted services exports based on GDP Non-ASEAN country ASEAN country Sources: World Development Indicators, Trade in Services Database, and Services Trade Restrictions Database (all from the World Bank) and Centre d’Etudes Prospectives et d’Informations Internationales (CEPII). Note: ASEAN = Association of Southeast Asian Nations. For the International Organization for Standardization (ISO) abbreviations, see appendix B. Figure 3.8. Actual Trade and Predicted Trade of Gravity Model, Estimated with Fixed Effects, 2008–09 10 Log of actual services exports THA 5 PHL VNM SGP MYS BRN Module 3 KHM 0 −5 −2 0 2 4 6 Log of predicted services exports 95 percent confidence intervals 45-degree line Bilateral relationships with ASEAN Indonesia’s bilateral relationships Sources: World Development Indicators, Trade in Services Database, and Services Trade Restrictions Database (all from the World Bank) and the Centre d’Etudes Prospectives et d’Informations Internationales (CEPII). Note: ASEAN = Association of Southeast Asian Nations, BRN = Brunei Darussalam, KHM = Cambodia, MYS = Malaysia, PHL = Philippines, SGP = Singapore, THA = Thailand, VNM = Vietnam. from reaching its full trade potential with a particular Trade Restrictiveness Index, for example) and a low country. Lower potential trade after properly controlling level of regulatory governance may be dampening trade for country-specific obstacles suggests that barriers are potential. deterring services trade with ASEAN and non-ASEAN Figure 3.8 shows actual and predicted bilateral export countries. This result could suggest that a high level relationships with Indonesia (given by the dyadic gravity of regulatory restrictions (as captured by the Services equation). If an observation is above (below) the 45-degree 88 Valuing Services in Trade line, the average observed export relationship in 2008–09 by structural trade determinants differed from the actual was more (less) than what the gravity model predicts—on intraregional trade values between 2008 and 2009. the basis of countries’ structural determinants—and the Indonesia appears to have overtraded with Cambodia exporter is said to be overtrading (undertrading) with its and Brunei, undertraded with Malaysia and Singapore, trading partner. Estimates from the gravity model indi- and traded at the predicted level with the Philippines and cate that the estimated potential trade volumes predicted Vietnam. Annex Table 3A.1. Gini Class and Tradability of Selected Services Sectors Industry Gini class Tradable? Information (51) Newspaper publishers 1 No Radio and television broadcasting and cable 1 No Libraries and archives 1 No Wired telecommunication carriers 2 Yes Data processing services 2 Yes Other telecommunication services 2 Yes Publishing except newspapers and software 2 Yes Other information services 3 Yes Motion pictures and video industries 3 Yes Sound recording industries 3 Yes Software publishing 3 Yes Finance and insurance (52) Saving institutions, including unions 1 No Banking and related activities 1 No Insurance carriers and related activities 2 Yes Module 3 Nondepository credit and related activities 2 Yes Securities, commodities, funds, trusts, and other financial investment 3 Yes Real estate and rental (53) Videotape and disk rental 1 No Other consumer goods rental 1 No Commercial, industrial, and other intangible assets rental and leasing 2 Yes Real estate and rental 2 Yes Automotive equipment rental and leasing 2 Yes Professional, scientific, and technical services (54) Veterinary services 1 No Accounting, tax preparation, bookkeeping, and payroll services 1 No Architectural, engineering, and related services 2 Yes Other professional, scientific, and technical services 2 Yes Legal services 2 Yes Specialized design services 2 Yes Computer systems design and related services 2 Yes Advertising and related services 2 Yes Management, scientific, and technical consulting services 2 Yes Scientific research and development services 3 Yes Management (55) Management of companies and enterprises 2 Yes Administrative support (56) Waste management and remediation services 1 No (continued on next page) Module 3: Assessing the Potential for Trade in Services 89 Table 3A.1. (continued) Industry Gini class Tradable? Business support services 1 No Services to buildings and dwellings 1 No Landscaping services 1 No Employment services 2 Yes Other administrative and support services 2 Yes Investigation and security services 2 Yes Travel arrangement and reservation services 2 Yes Education (61) Elementary and secondary schools 1 No Colleges and universities, including junior colleges 1 No Other schools, instruction, and educational services 1 No Business, technical, and trade schools and training 2 Yes Health care and social services (62) Hospitals 1 No Nursing care facilities 1 No Vocational rehabilitation services 1 No Offices of physicians 1 No Outpatient care centers 1 No Offices of dentists 1 No Offices of optometrists 1 No Residential care facilities, without nursing 1 No Child day care services 1 No Home health care services 1 No Other health care services 1 No Office of chiropractors 1 No Community food and housing, and emergency services 2 Yes Offices of other health practitioners 2 Yes Arts, entertainment, and recreation (71) Bowling centers 1 No Other amusement, gambling, and recreation industries 1 No Museum, art galleries, historical sites, and similar institutions 2 Yes Module 3 Independent artists, performing arts, spectator sports, and related 2 Yes Accommodation (72) Drinking places, alcoholic beverages 1 No Restaurants and other food services 1 No Recreational vehicle parks and camps, and rooming and boarding houses 1 No Traveller accommodation 2 Yes Other services (81) Automotive repair and maintenance 1 No Barber shops 1 No Religious organizations 1 No Commercial and industrial machinery and equipment repair and maintenance 1 No Dry cleaning and laundry services 1 No Car washes 1 No Electronic and precision equipment repair and maintenance 1 No Civil, social, advocacy organizations, and grant making and giving 1 No Nail salons and other personal care services 2 Yes Other personal services 2 Yes Business, professional, political, and similar organizations 2 Yes (continued on next page) 90 Valuing Services in Trade Table 3A.1. (continued) Industry Gini class Tradable? Labour unions 3 Yes Footwear and leather goods repair 3 Yes Public administration (92) Justice, public order, and safety activities 1 No Administration of human resource programs 1 No Other general government and support 1 No Elective offices and legislative bodies 1 No Military reserves or national guard 1 No Administration of economic programs and space research 1 No Administration of environmental quality and housing programs 1 No Public finance activities 2 Yes National security and international affairs 3 Yes U.S. armed forces, branch not specified 3 Yes U.S. Coast Guard 3 Yes U.S. Air Force 3 Yes U.S. Army 3 Yes U.S. Navy 3 Yes U.S. Marines 3 Yes Source: Jensen and Kletzer 2006. Note: Figures in parentheses are two-digit North American Industry Classification System (NAICS) codes. Table 3A.2. Probability Results from Comparative Advantage Regressions Dependent variable: observed exports (1) (2) (3) Item Trade costs Endowments Institutions log(distanceij) –0.202*** –0.214*** –0.221*** (0.000) (0.000) (0.000) contiguousij 0.026 –0.004 –0.003 Module 3 (0.534) (0.928) (0.952) common official languageij 0.027 0.049 0.048 (0.331) (0.120) (0.129) colonyij 0.168*** 0.135*** 0.125*** (0.000) (0.004) (0.010) RTAij 0.132*** 0.146*** 0.142*** (0.000) (0.000) (0.000) log(high-skill labori) * log(high-skill intensityk) 0.020*** 0.018** (0.006) (0.017) log(non-ICT capital stocki) * log(non-ICT capital intensityk) 0.041*** 0.039*** (0.001) (0.002) log(internet users per 100 peoplei) * log(ICT capital intensityk) 0.021 0.065** (0.407) (0.018) legali * complexityk 0.115*** (0.000) legali * Herfindahlk 0.061*** (0.000) Observations 39879 32008 31524 Pseudo-R-squared 0.389 0.379 0.380 Source: van der Marel and Shepherd 2013. Note: Estimation is by logit with fixed effects by exporter, importer, and sector. Probability values based on standard errors corrected for clustering by country pair are indicated in parentheses below the coefficient estimates. Coefficients are reported as marginal effects at the mean. RTA = Regional Trade Agreement. *p < 0.10, **p < 0.05, ***p < 0.01. Module 3: Assessing the Potential for Trade in Services 91 Notes Ghani, Ejaz, ed. 2010. The Service Revolution in South Asia. Oxford: Oxford University Press. 1. The Trade in Services Database is currently the best source of data Gwartney, James, and Robert Lawson. 2012. Economics Freedom of on global trade flows in services, but the quality of trade data for services is the World 2012 Annual Report. Economic Freedom of the World far from comparable to trade data for goods. Because of the long tradition Network, Fraser Institute, Vancouver, Canada. of tariff revenues, trade data for goods are of high quality; because of the IMF (International Monetary Fund). Balance of Payment Statistics. intangibility and nonstorability of services, at-the-border duties cannot Washington, DC: IMF. be applied, which has led to much less accurate data. Jensen, J. Bradford, and Lori G. Kletzer, 2006. “Tradable Services: 2. Broadly, a fixed-effects regression model aims at capturing omit- Understanding the Scope and Impact of Services Offshoring.” In ted variables by creating a different intercept for the regression using a Brookings Trade Forum 2005: Offshoring While-Collar Work, edited dummy variable. by Susan M. Collins and Lael Brainard, 75–134. Washington, DC: Brookings Institution. Magdeleine, Joscelyn, and Andreas Maurer. 2008. “Measuring GATS References Mode 4 Trade Flows.” WTO Staff Working Paper ERSD-2008-05, World Trade Organization, Geneva. Anderson, James E., and Eric van Wincoop. 2003. “Gravity with Gravitas: A Maurer, Andreas, Yann Marcus, Joscelyn Magdeleine, and Barbara Solution to the Border Puzzle.” American Economic Review 93 (1): 170–92. d’Andrea. 2008. “Measuring Trade in Services.” In Handbook of Baldwin, Richard, and Daria Taglioni. 2006. “Gravity for Dummies and International Trade in Services, edited by Aaditya Mattoo, Robert Dummies for Gravity Equations.” NBER Working Paper 12516, M. Stern, and Gianni Zanini, 169–220. Washington, DC: World Bank National Bureau of Economic Research, Cambridge, MA. and New York: Oxford University. Barro, Robert J., and Jong-Wha Lee. 2012. “A New Dataset of Educational Mishra, Saurabh, Susanna Lundstrom, and Rahul Anand. 2011. Attainment in the World, 1950–2010.” NBER Working Paper 15902, “Service Export Sophistication and Economic Growth.” Policy National Bureau of Economic Research, Cambridge, MA. Research Working Paper 5606, World Bank, Washington, Chor, Davin. 2010. “Unpacking Sources of Comparative Advantage: A Washington, DC. Quantitative Approach.” Journal of International Economics 82 (2): 152–67. Trade in Services Database. World Bank, Washington, DC. Costinot, Arnaud. 2009. “On the Origins of Comparative Advantage.” UN (United Nations). 2012. Manual on Statistics of International. Trade Journal of International Economics 77 (2): 255–64. in Services 2010 (MSITS 2010) New York. Eaton, Jonathan, and Samuel Kortum. 2002. “Technology, Geography, van der Marel, Erik. 2011. “Determinants of Comparative Advantage and Trade.” Econometrica 70 (5): 1741–79. in Services.” GEM Working Paper, Groupe d’Economie Mondiale, Feenstra, R. C. 2004. Advanced International Trade: Theory and Evidence. Paris. Princeton, NJ: Princeton University Press. van der Marel, E., and B. Shepherd. 2013. “International Tradability Francois, Joseph, and Bernard Hoekman. 2010. “Services Trade and of Services.” World Bank Policy Research Working Paper 6712, Policy.” Journal of Economic Literature 48 (3): 642–92. Washington, DC. Gervais, Antoine, and J. Bradford Jensen. 2013. “Are Services Tradable? World Development Indicators (database). World Bank, Washington, Evidence from US Micro Data.” NBER Working Paper 19759, DC. http://data.worldbank.org/data-catalog/world-development National Bureau of Economic Research, Cambridge, MA. -indicators. Module 3 Module 4 Policy Options for Increasing Competitiveness and Trade in the Services Sector This module examines broad policy areas and options Regulatory policy barriers affecting trade, investment, for addressing the constraints identified in Modules 1, 2, and labor mobility include the entire range of policies and 3. It distinguishes between sectoral policies that can be affecting cross-border trade, consumption abroad, for- implemented in the short to medium term (mostly regula- eign investment, the participation of multinationals, and tions and barriers to trade) and horizontal measures and the movement of individual services providers. These strategies related to the domestic economy and institu- policy barriers are common across services sectors. They tions, which can be implemented only over a longer time seek to address market failure problems. They are not horizon. The module addresses policies that affect trade in explicitly targeted to trade in services, but they neverthe- services, including investment policies, temporary migra- less affect trade. tion policies, and domestic policies that may have unin- Trade barriers explicitly affecting trade, investment, tentional negative impacts on trade. It does not provide and labor mobility in services include barriers related to prescriptive advice on specific policies that should be accessing a market or the differential treatment of foreign adopted. It does include case studies illustrating good and and incumbent domestic firms. Barriers in this category bad policy practices. include indirect barriers, such as domestic regulation or The focus is on identifying regulations that affect or can wider regulatory measures that affect trade. facilitate the importation of input services that are determi- nants of a country’s competitiveness, including regulations Reasons for Regulatory Policies that may affect the competitiveness of services exports. For example, regulations that limit the participation of foreign Analyzing the determinants of competitiveness is more direct investment (FDI) in a sector affect the export of complex for services than for goods because of the broad services. A liberal policy toward FDI has helped turn the definition of trade in services, which encompasses fac- Philippines into the world’s third-largest exporter of busi- tors of production; the complicated links between modes, ness process outsourcing (BPO). In Malaysia, the estab- inputs, and outputs; policies stemming from different lishment of branches of foreign universities has boosted areas and having different social objectives; and new trade exports of education services. patterns emerging as part of technological change. Trade Domestic enabling factors reflect a country’s endow- can take various forms other than cross-border trade, ments. They include human capital, natural resources including the movement of services and the movement (such as sights that attract tourists), infrastructure (such of FDI and skills transferred through temporary labor as a telecommunications network that facilitates the deliv- mobility schemes. ery of services), and institutions (such as the quality of a For some type of services, modes of supply may be sub- country’s rule of law or regulatory environment). These stitutes or complements. If modes are perfect substitutes, factors are an important complement to good trade poli- the liberalization of one mode would be enough to fully cies and regulations. reap the gains from liberalization. If modes are imperfect This module examines two sets of policy options substitutes, or complements, their provision normally that drive the performance of services competitiveness.1 requires a mix of cross-border trade through Mode 1, 93 94 Valuing Services in Trade Mode 2, or both and locally supplied services through people, technology, and long-term business relationships; Modes 3 or 4. Services provision may require complemen- and services that coordinate dispersed production in dif- tary inputs. Delivery of one service may thus depend on ferent locations. Miroudot, Lanz, and Ragoussis (2009) the use of another. Financial services, for instance, cannot find that intermediate inputs represent 56 percent of be provided without telecommunications and professional goods trade and 73 percent of services trade. Trade flows services. are thus dominated by products and services that are not Regulations affecting trade in services are linked to consumed but instead used in the production of other policies in a wide range of areas beyond the sector itself. goods and services. Exporting health care services, for example, requires Trade costs are higher for services than for goods—as addressing migration and labor market regulations that much as twice as high, according to Miroudot, Sauvage, prescribe wage payments to services providers, taxation and Shepherd (2010). Policies can raise these costs and requirements and social security regulations, and recogni- restrict the growth and expansion of trade in services, as tion of professional titles and experience. illustrated by countries that have liberalized the telecom- Technical changes have made services trade more munications sector but have not put in place an adequate interlinked with other areas and affected trade patterns. regulatory framework (box 4.1). Changes in information and communication technology Measures designed to reserve a market for domestic (ICT) reduce the need for proximity between producer incumbents explicitly insulate them from foreign com- and consumer. They also allow for the fragmentation of petition. Other measures, such as policies implemented production into tasks that may be performed in different in the pursuit of social objectives, may not be protection- locations (see, for instance, Jones 2000, Grossman and ist in intent but often unintentionally restrict trade in Rossi-Hansberg 2008, Feenstra 2010, and Helpman 2011 services. for the theoretical underpinnings). Fragmentation, which The difference between explicit and unintentional affects the production of both goods and services, means policy measures can be grasped by comparing tariffs with that a vertically connected production process that takes regulatory barriers. Tariffs are almost always protectionist place in one location can now be undertaken in different in intent and effect. In contrast, regulations affecting ser- regions or countries. vices sectors often promote other objectives such as equal These trends have increased the interdependence access or universal service provision. An antiexport bias on among trade, FDI, and the temporary mobility of both trade in services arises if these regulations are implemented high-skilled and low-skilled workers, complicating policy. in a way that is economically inefficient and increases cost, New trade patterns are emerging as a result of these tech- by restricting free trade in services. nological changes. Baldwin (2011) suggests that whereas Services regulation intended to achieve legitimate policy trade in the 20th century was mainly about selling goods objectives—in health, education, and the environment, made in factories in one country to customers in another for example—may become regulatory barriers to trade in country, trade in the 21st century is characterized by the services. Because services trade invades areas traditionally trade-investment-services nexus that reflects the inter- reserved to domestic constituencies (such as education and twining of trade in parts and components; the interna- health), removing regulatory barriers is politically difficult. tional movement of investment in production facilities, In addition, such measures are often technically complex Box 4.1. Weak Policies, Weak Investment Climate, and Weak Integration into World Markets in the Middle East and North Africa Module 4 The countries of the Middle East and North Africa are still poorly integrated into world markets, and they have joined global pro- duction chains at a lower level than other developing countries. These countries receive only a small share of global foreign direct investment (FDI) flows, and their participation in world trade is below its potential. The share of components in manufactured exports, for instance, remains far below that of Malaysia or Taiwan, China (Ng and Yeats 2000). The weak investment climate and the poor quality of backbone services have also contributed to poor performance in several strategic sectors and dampened the competitiveness boost of other measures, including trade liberalization initiatives and the sign- ing of association agreements with the European Union. Quantitative studies find that the quality of procompetitive regulation in the telecommunications sector increases trade in manufacturing goods and helps countries join global production chains, reducing the regulatory risk perceived by investors. Source: Rossotto, Sekkat, and Varoudakis 2003. Module 4: Policy Options for Increasing Competitiveness and Trade in the Services Sector 95 and difficult to understand. They are complex because of services sectors directly. Market interventions affecting the wide range of policy barriers that affect services and services sectors can be justified on welfare grounds if they the many types of market failures that characterize services correct a market failure. Market failures are often linked markets. to social objectives. In the health care sector, for instance, Market failures that need to be addressed through governments often enjoy a mandate to regulate with the policy include the following (figure 4.1):2 goal of achieving equal access by all. Regulations should correct market failures in the most • failures that limit competition efficient and effective way, where “efficient” means that • failures related to the public good characteristics of a they achieve their goals at the minimum cost in terms service and divergence with important and necessary of economic welfare and “effective” means that they noneconomic social goals achieve the goal in question. Policy makers must achieve • failures caused by asymmetric information social objectives in the least restrictive manner when • failures caused by divergent private and social utility designing, implementing, and applying regulations, so functions. that optimal allocation of goods and services production is guaranteed. Although there are good reasons to develop regula- One way to help overcome regulatory failures is to create tory policies to correct domestic market failure, these strong, transparent, and accountable regulatory bodies. policies have potential economic risks, as table 4.1 shows. Good regulatory policies can be designed through a pro- Regulatory policies can also impede trade because countries cess that reflects the main principles of good governance. differ in the choice and stringency of instruments, because Ideally, these processes and procedures should increase regulations are deliberately designed to protect domes- transparency, through consultation that increases the tic service providers, or because the administration and accountability of regulators. Transparency and account- enforcement of regulations is inadequate. ability reduce regulatory capture while contributing to When regulations are designed to protect domestic informed and well-designed regulations that are adequately incumbents or safeguard local interest groups, they are implemented. Expert advice and interagency coordination often accompanied by inadequate regulatory planning, mechanisms increase efficiency and enhance analytical weak administration, and poor public oversight. The risk support for the measures, reducing the scope for regula- of regulatory failures related to capture by interest groups tory failures (caused by bounded rationality, for example) is especially high when regulatory policies are motivated by (Molinuevo and Sáez 2014). A case of good practice is illus- noneconomic purposes. Such policies are often based on trated by the Asia-Pacific Economic Cooperation (APEC) instruments whose economic rationale is void of analytical legal service website initiative, discussed in box 4.2. or empirical support. Regulation to correct market failures may lead to mar- Natural Monopoly or Oligopoly ket distortions and welfare losses, even when government or regulatory bodies lack the capacity and means to reg- Because a monopoly can restrict supply and keep prices ulate effectively. This problem is most likely to arise as a high, consumers suffer and economic welfare is reduced. result of laws and regulations that are not aimed at the Oligopolies can result in similar effects, through price Figure 4.1. Types of Market Failures Services with public good Divergence between private Limited competition Asymmetric information characteristics that achieve Module 4 and social utility functions noneconomic goals • Natural monopoly • Between consumer and • Education, health, finance, • Underdeveloped domestic • Oligopoly producer telecom, public good services sector • Exclusive or controlled • Between consumer and features • Lack of technological access to natural financial services • Health and safety protection upgrading resources provider • Environmental safety and protection • Equal service access • Unemployment reduction Module 4 96 Table 4.1. Instruments for Addressing Market Failures Affecting Trade in Services Type of market failure or policy goal Policy objective Instruments Affected sectors Economic risk Market failure Natural monopoly or • Avoid duplicating fixed costs. • Information on price setting Network industries, such • Prices set so low that monopolist oligopoly • Create optimal environment given • Cross-subsidies as water, electricity, and unable to recoup fixed costs limited competition. telecommunications • Costly to obtain perfect • Splitting of networks from service provision information Necessity of exclusive • Optimally allocate exclusive access. • Licensing for usage, through auctioning Radio transmission, mobile • Negative spillover effects because or controlled access to • Ensure competition without regulation. telephony, and other utilities of overuse of bandwidth natural resources • Avoid negative externality through • Insufficient number of firms given spectrum signal interference. license Public goods • Achieve social objectives (benefits to • Cross-subsidies Health care, education, • Underproduction of positive characteristics of society at large). • Prudential regulation utilities, and financial services spillover effects service • Underproduction of the service • Free-riding Asymmetric • Protect consumers. • Licensing Financial, insurance, and • Overregulation information between • Reduce consumer safety concerns. • Education, including demonstration of work professional services • Moral hazard consumer and • Guarantee quality of service. experience • Strict licensing requirements, producer • Minimum skill requirements for services which may restrict supply providers • Quality control • Compulsory insurance schemes Asymmetric • Increase stability of financial sector. • Financial information transparency Investment banking, retail • Overregulation information between • Licensing banking, private equity firms, • Moral hazard consumer and hedge funds, insurance • Minimum standards or capital requirements financial services services provider • Strong central bank • Capital adequacy tests No market failure • Achieve other legitimate policy • Promotion of universal access in remote areas Network industries; services • Increased costs because of objectives, such as equal services access • Measures to improve redistribution provided through Mode 3 higher factor costs or inefficient or reduction of unemployment. (investment) organization • Job creation Policy goal Environmental • Promote social welfare rationale • Effective rural and urban planning Tourism, retail, • Regulations that are more protection • Reduce negative environmental transportation sectors burdensome than necessary to externalities. achieve desired goals • Conserve the environment. Development of • Generate positive spillover effects for • Restrictions on legal form of firms All services sectors • Overregulation of FDI domestic services firms in rest of sector. • Rules on local employment of personnel/ • Risk aversion of foreign investors sector; upgrading of directors • Barriers to trade in services technology Module 4: Policy Options for Increasing Competitiveness and Trade in the Services Sector 97 Box 4.2. Good Policy Practice: Increasing the Transparency of Regulations Governing Lawyers Working in Asia-Pacific Economic Cooperation (APEC) Member Economies A website of the Asia-Pacific Economic Cooperation (APEC) (http://www.legalservices.apec.org/) enhances the transparency of regulation of legal services. It explains the requirements and procedures for practicing foreign law in the organization’s 21 member economies. The site provides information on the licensing and rights of foreign lawyers providing foreign and international legal services. It also provides information for foreign lawyers who work in association with host economy lawyers. The website provides information about APEC members’ regulatory regimes for foreign lawyers, compares regulatory arrangements, and helps develop best practice for the regulation of foreign lawyers. Figure 4.2. Options, Constraints, Levers, and Risks Associated with Regulating Natural Monopoly or Oligopoly Responses/policy Levers Risks Issues for policy Constraints makers options Efficiency costs Regulation of Costly to Technological Profitability of monopolist to eliminate progress to operator Rents ensure pricing information gap expand Costly to Supply close to between information, monitor restrictions economic value monopolist/ competition, the quality of High prices Universal access oligopolist and and unbundling service Welfare costs through regulator opportunities Price fixing universal service Need to ensure obligation, profitability of Collusion tolerance for operator Lack of incentive cross-subsidization, to innovate or both fixing or collusion. Both setups can have negative effects given governments’ imperfect information on incumbents’ on the economy, because they reduce the incentive to cost structures, natural monopolists charge more than innovate, dampening long-run growth prospects. competitive firms. In some industries, competition is “naturally” limited. A second policy response to a natural monopoly has In network industries, such as gas, electricity, and telecom- been to try to ensure universal access to the network in munications, the fixed costs of creating a network are enor- question (box 4.3). One way of doing so is to impose a uni- mous. In such cases, it is often efficient for a single firm to versal service obligation on the supplier. Such an obligation create the network, so that the economy avoids paying for is generally combined with tolerance of cross-subsidized more than one network. This situation is referred to as a pricing to support the extension of the network to remote Module 4 natural monopoly. Figure 4.2 summarizes the options for areas. Under this model, consumers in urban areas, where regulating this type of market failure. network construction is relatively inexpensive, pay some- Policy responses to natural monopolies cover two what higher prices than they otherwise would in order to important areas. One is regulation to ensure that pricing subsidize network construction in rural areas. Some con- remains as close as possible to competitive levels. In a situ- sumers therefore end up paying somewhat less than they ation of perfect information, it is possible for the govern- would if they had to fully cover the costs of network con- ment to implement regulated pricing. However, prices still struction themselves. need to be high enough for the monopolist to recoup its The regulatory response to natural monopolies has initial investment in setting up the network. In practice, changed over time, at least in developed countries, 98 Valuing Services in Trade Box 4.3. Good Policy Practice: Achieving Universal Postal Coverage in Trinidad and Tobago Before reform, the postal service sector in Trinidad and Tobago was characterized by weak financial results, low investment, declining mail volumes, and poor service. On average, mail delivery took 7–10 days, and services reached only half of all households. The government’s reform strategy was to outsource management of TT Post, in the hope of transforming the company into a viable supplier. As part of a two-phased reform process, a delegated management arrangement was set up for five years with the goals of increasing customer satisfaction, achieving universal delivery, improving the quality of service, and raising total revenue and net income. All targets except net income were achieved or nearly achieved (daily coverage rose from 50 percent to 94 percent). To provide the legislative and regulatory framework needed to support a management contract, the Postal Act of 1999 was established. It provides the new public postal corporation with more management autonomy and commercial flexibility. By engag- ing stakeholders, including major mailers, labor unions, customer groups, and private postal operators, the government built a broad consensus for reform. Source: Guislain 2004. primarily as a result of technological progress. The domain or oligopoly. The problem in this case is not how to regu- of natural monopolies has narrowed, as activities such late an environment with limited competition but how as network creation and maintenance have been split off to ensure that competition without regulation does not from activities such as the provision of services to end- impose negative externalities on consumers through signal users. Electricity markets were traditionally considered interference. local natural monopolies, but the activities of generation Licensing is a common solution to this problem. The and distribution can now be technically separated, with spectrum is broken up into bands, each of which is given distribution based on a competitive service delivery model. to a single service provider, which is granted a license to Even the local monopoly of generation services has come use that particular bandwidth. A regulator decides how into question, as national or regional grids and “smart” much bandwidth is necessary for commercial operations grids have made it possible for generators to compete for and then allocates it in a way that prevents signal interfer- market share over a wider area. ence. The main economic question is how best to allocate Regulation has kept pace with these technological licenses to ensure effective competition among licensees changes by focusing on the need to expand competition (this issue is discussed below in the context of regulatory wherever possible, in particular through the division of strategies). One model involves identifying a fixed num- activities. The crucial role for policymakers and regulators ber of licenses based on bandwidth requirements and then once competitive activities have been divested from his- auctioning them to firms. The regulator can use the auc- torical operators is to regulate network access. The goal is tion to ensure that a sufficient number of firms enters the to create equal access to all competitors on conditions that market to ensure a reasonable level of competition. are neutral, even if some local monopolies remain. Public Good Characteristics Exclusive Access to Natural Resources Only a few services have a pure public good characteristic. Limited competition necessarily arises when firms require Many services have strong public good characteristics, exclusive access to a scarce resource in order to do business however. Public goods are defined in terms of nonexclud- (figure 4.3). Sectors such as radio transmission or mobile ability (the fact that consumers cannot be individually telephony are good examples: both require exclusive access charged based on their consumption) and nonrivalry (the Module 4 to parts of the electromagnetic spectrum. If two firms use fact that one person’s use of a good does not diminish the the same part of the spectrum, they will be unable to do ability of another to consume it). Public goods create posi- business, because interference will make signal transmis- tive externalities, which entail a social benefit. sion impossible. It is therefore in the public interest that Many services have public characteristics in some situ- scarce resources such as the electromagnetic spectrum be ations. Health services, for example, have public good allocated so that each firm has exclusive rights to a particu- characteristics only in limited cases, such as vaccinations lar part of the resource. or control of epidemics. For most types of health care ser- The policy response to this kind of limited competi- vices, the consumer is the primary beneficiary. Many coun- tion is different from the response to a natural monopoly tries nevertheless treat health care services as a public good, Module 4: Policy Options for Increasing Competitiveness and Trade in the Services Sector 99 Figure 4.3. Options, Constraints, Levers, and Risks Associated with Regulating Exclusive or Controlled Access to Natural Resources Responses/policy Levers Risks Issues for policy Constraints makers options Efficiency costs Licensing Limited market Technological Collusion Rents Establishment of size may not progress and among an authority for allow sufficient convergence to operators Supply number of expand restrictions competition in Costly to monitor the sector licenses competition quality of service High prices Need to ensure Welfare costs profitability of Price fixing operators Collusion Lack of incentive to innovate Box 4.4. Good Policy Practice: Increasing Access to Health Services in Thailand through Regulation Rather Than Provision In Thailand, the government’s role has changed from providing health services to regulating the health services market. During the last two to three decades, the government has moved toward a policy of contractual arrangements in which independent public and private institutions provide services. The shift happened against the background of economic growth during the 1990s. The introduction of the Social Security Scheme in 1990, medical benefits provided by employers, and rising household demand spurred the rapid growth of private hospitals. In response, the government adopted policies regulating entry and quality control. Although problems—such as oversupply in urban areas—remain, Thailand seems to have moved in the right direction, giving more and more people access to health services. Regulation is divided into regulation of health care institutions and regulation of professionals. The Ministry of Public Health is responsible for health institutions, controlling the licensing and renewals of private clinics and hospitals. The Social Security Scheme plays an active role through the price system: providers are paid a flat rate per registered worker regardless of the services they require. Professional organizations play a key role in granting certification and licensing for health professionals. The government created strong regulatory rules that require professionals to work a minimum number of years in community hospitals. Price con- trols for fees, especially for doctors in private hospitals, are weaker than they are for institutions. Quality performance and enforce- ment seem to require improvements for both health institutions and professionals. Source: Teerawattananon and others 2003. even though the real justification for regulating them between private and social costs and benefits can result is to ensure the social objective of caring for sick people in underproduction of the good. The high costs of higher Module 4 (box 4.4). education, for instance, dampen demand, reducing the Education, which is widely regarded as a public good, is positive externalities created. The second is the free-rider actually only partially a public good, as a large share of the problem (the fact that some people will consume a good benefits accrue to the individual being educated (as wage without paying for it). For example, some neighbors may differentials by level of education indicate). Society at large not be willing to pay for garbage collection, assuming that also benefits from education, however, making it a public others will. This problem removes the incentive to produce good as well. the service. Regulation that aims to create adequate incen- Regulation of pure or partial public goods is necessary tives to correct these market failures is therefore neces- because of two problems. The first is that the difference sary to ensure that public goods are produced in socially 100 Valuing Services in Trade optimal quantities. To do so, developing countries have Dealing with information asymmetries is a strong eco- adopted various policies over the last decade, including nomic rationale for licensing restrictions in some services public-private partnerships (such as the contracting out sectors. The question is how strict these restrictions should of selected services or facilities), the development of pur- be. A related concern is the extent to which regulatory chasing arrangements, franchising, and the use of vouchers procedures should prescribe particular activities, such as (figure 4.4). completion of training programs or demonstration of cer- tain competencies. Many licensing regimes impose both sets of requirements. Aspiring lawyers in the United States, Consumer Protection for example, must demonstrate that they have completed In professions such as law and medicine, failure to provide a period of study at a U.S. or foreign law school and then adequate services can have serious consequences for con- pass an exam to demonstrate competency. The nature and sumers. Low-quality services in the transport sector also scope of these types of requirements is crucial in assess- pose safety problems. ing the extent to which they achieve the social objective of The problem can be analyzed through the lens of asym- minimizing information asymmetries in an economically metric information: consumers want to use “good” ser- efficient way. vices providers but have no way of assessing quality until Assessing the quality of a service is even more difficult after they consume the service. Regulation has a role to than assessing the skills of a provider. Even highly trained play in establishing mechanisms that allow only “good” surgeons occasionally make errors of judgment or perform providers to operate. an operation in a suboptimal way. Some countries and sec- Licensing is one way of dealing with this market failure. tors deal with this problem through compulsory insurance Only people who meet certain standards of education and schemes. In professions such as law, for example, services training are authorized to practice law or medicine in most providers are often required to purchase liability insurance countries (figure 4.5). Commercial drivers are required to to ensure that if the service is provided in an unsatisfactory be licensed in most countries. Regulation can never elimi- manner, the consumer has financial recourse against the nate information asymmetries: service providers of varying provider and can be compensated for any losses caused as skill levels will always be present in the market. Licensing a result. requirements can help exclude providers that fail to meet Governments can also require providers to be nationals or certain standards, however. residents of the country in which the consumer is located. Figure 4.4. Options, Constraints, Levers, and Risks Associated with Regulating Public Goods Responses/policy Levers Risks Issues for policy Constraints makers options Difference Public-private Lack of private Collaboration Quality between private partnerships sector interest in with private performance and social costs Purchasing providing the sector and and benefits arrangements service organizations enforcement may result in for granting underprovision Franchising certification and of the service Vouchers licensing Module 4 Free-riding may Licensing reduce incentive schemes to produce and Price controls innovate Public service obligations Module 4: Policy Options for Increasing Competitiveness and Trade in the Services Sector 101 Figure 4.5. Options, Constraints, Levers, and Risks Associated with Addressing Health and Safety Concerns in the Presence of Asymmetric Information Responses/policy Levers Risks Issues for policy Constraints makers options Asymmetric Licensing Impossibility of Increasing Reducing information schemes obtaining quality market between Certification perfect competition by contestability provider and requirements information liberalizing Creating a user market closed Compulsory to foreign Risk of moral insurance profession hazard providers schemes Nationality restrictions or residency obligations This type of restriction helps ensure that the consumer can consumers may not be able to identify good banks. Most sue the services provider for inadequate service (the cost governments therefore also establish a licensing system and complexity of suing a defendant abroad make doing for the banking sector (figure 4.6). This kind of regu- so very difficult). Alternatively, governments can require lation forces banks that accept deposits from the pub- foreign providers to keep a minimum level of assets within lic to meet minimum standards of capital adequacy. the jurisdiction (capital requirement). Doing so provides To limit the possibility of bank runs, most economies consumers with an asset pool that can be drawn from if also require banks to participate in a deposit insurance necessary. scheme. Banks are usually also subject to a sectoral regu- lator, which can be the central bank or an independent body. This body needs to have wide powers of over- Financial Sector Stability sight and intervention in order to force banks to remain Financial services are a topic of special public policy inter- financially sound. est because of their centrality in the economic system. The Outside banking, particularly the retail banking sec- importance of this sector was made clear during and after tor, regulation is also required. Insurance companies are the global financial crisis of 2008–09. One justification for generally subject to licensing requirements with capital special policy intervention in the financial sector has to do adequacy tests. They are also supervised by regulators with information asymmetry, which is particularly great with broad powers. Other financial services providers are in retail banking. Consumers want to invest their money sometimes subject to licensing requirements, particu- Module 4 in good banks and good financial instruments, but they larly when they deal with small retail investors or engage cannot judge for themselves whether a bank is financially directly in market transactions. Wholesale financial insti- stable. The flip side of this problem is that sudden distrust tutions, such as investment banks, private equity firms, of a bank can create a run that could lead to the failing of and hedge funds, are less heavily regulated than retail the bank, even if it was healthy. banking, because the problem of information asymme- A variety of policy responses address this problem. try is assumed to be less severe. Unlike retail consumers, Publicly listed banks can be required to disclose substan- wholesale investors are assumed to have sophisticated tial financial information. Even with such disclosures, knowledge of finance. They are expected to calculate however, which are usually difficult to understand, risks they take. From a moral hazard vantage point, it is 102 Valuing Services in Trade Figure 4.6. Options, Constraints, Levers, and Risks Associated with Dealing with Asymmetric Information between Providers and Consumers of Financial Services Responses/policy Risks Issues for policy Constraints makers options Moral hazard Prudential Risk of bank runs Reduced lending Negative regulation Limited room capacity externalities to Strict oversight for manoeuver Risk aversion the rest of the Licensing because of risk Overregulation economy of contagion and financial meltdown (bank runs and existence of "too big to fail" institutions) Box 4.5. What Are the Effects of Planning and Zoning Regulations? Many studies have been conducted on the effects of regulation related to planning and zoning of the retail sector, most of them on developed countries. After reviewing these studies, the Australian Productivity Commission concluded that land use regulation that centralizes retail activity can either enhance or reduce competition. The effects depend on how regulatory policies are designed and implemented. Research on the impact of regulations that restrict the establishment of large-format stores suggests that such regula- tions reduce retail productivity and retail employment and raise consumer prices. Other evidence suggests that land use regulations raise residential and commercial property prices. The Productivity Commission concluded that in order to reduce the anticompeti- tive effects of zoning, policy makers need to make sure that sufficient floor space is allocated to retailers and that there are larger zones. These measures can increase productivity and eventually welfare by allowing new and innovative firms to enter the local market while allowing existing firms to expand. Source: Australian Productivity Commission 2011. appropriate that they be left to bear the consequences of Planning restrictions are particularly important in the their decisions. retail sector. Many jurisdictions limit the urban areas in which supermarkets and other large retailers can be estab- lished, in an attempt to reduce environmental externalities. Environmental Protection Module 4 Restrictions that prescribe the use of roads at specific hours Environmental protection is frequently used to justify of the day or the size of the vehicle affect market participants regulation of products and packaging. Although it is a less who need to establish transportation and distribution cen- obvious reason for restricting trade in services, it plays an ters, which are necessary for supermarkets and large retailers. important role in some situations. One is town planning. Another rationale for regulation is the desire to protect small Many localities impose restrictions on the types of busi- retailers from competition (box 4.5). On economic grounds, nesses that can establish facilities in particular areas. Other this rationale is weaker, because there is no market failure examples include regulations affecting tourism and trans- that needs to be corrected. Indeed, such regulation can have port services. negative consequences on consumer welfare. The question, Module 4: Policy Options for Increasing Competitiveness and Trade in the Services Sector 103 therefore, is not how to implement these regulatory restric- To stimulate technology spillover, governments often tions in the most efficient way but how regulators can put impose an obligation that a foreign investor work with a the interests of the majority (consumers) ahead of the local firm, through a joint venture or similar arrange- interests of an organized minority (store owners). ment (figure 4.7). A joint venture is a business agreement in which the foreign and local firm agree to develop a new product, service, or production procedure. Restrictions on Technology Transfer the legal form of foreign investment that stimulate coop- The presence of foreign investors can generate positive eration with local firms can increase technology spillover. spillover effects or externalities for domestic firms in the Another way of ensuring that firm-specific technologi- same sector, particularly through technology upgrad- cal factors of a foreign company become embedded in the ing and the development of workforce capabilities. Some local economy is to require that the foreign firm hires local developing countries believe that multinational corpora- workers, who learn from exposure to foreign technologies tions create such spillovers. and practices. Restrictions can also be imposed that require For services, the term production technology needs to be foreign firms to employ some local directors or managers understood in terms that include all means by which a firm (box 4.6). These regulations are usually considered to be produces its services, including management competence, barriers to services trade. Trade agreements aim at elimi- improvement of organization structures, and the methods nating them. used to produce a tailored service for consumers. Some types of “technology” increase productivity within the firm Trade Policy Barriers and, in the longer run, the sector. Triplett and Bosworth (2004) find that services sectors accounted for most of the Limitations or restrictions on trade in services take many improvements in labor productivity and total factor pro- different forms, discriminatory and nondiscriminatory, ductivity in the United States after 1995. They note that direct and indirect. These measures are usually more com- especially in the financial and distribution sectors, produc- plex than trade restrictions on goods. tivity accelerated, in large part because of managerial inno- Barriers may be aimed to the service itself or the services vations such as outsourcing and specialization, as well new provider. If, for example, credit cards can be provided only by concepts of retailing. established institutions in a particular country, the credit card Figure 4.7. Options, Constraints, Levers, and Risks Associated with Promoting Technology Transfer Responses/policy Levers Risks Issues for policy Constraints makers options Enhancing Restrictions on Foreign Leapfrogging Reduction in management legal forms of operators’ opportunities technology competences FDI that require interest in the transferred to Enhancing ICT foreign firms domestic market domestic empowerment work with services sector domestic Increasing use operators of efficient production Local methods, employment Module 4 including requirements outsourcing of noncore actitivities and specialization Note: FDI = foreign direct investment, ICT = information and communication technology. 104 Valuing Services in Trade Box 4.6 How Can Developing Countries Increase the Technological Capabilities of Their Services Sector? Several routes can increase technological capabilities in the host country: • encouraging trade and providing aid to strengthen production by domestic firms • providing FDI and contracting with domestic firms to build local export-oriented companies • developing local subcontracting capacity. Sector development, and eventually higher productivity and growth, increase only slowly, especially for slow and very late industrializers. Slow industrializers are countries in which the manufacturing sector accounts for no more than 15 percent of GDP (examples include Algeria, India, and South Africa). Very late industrializers are low-income countries with even smaller manufactur- ing bases. Experience in East Asia suggests that acquiring established technologies and building up technological capabilities based on their application can spur growth. For some very late industrializers, the conventional route of industrial development may not be appropriate. Low-income countries often lack an industrial base or have an agriculture sector that contributes only marginally to economic output. Some developing countries, therefore, have adopted the strategy of leapfrogging—skipping more “traditional” development paths. Firms in China, for example, have moved from producing products for the domestic market to production activities through joint ventures that are organized around global supply chains. Leapfrogging often takes place in services sectors. Low-tech leapfrogging is found in the tourism industry. Developing countries have capitalized on their good weather and labor resources. Governments have invested in and imported management skills. Recent examples of leapfrogging include software development and lower-tech back office functions, such as call centers. What is needed in such cases are good technological infrastructure and telecommunication systems, which can be obtained through foreign participation. A crucial difference between goods and services is that for services, countries need only to adapt the technology to provide the service; they do not need to replicate or develop technologies. Adapting the technology requires good-quality infrastructure, the ability to add value, and the availability of technically trained labor. Source: Bennet 2002. Table 4.2. Examples of Restriction on Trade in Services, by Mode Mode 1 Mode 2 Mode 3 Mode 4 (cross-border trade) (consumption abroad) (commercial presence) (movement of natural persons) • Local registration; • Local registration of • Full or partial acquisition of existing • Permission for specific categories local agent; local offshore provider— businesses not permitted. of personnel subject to approval presence; and applied on a transparent, • Restrictions on establishment of new and labor market tests. authorization, readily accessible, and businesses. • Approval for intercorporate license, or permit nondiscriminatory basis— • Reservation of some sectors or transferees and specialists required to market required to market services. activities, state-owned enterprises subject to general economic or supply services. • Transfer of capital, payments, to be privatized, and government- needs test. or use of credit cards for contracted services, for investment • Only intracorporate transferees transactions subject to only by nationals or permanent permitted, subject to a limit authorization. residents. of two foreign transferees per • Trade permitted only • Requirement that providers operation and mandatory through firms with established in one part of a country training of local staff. commercial presence in have a minimum number of • Provision of services by country or specific “brand- resident providers of their agents self-employed persons not name” entities. for provision in another part of a permitted. country. Source: Adapted from Thomson and Nielson 2001. service is affected. If instead a regulatory policy prescribes market access through all modes of supply. Table 4.2 pro- Module 4 that credit card services can be provided only by financial vides other examples of some restrictions by mode. institutions owned or controlled by nationals, the restriction Barriers to trade can be discriminatory or nondiscrimi- affects both the service and providers of credit card services. natory. Discriminatory measures target foreign providers; The provision and trade of a service can be limited through nondiscriminatory measures affect all firms, foreign and specific modes of supply. When, for example, the national domestic. An example of a discriminatory barrier is a tax government is the sole provider of a service (as in the case of imposed only on foreign services or providers. An example education and health care services in some countries), for- of a nondiscriminatory barrier is the granting of a market eign and domestic providers are not allowed to provide the to a monopoly, which prevents all other firms, foreign and service through any mode of supply. This measure affects domestic, from accessing the market (table 4.3). Module 4: Policy Options for Increasing Competitiveness and Trade in the Services Sector 105 Table 4.3. Policies Affecting Trade in Services • the total number of service operations or volume of ser- Policies on entry/ Policies on ongoing vice output Nature of policy establishment operations • the total number of natural persons that may be Nondiscriminatory Licensing Quality and safety employed in a sector requirements requirements Discriminatory Special Requirement that • restrictions or requirements specific to the type of legal requirements for foreign operators entity through which a service supplier can supply a foreign operators comply with additional service requirements such as rules governing • limits on foreign ownership (participation of foreign technology transfer capital). Source: Deardorff and Stern 2008. Barriers to services trade distort trade patterns and reduce welfare gains. Based on data from the World Barriers affecting trade in services can be explicit Bank’s Services Trade Restrictions Index (STRI) Database (de jure) or implicit (de facto). Explicit measures are (box 4.7), Borchert, Gootiiz, and Mattoo (2012a, 2012b) clear, known by firms, and comprehensible; they are for- find that merger and acquisition activity by foreign firms mulated in a transparent way. When a regulatory barrier is lower in countries that restrict ownership to a minor- is implicit, it may be created by the effect of the regula- ity stake. Other significant barriers include discrimina- tions on the provider. A language requirement, for exam- tory licensing criteria, restrictions on the repatriation of ple, may prevent a foreign entity from providing a service. earnings, and the inability to appeal licensing decisions. Both explicit and implicit regulations increase costs, rais- Surprisingly, the authors do not find limits on the number ing prices for consumers. of licenses to be a significant barrier. Discriminatory enforcement of a trade barrier or regu- Using detailed cross-border data, Shepherd and van latory policy may also limit trade. De facto discrimination der Marel (2013) investigate the role of the STRI barriers may also occur when the absence of a regulatory policy on services trade by sector. They find that restrictions in may affect a provider’s ability to deliver a service. In net- Mode 1 have a detrimental effect on total services (when all work industries such as telecommunications, for example, services are taken together) and on transportation services providers must be able to access the infrastructure to pro- specifically, but surprisingly not on other specific services vide a service. The absence of a legal requirement requiring sectors. In contrast, regulatory restrictions on Mode 3 have access and use of the infrastructure may affect the provi- a greater negative effect than Mode 1 restrictions on services sion of the service. trade flows for business, financial, and insurance services. The authors provide evidence of cross-modal substitution effects in total services trade but complementarity effects in Typology of Services Barriers business, financial, and insurance services. A wide range of regulations restrict trade in services. Table 4.4 provides a typology of barriers and restrictions National treatment that are subject for negotiations in trade agreements. National treatment measures are domestic regulations Agreements reflect the international efforts to classify and that discriminate, in a de jure or de facto manner, against eliminate restrictions that affect trade in services. The fol- foreign services and services providers. GATS Article XVII lowing sections discuss all three types of barriers and pro- requires that once foreign services providers enter the vide some empirical findings showing their impact on trade. domestic market, they should not face less favorable terms than domestic providers. The GATS does not contain an Market access exhaustive list of measures that may limit national treat- Module 4 Under the General Agreement on Trade in Services (GATS) ment. Measures that fall within national treatment restric- and other trade agreements, market access restrictions tions include prior residency requirements for the issuance include barriers that may or may not discriminate against of a license to supply a service; subsidies or tax rules that foreign firms. The GATS defines six types of market access discriminate against foreign providers; and restrictions limitations, which can affect all modes of supply: on the purchase, lease, or use of real estate. Market access and national treatment limitations are the main focus of • the number of service suppliers, monopolies, or exclu- services negotiations at the multilateral level. sive service suppliers Empirical studies explicitly measuring the role of • the total value of assets or service transactions national treatment are rare. Some studies analyze the 106 Valuing Services in Trade Table 4.4. Typology of Barriers to Trade in Services Type of effect Sources of information on Type of barrier (direct or indirect) Examples restrictions Restricted Direct: Restrictions (usually • Number of service suppliers, monopolies, • WTO/GATS (http://i-tip.wto.org market access quantitative) on foreign or exclusive service suppliers’ total value of /services/) entry into domestic assets or service transactions • World Bank Services Trade markets • Total number of service operations or volume Restrictiveness Index (STRI) of service output • Total number of natural persons that may be employed in a sector • Restrictions or requirements specific to the type of legal entity through which a service supplier can supply a service • Limits on foreign ownership (participation of foreign capital) • Restrictions on the movement of people National Direct: De jure • Discriminatory requirements • WTO/GATS (http://i-tip.wto.org treatment discrimination against • Discriminatory taxes or limiting of subsidies /services/) requirements foreign firms or de facto to nationals only • OECD (http://www.oecd discriminatory impact • Restrictions related to competition .org/economy/growth /indicatorsofproductmarket • Restrictions on property or land use regulationhomepage.htm) • Residency requirement for obtaining license to provide a service • Domestic preference on the allocation of bandwidth frequencies for transmission within the national territory Domestic Indirect: Nondiscriminatory • Qualification and licensing requirements and • WTO/GATS (http://i-tip.wto.org regulation regulations that do not (licensing or qualification) procedures that /services/) explicitly target foreign are lengthy, complex, and nontransparent, • World Bank Doing Business firms but have an impact which discourage foreign service providers • OECD Product Market Regulations on trade. These measures from seeking access to markets (http://www.oecd.org/economy can be categorized as • Technical standards /growth/indicatorsofproductmarket “entry” or “conduct” • Lack of objective and transparent regulationhomepage.htm) measures (measures that authorization criteria affect the operation of the firm). • Public service obligations for quality and social policy reasons • Rules governing conduct and operations, including limits on advertising and store opening hours, regulations on prices and fees, burdensome migration requirements, and regulations on the form of business and interprofessional cooperation • Restrictions on ownership, including barriers to vertical integration Note: GATS = General Agreement on Trade in Services, OECD = Organisation for Economic Co-operation and Development, WTO = World Trade Organization. discriminatory and nondiscriminatory effects of services of services plus tariff barriers are about three times the policies on the welfare of developing countries using gains from a preferential liberalization of goods and ser- computable general equilibrium models. vices with the European Union alone. They also find that Module 4 Jensen and Tarr (2012) estimate the gains to Armenia more than 85 percent of the gains from unilateral liberal- from creating a deep and comprehensive trade agreement ization come from liberalization of services rather than tar- with the European Union that also includes further liber- iff liberalization. alization of barriers in services. They estimate these barri- Using similar modeling techniques, Jensen, ers (a) using an index in which regulatory barriers impose Rutherford, and Tarr (2008) estimate that the welfare costs on both domestic and multinational firms in a non- gains for Tanzania of full reform of services are equal to discriminatory manner and (b) examining discriminatory 5.3 percent of Tanzanian consumption in the medium barriers against multinational services providers only. They run and 16 percent in the long run. The medium-term find that the gains from nondiscriminatory liberalization gains derive primarily from the removal of inefficient Module 4: Policy Options for Increasing Competitiveness and Trade in the Services Sector 107 Box 4.7. The World Bank’s Services Trade Restrictiveness Index (STRI) Database The STRI Database includes information from 103 countries, including 79 developing countries and 24 Organisation for Economic Co-operation and Development (OECD) countries. The data were collected during 2007–08. An update using the original method- ology and 2012 data did not reveal significant changes. The database focuses on five sectors: financial services (banking and insurance), telecommunications, retail distribution, trans- portation, and professional services (accounting and legal services only), with each sector disaggregated into subsectors as appli- cable. Within each subsector, the database covers the most relevant modes of supply. The measures affecting commercial presence are classified under the following broad categories: • requirements regarding the legal form of entry and restrictions on foreign equity • limits on licenses and discrimination in their allocation • transparency and accountability of licensing • restrictions on ongoing operations • relevant aspects of the regulatory environment. For some sectors, this information is supplemented with specific issues, such as regulation to ensure access to the market in tele- communications. For cross-border transactions, the focus is on conditions under which trade takes place. The temporary movement of people is covered only in professional services. Within each subsector and mode, policy regimes are rated using the scale shown in table B4.7.1. Table B4.7.1. Levels of Trade Openness in the World Bank’s Services Trade Restrictiveness Index (STRI) Database Level of openness Score Completely open 0 Virtually open but with minor restrictions 25 Major restrictions 50 Virtually closed, with limited opportunities to enter or operate 75 Completely closed 100 Source: http://iresearch.worldbank.org/servicestrade/. nondiscriminatory regulatory barriers against service pro- which regulators grant access to the domestic market can viders and of regulatory barriers against multinational be viewed as a hidden protectionist barrier. To prevent service providers. In a similar study of Kenya, Balistreri, domestic regulations from reducing trade in services, they Rutherford, and Tarr (2009) find even larger welfare gains. should be based on objective and transparent criteria, be These gains are explained largely by the liberalization of no more burdensome than necessary to ensure the qual- services policies. ity of the service, and not themselves restrict supply, in the case of licensing procedures (box 4.8).3 Domestic regulation In contrast to market access and national treatment, Restrictiveness of Services Barriers domestic regulatory barriers are not a primary concern of GATS negotiations. However, the GATS does call upon Triplett and Bosworth (2004) show that in the Module 4 World Trade Organization (WTO) members to develop United States, services are a critical source of labor any necessary disciplines to ensure that measures relating productivity and total factor productivity (TFP) growth. to qualification requirements and procedures, technical Inklaar, Timmer, and van Ark (2008) confirm this conclu- standards, and licensing requirements and procedures do sion across countries by showing that observed differences not constitute unnecessary barriers to trade in services. For in productivity growth and levels are explained mainly by example, excessively lengthy, complex, and nontransparent services, particularly business services, rather than goods licensing procedures may discourage foreign firms from or ICT investments. entering a market even if there are no explicit discrimina- Foreign competition in the form of trade in services tory barriers. Lack of objective and transparent criteria on can increase productivity in the wider economy. This 108 Valuing Services in Trade Box 4.8. Domestic Regulations and the World Trade Organization’s Necessity Test The General Agreement on Trade in Services (GATS) provides for a “necessity test” in Article VI:4 (domestic regulations) and in the Article XIV exceptions. Broadly speaking, a necessity test examines four issues: • Legal rulings relating to the objectives of the measure. The Appellate Body has consistently ruled that it is not the necessity of the policy objective that is to be examined but the necessity of the measure to achieve the intended policy objective. • The level of attainment a member is seeking to achieve with the measure. The Appellate Body has ruled that the choice of a measure can indicate the objective sought by it as well as the level of protection the measure intends to achieve. • The necessity of the measure. • The burden of proof. Necessity figures in three of the exceptions to General Agreement on Tariffs and Trade (GATT) Article XX. Panels have relied on a balancing approach in analyzing necessity. In the leading World Trade Organization (WTO) case on this issue, analyzing a discriminatory Korean regime for imported beef, the Appellate Body noted that claims of necessity must be evaluated in relation to the circumstances and that the evaluation involves a process of weighing and balancing a series of factors. These factors include the actual contribution made by the measure to achieving the stated objective within Article XX, the importance of the common interests or values protected, and the restrictive impact of the measure on trade. In this and other cases, the Appellate Body looked for a rela- tion between the measure and the end pursued that is not just a contribution to accomplishing the objective but is closer to being indispensable to accomplishing that objective. The party must demonstrate that its measures are necessary by providence evidence establishing that the measures contribute to the achievement of the objectives pursued. Evaluation of a measure’s necessity also requires evaluation of its restrictive effect on trade (or on the behind-the-border sale or distribution of imports, if the issue is behind- the-border discriminatory regulations). The less restrictive a nontariff measure is, the more likely it is to be deemed “necessary.” In the Korean beef (WTO 2000) and U.S. gambling (WTO 2005) cases, the Appellate Body clarified that as a panel evaluates necessity, it must examine whether the defending party could reasonably be expected to employ an alternative measure that is WTO consistent (or less WTO inconsistent) that would achieve the objectives pursued by the measure at issue. An alternative measure may not be “reasonably available” (that is, it may be relevant only in theory) or it may impose an undue burden on a member (examples include prohibitive costs or technical difficulties in implementation). Moreover, an alternative measure that is “reasonably available” must preserve the defending party’s right to achieve its desired level of protection with respect to the objec- tive pursued under Article XX. Where the complaining party identifies an alternative measure, the defending party has the burden of demonstrating that its GATT-inconsistent measure is “necessary.” To determine whether an alternative measure exists, the panel must evaluate whether the measure is economically and techni- cally feasible from an economic and technical point of view, whether the alternative would achieve the same objectives as the origi- nal measure, and whether it is less trade restrictive than the measure analyzed. If any of these conditions is not met, the alternative measure is deemed to be incompatible with WTO obligations. Sources: WTO 2000, 2003, 2005, 2011; Cadot, Maliszewska, and Sáez 2011. finding is reflected in trade studies analyzing the effect services for both domestic and foreign providers, the pres- of reducing domestic regulatory policies and trade barri- ence of foreign providers in the services sectors, and the ers in services on productivity or TFP, particularly in the level of privatization. Allowing foreign entry into services manufacturing sector. industries appears to be the key channel for higher TFP in Arnold, Mattoo, and Narciso (2008) analyze the rela- goods industries. tionship between communications, financial, and electric- Arnold and others (2010) assess the effect of services ity services on firm TFP in the manufacturing sector. They policy reforms in India on TFP in the manufacturing sec- show that countries with competitive and efficient telecom tor. They find that private sector participation, liberaliza- markets can provide new phone line connections in a few tion of decisions on operational activities, and the scope for days; in countries with inefficient public monopolies, it foreign participation increase downstream productivity. may take months to establish a new connection. According Barriers that are not explicitly addressed in trade Module 4 to the authors, efficient telecom services have had a posi- agreements may constitute barriers to trade in services. tive effect on the productivity of firms in Africa. Migration restrictions that include visa requirements and Other econometric country case studies corroborate other administrative necessities may form a barrier to evidence that reducing domestic regulatory barriers to for- trade. Interdiction of vertical integration of the production eign competition increases manufacturing TFP. Examining process may also act as a barrier for services firms. the reform process in the Czech Republic, Arnold, Javorcik, The OECD, the World Bank, and the Australian and Mattoo (2011) find a positive and statistically sig- Productivity Commission have each developed method- nificant relationship between firm performance in manu- ologies to assess the importance of a variety of regulatory facturing and policy measures such as overall reform in policies (boxes 4.9 and 4.10). This research has attempted Module 4: Policy Options for Increasing Competitiveness and Trade in the Services Sector 109 Box 4.9. The Australian Productivity Commission’s Trade Restrictiveness Index The Australian Productivity Commission has developed indexes of policy reform in services for economies in Europe, Asia, and North and South America. It defines two broad sets of regulations: establishment (or entry) barriers and barriers on operational activities. Depending on the sector, establishment barriers normally refer to the form of establishment (foreign partnership or joint ven- ture), nationality or citizenship requirements, residency or local presence, quotas or economic needs test, licensing, and regula- tions affecting the permanent movement of people. Operational barriers are activities reserved to certain professions or providers. They include limits on advertising and fees and restrictions on the temporary movement of people. For both types of barriers, the Australian Productivity Commission distinguishes between discriminatory and nondiscriminatory measures. A Services Trade Restrictiveness Index score is calculated for each economy. Scores are assigned to each restriction on the basis of a judgment about how stringent it is. The more stringent the restriction, the higher the score (scores range from 0 to 1). The restriction categories are then weighted based on a judgment about their relative economic cost. Separate indexes are calculated for domestic and foreign providers. The foreign index measures all restrictions, discriminatory and nondiscriminatory, that hinder foreign firms from entering and operating in an economy. The domestic index includes restric- tions applied to domestic firms. It generally covers only nondiscriminatory restrictions. The difference between the foreign and domestic scores is a measure of discrimination against foreigners. Box 4.10. Static and Dynamic Impact of Restrictions on Entry and Operations There are important differences between measures restricting market access and measures affecting operations. On a purely static basis, entry barriers tend to create rents for incumbents that might be partly shared with the state if, for example, licenses are sold rather than being given out. Removal of such restrictions increases consumer surplus and reallocates rents away from producers and the state. Measures that increase the cost of doing business “waste” economic resources, in the sense that goods and services must be allocated to paying for activities that produce no gains for the national economy. In a dynamic sense, it is not obvious that the gains from removing cost-creating measures outweigh the gains from removing market entry barriers. Market entry barriers are often associated with dynamic inefficiencies—a waste of economic resources—that are greatly reduced when markets are made more contestable. When a services market with a domestic monopolist is opened up to international competition by the elimination of an unduly strict licensing requirement, for example, there is not just an increase in consumer surplus, there is also a reduction in the cost of doing business because of the need to be more efficient in the face of competition from foreign service providers. Keeping the dynamic gains in mind suggests the need to sequence reform. In general, market entry barriers should be removed first, because they stimulate competition, at least in markets that are not freely contestable, as is the case in services markets in many developing countries. An additional reason for preferring market entry barriers as the starting place for reform is that the gains from reducing the costs of doing business are unlikely to be fully passed on to consumers in noncompetitive environments. Dealing effectively and efficiently with measures that restrict market entry is therefore often a necessary prerequisite to reducing the cost of doing business for firms already in the market and ensuring that the benefits are passed on to consumers. to go beyond the classifications employed in trade negotia- business and interprofessional cooperation. Van der Marel tions by including a wider range of limitations. These stud- (2012) finds that conduct regulation within each services ies have tried not to assess the letter of the law but rather sector is a more robust factor in explaining TFP growth whether access is provided in practice. than entry barriers. Little quantitative analysis has been conducted on whether wider policy barriers in services have an effect Quantifying Regulatory Barriers on performance. Van der Marel (2012) finds that services Module 4 trade—particularly inward FDI and services imports—has The aim of a quantitative assessment is to calculate the a significant effect on TFP in services. These effects are benefits and costs of maintaining or implementing services inhibited by domestic and FDI regulation. Regulation is regulations. It is conducted for policy makers in order to the main factor explaining cross-country differences in choose the most efficient regulatory mix. TFP growth over time while domestic and FDI-conduct Molinuevo and Sáez (2014) review the literature on ser- restrictions are the most important factors explaining vices trade. They describe some methods used to assess the TFP growth in services. These regulatory barriers include impact of regulatory frameworks on services, particularly limits on advertising, regulations on store hours, regu- the impact of regulations on the provision, price, and cost lations on prices and fees, and regulation of the form of of services. 110 Valuing Services in Trade Quantitative methods can be classified according to Formally, the tariff equivalent of importing country j, tj, three not mutually exclusive taxonomies: is calculated as • whether they employ direct or indirect measures 1/ −σ • whether they focus on sectoral or economywide impacts ⎛ ⎛ ⎜ ln(1 + τ j ) = ⎜ ln ⎜ ∑ i≠ j M actual j ⎞ ⎛ ⎟ − ln ⎜ ∑ i≠ j actual M benchmark ⎞⎞ ⎟⎟ • whether they use a retrospective or prospective approach. ⎜ ⎝ ⎝ ⎜ ⎜ ∑ i≠ j M jpredicted ⎟ ⎟ ⎠ ⎜ ⎜ ⎝ ∑ i≠ j predicted ⎟ ⎟ M benchmark ⎟ ⎠⎟⎠ Direct methods gather information on restrictions and policy, to include in econometric analyses in order to assess the impact on outcomes of interest. Researchers lacking where ∑ i≠ j M is the sum of imports over all trading j direct information on restrictions use indirect methods. partners, M and M jpredicted are actual and predicted actual j They estimate or infer the level of openness, restrictive- imports from the gravity model of trade, M benchmark ness, or contestability of a market by comparing countries is imports by the free-trade benchmark country (the against a benchmark (a country with no or few restric- country with the lowest ratio of actual to predicted tions). The main limitations of indirect methods are that imports), and s is the elasticity of substitution. The same the resulting estimated restrictions may well be capturing gravity model specification is estimated in table 3.11, more than trade barriers and it usually proves difficult to except that the STRI variables are excluded, in order to link the estimates to a specific policy measure, making this be consistent with the literature in allowing the residuals approach less suitable than direct methods for guiding pol- to fully capture the barriers to services trade. For robust- icy decisions. ness, two separate values of the elasticity of substitution Because a wide range of regulations affects the provision are assumed, s = 1.95 and s = 5.6, adopted from the of services, comparing the contestability of markets across literature.4 countries can be difficult. Studies normally deal with this This method is used to estimate the tariff equivalent problem by constructing indexes that try to summarize in of services barriers in selected countries (table 4.5). In one indicator a set of policies affecting the provision of a general, countries with more restrictive services regula- services category in a country. Once the methodology is tory environments are much less likely to export services. applied to several countries, the market contestability of Undertrading in services may thus suggest untapped a category can be compared across economies (see box potential to increase exports through the removal of 4.7 and Borchet, Gootiiz, and Matoo 2012a for an in- trade-related obstacles. depth discussion of the World Bank’s restrictiveness index Barriers to trade not only dampen services export methodology). potential, they also act as barriers to services imports. Gravity models can be used to quantify country- Table 4.5 presents the average tariff equivalents for specific barriers. Various methodologies can be used to Belarus, the Russian Federation, and each Eurpoean quantify the level of regulatory protection in domestic Union (EU) country in the sample over the period markets for services of an importing country, includ- 2009–11 for the two values of the elasticity of substitu- ing a quantity-based approach that measures barriers tion. Although the magnitudes of the tariff equivalents by estimating tariff equivalents. The methodology most are quite sensitive to the elasticity of substitution, the commonly applied in the literature, as proposed by Park country rankings are preserved. (2002), computes the average protection applied by Belarus shows high levels of regulatory protection each importer from the residuals of an estimated gravity of its domestic services. Its estimated trade barriers in model of trade for services. The approach compares services—an estimated tariff equivalent of 65 percent actual levels of trade flows with potential levels of trade, assuming the lower elasticity value and 120 percent assum- Module 4 given the physical and economic characteristics of coun- ing the higher value—are higher than all EU countries tries and their trading partners. As the residuals may be and Russia. Belgium’s average protection of 63 percent capturing factors other than trade barriers, actual and (118 percent at the higher elasticity) is the highest among predicted trade are normalized relative to a theoretical the EU countries; Ireland’s rate of 48 percent (106 per- situation considered to be the free-trade benchmark. cent) is the lowest. The average EU country’s tariff equiva- This benchmark is the country in the sample with the lent is 55 percent (112 percent). Russia’s tariff equivalent highest level of actual imports relative to predicted of 54 percent (111 percent) is just below the average for imports. the European Union. Module 4: Policy Options for Increasing Competitiveness and Trade in the Services Sector 111 Table 4.5. Tariff Equivalents of Barriers to Services Trade in are difficult to change. Governments can reduce the costs Selected Countries, 2009–11 of other measures through good regulatory policies on Tariff equivalent (percent) the domestic economy and institutions. Horizontal Importer s = 1.95 s = 5.6 measures—measures affecting a wide range of services Austria 57.7 113.9 sectors—complement the specific trade and domestic Belarus 65.2 119.6 policies discussed above. These factors are referred to as Belgium 62.9 117.9 Bosnia and Herzegovina 60.6 116.2 domestic enabling factors. Bulgaria 53.8 110.8 Domestic enabling factors (or fundamentals) affect Cyprus 49.4 107.1 trade in services. None of these factors can be changed in Czech Republic 56.8 113.2 the short run. They reflect a country’s endowments of four Denmark 51.4 108.7 factors: Estonia 51.5 108.8 Finland 53.0 110.1 France 60.7 116.3 • human capital, including skills and entrepreneurial Germany 59.4 115.3 ability Greece 54.2 111.1 • natural resources, such as sights that attract tourists Hungary 53.3 110.4 • infrastructure, such as a telecommunications network Ireland 48.3 106.1 that facilitates the delivery of services Italy 56.9 113.3 Latvia 54.7 111.5 • institutions, such as the quality of a country’s rule of law Lithuania 54.1 111.0 or regulatory environment. Luxembourg 51.9 109.2 Malta 47.7 105.6 Table 4.6 lists some of the domestic enabling factors Netherlands 57.8 114.0 governments can target to reduce trade costs. The gains Poland 56.5 112.9 from good regulatory practices can be increased by putting Portugal 54.4 111.2 Romania 56.3 112.8 in place the right institutional and economic environment Russian Federation 54.3 111.2 for firms. These domestic enabling factors are as impor- Slovak Republic 58.3 114.4 tant for goods as they are for services. However, particular Slovenia 56.3 112.9 features of services may make them more sensitive to these Spain 54.0 111.0 enabling factors. For example, good domestic institutions Sweden 53.3 110.4 United Kingdom 56.4 112.9 in the form of a strong rule of law, in particular regarding contract enforcement, affect both services and goods firms. Sources: World Development Indicators, Trade in Services Database, and Centre d’Etudes Prospectives et d’Informations Internationales (CEPII). However, some goods are less sensitive to contractual Note: s = is the elasticity of substitution. For robustness, two values, enforcement, simply because there are fewer information adopted from the literature, were used. asymmetries regarding their inherent characteristics and features. In contrast, many services are differentiated to a These estimations are based on cross-border trade. very high extent to meet clients’ needs. For these services, They do not capture services trade flows from FDI or the there are more information asymmetries, and the many movement of people. These flows remain an important tailored inputs and features are harder to write down in a channel through which foreign providers supply services. contract. If restrictions on services affect commercial presence more Some enabling factors are easier to implement than than cross-border trade, the tariff equivalent will be under- others. It takes years to strengthen the rule of law in a reported. This analysis does not reveal which sectors are country where corruption is deeply rooted or to improve more protected or what the underlying causes of these high education enough to create a large pool of skilled workers. Module 4 levels of protection are. Other policies, such as tax breaks for inward FDI, have more immediate effects. It may be easier for policy makers to improve enabling Domestic Factors that Enable Trade in factors than to adopt regulatory policies that have a direct Services effect on firms. For instance, government policies encour- Miroudot, Sauvage, and Shepherd (2010) ask whether aging entrepreneurial skills or increasing the number of and how trade costs can be reduced. They find that some university graduates are likely to encounter less opposition trade barriers, such as geography or cultural barriers, than reforms aimed at opening up markets. Governments Module 4 112 Table 4.6. Domestic Factors that Enable Trade in Services Policy area Objective Indicator Relation to services Labor skills Increase capacity to produce sophisticated • Share of population with tertiary education • On average, services are more skill intensive than services exports; climb up ladder of • Tertiary enrollment goods. comparative advantage in services (or from • Average level of schooling • Level of skills of both exporter and importer is goods to services), especially in professional, related to level of two-way trade in services, computer and computer-related, and business • Average years of schooling which takes place mostly between OECD services. • Share of population with technical education countries. • Number of IT graduates Management Enhance adoption and use of modern • Number of management graduates • Highly differentiated nature of services makes and technologies that are essential for producing a • Average level of management diploma them particularly sensitive to management entrepreneurial service or good. competence and organizational structure. • Ranking of management school of employed managers skills • Services firms are generally smaller than • Type of management practice (Bloom and van Reenen 2010) manufacturers, which makes it easier for • Number of entrepreneurs entrepreneurs to start up an enterprise. Trade-related Reduce costs related to delivery of services • Density of telecommunications • Cross-border trade (Mode 1) can be conducted infrastructure (transportation, telecommunications, export, • Quality of telecommunications infrastructure over the Internet. Internet access reduces search transaction, and search costs). costs and can reduce asymmetry of information. • Number of computers per 1,000 people • Telecommunications services are an essential • Number of Internet users per 1,000 people input for other services. Institutions Establish governance arrangements to foster • Quality of rule of law (World Bank Governance Indicators) • Services are complex, highly differentiated, and relationships between private parties rather • Quality of legal system and security of property rights relationship specific, requiring strong contract- than those between private parties and the (Gwartney, Lawson, and Hall 2012) enforcing institutions. government (Acemoglu and Johnson 2005). Governance Increase ability of governments to formulate • Quality of regulation (World Bank Governance Indicators)/ • Services require strong independent regulators and implement sound policies and regulations independence of regulator that can implement good policies that open that allow and promote private sector • Government efficiency (World Bank Governance Indicators) markets for services. development. • Voice and accountability (World Bank Governance Indicators) • Well-governed regulatory bodies are more likely to resolve market failures in services. • Level of corruption (World Bank Governance Indicators) Business Attract FDI and multinational corporations in • Tax breaks to attract multinational corporations • Complementarity between outward FDI and environment order to benefit from export expansion and • Mix of policies to attract FDI (including reducing tariffs on services exports. increased domestic competitiveness. inputs for downstream users) • Inward FDI is positively correlated with exports in the goods sector. Note: FDI = foreign direct investment, IT = information technology, OECD = Organisation for Economic Co-operation and Development. Module 4: Policy Options for Increasing Competitiveness and Trade in the Services Sector 113 need to have the means to address the policies that stimu- among firms, which is harder to quantify in contracts and late services activities (box 4.11). therefore requires strong institutions. Empirical research finds a robust and significant relationship between institutions and the competitive- Institutions ness of services. Amin and Mattoo (2006) find that Institutions matter a great deal for economic activity. They countries with better institutions have larger and more relate to the optimal governance of relationships between dynamic services sectors. They suggest that because of private parties rather than between private parties and the the complex web of transactions involved in the produc- government, as Acemoglu and Johnson (2005) show. They tion of services, regulatory and contract-enforcing insti- are essential for free trade. tutions (such as the rule of law) play a key role in their Differences in institutions are a critical factor determin- development. They hypothesize that services are more ing comparative advantage. Levchenko (2007) shows that relationship-specific than goods because they involve sectors with complex input use are more dependent on investments in customization, which increase the costs strong rule of law than other sectors. Costinot (2009) shows of switching from one supplier or customer to another. that more complex sectors that require a longer training Services therefore require stronger contract-enforcing period for employees (fixed training costs) need stron- institutions than goods. For example, a firm’s willing- ger domestic institutions for optimal production. Nunn ness to outsource the transcription of confidential client (2007) points to the character of business relationships information is much greater if there is confidence in the Box 4.11. Challenges Facing Potential Services Exporters in Kenya Lack of knowledge about exporting opportunities, markets, and processes and a lack of awareness as how to acquire such knowl- edge are widespread in Kenya. According to a 2009 survey, 48 percent of Kenyan services exporters do not have a plan for export- ing, and 54 percent do not conduct any market research before exporting. These findings suggest that, although Kenyan services firms may be innovative and successful domestically, many do not engage in any systematic attempt to export their services. If nearly half of Kenyan exporters do not have an export business plan, there are likely many other firms that are able to export but do not. This lack of knowledge is an important constraint that must be overcome. Many services providers in Kenya, especially small ones, lack international networks and find it very difficult to obtain market intelligence on foreign markets. Lack of contacts, knowledge of foreign markets, tax implications, and other market intelligence prevents them from taking advantage of potential export opportunities. Services firms in Kenya are also unaware of the support services the government provides, the institutions that support trade, and the services they offer. For example, surveyed exporters were unclear about the activities and role of the Kenya Chamber of Commerce. Another important constraint is the difficulty of penetrating foreign markets. The problem is caused partly by the domestic regulatory restrictions on Kenyan firms in local and foreign markets. But global perceptions of Kenya also act as a disincentive for potential clients to work with Kenyan exporters. Current and potential services exporters face regulatory barriers that affect their operations and export opportunities. Surveyed firms cite taxation-related restrictions, burdensome procedures, and licensing requirements to operate in Kenya; outdated sector- specific regulatory measures; and sectoral restrictions such as advertising prohibitions in accounting, architectural, engineering, and legal services. Burdensome requirements for academic and professional qualifications and licensing requirements restrict Kenyan firms from entering foreign markets. Kenya has low international brand equity as a business service provider. In contrast to its counterparts in India or South Africa, the government does not sponsor international conventions or events to showcase the business process outsourcing (BPO) services available in Kenya. The foreign perception of Kenya’s government as unstable also deters foreign companies from using Kenyan BPO. However, Kenya does have high visibility as a business service provider in the East African Community region. Skills mismatches and skills shortages pose a significant challenge to many Kenyan exporters. Kenya is relatively well endowed with graduates who could work in various business services firms, including in the BPO sector. Recent graduates need to receive Module 4 substantial training to catch up with international standards, however. Such training costs represent a substantial part of BPO con- tact center costs. Kenyan institutions produce good programmers, but information technology (IT) firms lack exposure to foreign markets, promotional expertise, and skills in exporting their services. Kenya faces an acute shortage of engineers, particularly mechanical engineers. In the insurance sector, technical skills are weak. It is so difficult to find individuals with the necessary skills that firms devote substantial resources to in-house training and sponsor- ing staff to attend courses. Kenya has many lawyers, but legal training grasp is very theoretical, and few lawyers are qualified to export their services. Another frequently cited constraint is the lack of information and communication technology (ICT) infrastructure. It is being overcome by two new undersea fiber optic cables, which have given Kenya access to high-speed Internet. Source: Dihel and others 2012. 114 Valuing Services in Trade privacy and data protection laws in the country in which trade within a country or economic union. They con- the services are provided. trast the determinants of bilateral services trade within Figure 4.8 shows the results of a cross-country regres- Canadian provinces and within EU member states. sion of the relationship between the complexity of services Economic institutions and regulations should have no and the rule of law. The horizontal axis shows the rule of law effect on trade within a country (intraprovincial trade in index, taken from the World Bank’s Governance Indicators, Canada) but should affect trade across countries (within which the empirical trade literature often uses to capture the the European Union). Their measure for economic strength of domestic institutions. The vertical axis shows the institutions is the overall product market regulations indi- share of trade in “complex” services sectors (professional ser- cator for services sectors. The analysis reveals that exports vices such as accounting and legal services, finance, and insur- are very sensitive to economic institutions in the export- ance; see also table 3.9). This measure is based on Costinot’s ing country and to regulations in the importing country. (2009) calculations. The relationship between the two indi- Tentative estimates suggest that intra-EU services trade cators is positive, albeit with extreme outliers. Countries could be much greater if the internal market functioned such as the United States, the United Kingdom, and Ireland like the Canadian market. have large export shares of finance and professional services partly because of historical reasons and similar legal systems. Governance Lebanon and Mexico have large export sectors in complex services given the level of their domestic institutions. Governance structures are of major importance for Using a different approach, Lejour and Verheijden economic activity. Undertaking reforms in domestic gov- (2004) show that differences in institutions can explain ernance signals to investors and exporters that the business Figure 4.8. Complexity of Services versus the Rule of Law in Selected Countries, 2010 0.25 Complex services exports as a percent of total services exports GBR 0.20 USA IRL 0.15 LBN 0.10 MEX BRA DEU IND 0.05 MKD RUS CHL IRN JPN VEN PHLMDA EGY 0 MYS Module 4 1 2 3 4 5 Rule of law Fitted values High-income country Upper-middle-income country Low-income country Lower-middle-income country Sources: World Bank Governance Indicators and Trade in Services Database. Note: The rule of law captures perceptions of the extent to which agents have confidence in and abide by the rules of society. It covers contract enforcement, property rights, the police, the courts, and the likelihood of crime. It is measured on a 1–5 scale, where 5 represents the strongest rule of law. BRA = Brazil, CHL = Chile, DEU = Germany, EGY = Arab Republic of Egypt, GBR = United Kingdom, IND = India, IRL = Ireland, IRN = Islamic Republic of Iran, JPN = Japan, LBN = Lebanon, MDA = Moldova, MEX = Mexico, MKD = Macedonia FYR, MYS = Malaysia, PHL = Philippines, RUS = Russian Federation, USA = United States, VEN = Venezuela, RB. Module 4: Policy Options for Increasing Competitiveness and Trade in the Services Sector 115 climate is competitive, reduces the regulatory burden on Where private goods markets are functioning, competi- businesses, and increases transparency. tion serves as a substitute for regulation. With good gover- Governance can be measured in various ways. The nance structures in place, government regulators take on World Bank Governance Indicators include the level of the function of enablers of competitiveness and economic corruption, regulatory quality, voice and accountability, growth (Reis and Farole 2012). and other measures (Kaufman, Kraay, and Mastruzzi 2009). To some extent, competition plays this role for services, Reform of these policies can best be achieved through too. Figure 4.9 shows a negative relationship between the simplification and gradualism. Simplification means that governance index of regulatory quality and the STRI for the process of governance should avoid complexity and both services restrictions overall and financial services be efficient. Simplifying new policies, regulators have to through Mode 3. Governments with good regulatory poli- ensure that the potential for any reintroduction of cum- cies facilitate well-functioning and competitive services bersome or ineffective changes is minimized. A gradual markets. reform process is more realistic than radical reforms. This relationship suggests that addressing trade restric- Partial reform can set off positive feedback experience, tiveness is a necessary but not sufficient condition for which encourages policy makers to reform other elements increasing competitiveness: domestic institutions play of the governance structure. a complementary role in creating competitive markets. Figure 4.9. Services Trade Barriers versus Regulatory Quality in Selected Countries, 2008–10 a. All modes of trade b. Finance through Mode 3 5 5 IRL DEU DEU CHL USA IRL CHL 4 4 USA JPN JPN MUS MUS BHR BHR Regulatory quality Regulatory quality MYS QAT MYS QAT CRI OMN CRI OMN PAN PAN 3 3 BRA MEX SAU MEX BRA SAU GHA LBN GHA LBN UGA EGY UGA EGY CHN PHL CHN PHL PRY MOZ RUS IND MOZ PRY IND KHM KHM RUS PAK LSO LSO PAK 2 CMR 2 CMR BOL ETH BOL ETH ECU BDI DZA ECU DZA BDI UZB UZB VEN ZAR IRN VEN ZAR IRN 1 1 ZWE ZWE Module 4 0 20 40 60 80 0 20 40 60 80 100 Services Trade Restrictiveness Index (STRI) Services Trade Restrictiveness Index (STRI) Fitted values High-income country Upper-middle-income country Low-income country Lower-middle income country Sources: World Bank Governance Indicators Database and Services Trade Restrictiveness Index. Note: Regulatory quality reflects perceptions of the ability of the government to formulate and implement sound policies and regulations that permit and promote private sector development. It is measured on a 1–5 scale, where 5 represents the highest value. The STRI is measured on a scale of 0–100, where 100 represents the highest value. For the International Organization for Standardization (ISO) abbreviations, see appendix B. 116 Valuing Services in Trade Van der Marel (2011) finds that countries with more its technical understanding of the regulatory field and sophisticated governance frameworks are better able to services sector export services sectors in which behind-the-border barri- • the financial resources allocated to regulation of the ers are limited. sector. In the same vein, Molinuevo and Sáez (2014) propose that when assessing the regulatory governance frame- The Business Environment work for services, policy makers should examine both institutions and their capacity to conduct their regula- Developing countries compete for FDI by offering tax tory functions. Services markets often suffer from mar- incentives to large multinationals or by reducing barriers ket failures that require regulations. Regulators need to to foreign investment (box 4.12). They do so because of walk a fine line between the need to achieve social objec- the widely held view that multinational corporations are tives and the need to minimize the burden of regulation a catalyst for export expansion and improve the competi- on market participants. Molinuevo and Sáez (2014) note tiveness of the host country. The business environment is that a poor regulatory environment may arise from a poor also very much influenced by governance, institutions, and regulatory setting. In this situation, the government bod- overall regulatory policymaking. ies and agencies charged with regulating services markets FDI promotes exports of a host country by augment- lack a mandate to regulate. They are unable to resist pres- ing domestic capital for exports, transferring technology sures from other government bodies or private stakehold- for new products for exports, facilitating access to new ers. Regulatory agencies must have adequate resources to and large foreign markets, and training and upgrading understand and evaluate fully the complexity of the market the technical and management skills of the local work- and the impact of regulatory policy. force. Empirical evidence suggests that FDI promotes An assessment of regulatory governance and its likely trade in the host country by introducing new technology impact on services trade must examine at least three aspects and management techniques, diffusing market informa- of the institutional setting: tion and increasing market access, and stimulating com- petition in the host country (Blomström and Kokko 1998; • the mandate and independence of the regulatory agency Borzenstein, De Gregorio, and Lee 1998; UNCTAD 1999). in the broader context of the general institutional As a result of FDI, East Asian economies witnessed tremen- framework of the government dous growth before the crisis of 1997–98. • the institutional capacity to fulfill that mandate, which Another area of policy making related to the business is determined mainly by the regulator’s staff, especially environment is the promotion of trade and investment. Box 4.12. Has the Philippine Economic Zone Authority (PEZA) Boosted Services Exports? In 2004, the Philippine government formally recognized service exports as one of its top priorities for employment generation and foreign exchange earnings. An executive order issued that year resulted in the formation of a public-private partnership task force. The private sector’s proactive involvement seems to be key to the Philippines’ successful export performance. As a result, the business process outsourcing (BPO) sector has performed remarkably well. The Philippines has the world’s third- largest market share of BPO. The rapid growth of the sector partly reflects investment incentives provided by the government, which allows 100 percent foreign ownership in a BPO firm. Since 2000, ICT has been one of the Investment Priority Plan sectors. BPO firms registered with the Philippine Economic Zone Authority (PEZA) and located in designated information technology (IT) parks and IT buildings have been eligible for PEZA incentives. PEZA has created a more business-conducive environment. For companies registering under PEZA, it offers one-stop-shop services for business registration and provides an exemption from local government business permits, licenses, and fees. PEZA issues permits related to building and occupancy, import and export, and environment clearance. Such services reduce business start-up Module 4 time and costs. In an effort to reduce graft and corruption, PEZA rotates staff members on a regular cycle. Because data are limited, one cannot conclusively determine whether firms receiving PEZA incentives have generated more BPO exports than non–PEZA firms. PEZA reports that exports by PEZA IT-BPO firms amounted to $4 billion in 2009. This figure seems low, given that planned BPO investment in PEZA areas from 2003 to 2009 was about 85 percent of total planned BPO investment, according to the Board of Investment. A possible underestimation may have occurred because of a discrepancy in the classification of IT-BPO firms by PEZA and the Business Processing Association of the Philippines. Additional research needs to be conducted to determine how effective PEZA was in boosting BPO exports. The Board of Investment has been building a statistical database on the performance of the BPO industry. Completion of this work would shed some light on this subject. Source: Yi 2012. Module 4: Policy Options for Increasing Competitiveness and Trade in the Services Sector 117 Export and investment promotion agencies are responsi- proportion of services trade is facilitated through foreign ble for institutional reform, capacity building, and private affiliates. sector participation. These agencies need to develop a tar- Figure 4.10 illustrates the complementarity between geted export promotion or investor strategy to optimally Mode 1 (and 2) services trade and Mode 3 trade barriers. profit from the interaction between modes, especially The negative relationship between the STRI and trade in where complementarities between Mode 1 and Mode 3 services is evident in both the finance and telecommunica- trade exist. Domestic firms can be linked to and therefore tions sectors (it is somewhat stronger for telecommunica- profit from attracting foreign FDI. Costa Rica attracted tions). Laxer restrictions on Mode 3 services are associated foreign FDI, in particular from Intel, in the 1990s that with greater cross-border trade through Mode 1 (and 2), dramatically changed its export and economic structure. an indication that at least in these sectors, the two modes Special agencies developed a program to enhance link- reinforce each other. ages between local small and medium-size enterprises and More recent studies using firm-level services trade data foreign investors (see Reis and Farole 2012). examine the determinants for choosing a mode of supply in Grünfeld and Moxnes (2003) try to identify the deter- a deeper manner. Kelle and others (2013) show that firms minants of services trade through cross-border (Mode 1) choose foreign affiliates based on their productivity and and foreign affiliates (Mode 3). Their study, which focuses that more productive firms are more likely to trade through on OECD countries, reveals strong links between FDI and Mode 3. Other important factors include wages, which are cross-border trade in services. It also finds that outward negatively correlated with Mode 3 trade, and costs related to FDI and services exports are complements. Lennon (2009) trade through Mode 1. Higher organizational costs for cross- also finds complementarity between Mode 1 and Mode 3 border trade make firms more likely to set up foreign affili- trade. These findings are policy relevant because a large ates. It would be useful for investment promotion agencies Figure 4.10. Mode 3 Barriers versus Modes 1 and 2 Trade, 2008–10 a. Telecommunications sector b. Financial services sector 0.5 0.5 IRL IRL 0.4 0.4 LBN LBN Sector exports/GDP Sector exports/GDP 0.3 0.3 PAN PAN 0.2 0.2 KHM KHM MYS MYS EGY 0.1 EGY PRY 0.1 DEU MOZUGA PHL IND DEU PRY IND CHL UGA PHL MOZ LSO CHL JPN BOL USAOMN PAK RUS USA PAK LSO BRA SAU QAT BRA BOL JPN RUS OMN 0 MEX ECU MUS VEN CRI GHA CMR CHN BHR DZA ETH Module 4 SAU QAT ECU DZA VEN 0 CMR MUS GHA MEXCRI BHR ETH 0 20 40 60 80 100 0 20 40 60 80 100 Services Trade Restrictiveness Services Trade Restrictiveness Index (STRI) for Mode 3 Index (STRI) for Mode 3 Fitted values High-income country Upper-middle-income country Low-income country Lower-middle-income country Sources: World Bank Services Trade Restrictiveness Index, World Development Indicators, and Trade in Services Database. Note: For the International Organization for Standardization (ISO) abbreviations, see appendix B. 118 Valuing Services in Trade to identify services sectors that engage in many small trans- IT and information technology–enabled services (ITES) actions rather than a few larger ones, so that policies can be sectors. Institutional mechanisms for aligning skills develop- tailored to specific activities. ment with the needs and requirements of the industry are critical (box 4.13). Figure 4.11 illustrates the relationship between higher Labor Skills education and services exports. The indicator of education High-skilled labor is strongly associated with economic devel- measures not merely the number of people with degrees; it opment (Hall and Jones 1999). This relation is documented also takes into account the qualitative aspects of education in various studies analyzing trade patterns: skill-intensive and includes information on both the current and future industries are found in skill-abundant countries that stimu- workforce. The figure shows a correlation between the two late productivity; such industries are correlated with higher variables, indicating that countries that are better endowed economic growth (Findlay and Kierzkowski 1983). with high-skilled labor and qualitatively better education Economists debate the causal direction of the relation- export more services (box 4.14). ship between trade and labor skills. Depending on the type Shingal (2010) documents the correlation between of labor (skilled, semi-skilled, low-skilled), opening up skills and services exports. Using gross tertiary school domestic markets can also have distributional consequences enrollment as a proxy for human capital, mainly in for workers. Many services sectors are significantly more OECD countries, he shows that human capital is critical skill intensive than many goods sectors (Nusbaumer 1987; for aggregate services exports. In a study of mostly OECD Gibbs 1986; Jensen 2008; van der Marel 2011). Economists economies, Lennon (2009) finds that years of schooling, believe that skilled labor can help increase the sophistication secondary school enrollment, and high school educational of exports. In the long run, a skilled labor base will allow attainment in both the importing and exporting country countries to climb the ladder of comparative advantage. affect services trade. After access to high-bandwidth telecommunica- Through a theoretical model of economic develop- tions infrastructure, the availability of skilled labor is the ment, Lennon, Mirza, and Nicoletti (2009) hypothesize most important determinant of the long-term growth of that services trade is characterized by tasks that need Box 4.13. What Are Developing Countries Doing to Develop Skilled Labor in High-Tech Services Sectors? Governments, in collaboration with the private sector, have undertaken a number of initiatives to address skills mismatch. In 2008, the government of Mexico established MexicoFIRST as a partnership between the Asociación Mexicana de la Industria de Tecnologías de Información (AMITI) and the Asociación Nacional de Instituciones de Educación en Tecnologías de la Información (ANIEI). ProSoft, a government agency tasked with promoting information technology (IT) and information technology–enabled services (ITES) industries, facilitated and supported its creation. MexicoFIRST interfaces with industry and Mexican universities to facilitate training programs at the universities that meet industry needs. In India, the National Association of Software and Services Companies (NASSCOM) assessment of competency (NAC) framework was developed in consultation with ITES companies. It has emerged as India’s national standard for generic skills and recruitment of entry-level talent for the ITES industry. NASSCOM rolled out the framework in partnership with a number of state governments in India. Assessment scores indicate areas for improvement, allowing customization of training. NASSCOM has also developed a NAC-Tech certification, which benchmarks engineering skills for the IT industry. It, too, is being rolled out in partnership with state governments. Applying and enforcing common industry certification not only helps align skills with industry requirements, it also provides IT and ITES companies with an estimate of the talent pool available and reduces their recruitment costs. Several countries are providing training grants for this purpose. In November 2007, the president of the Philippines directed the Technical Education and Skills Development Authority (TESDA) to allocate = P350 million (about $8 million) to provide scholarships for training 70,000 call center agents. Singapore has a national Skills Development Fund for upgrading worker skills. It launched the Initiatives in New Technology scheme to establish new capabilities within companies or industries by encouraging manpower development in the application of new technologies, industrial research and development, and know-how. South Africa offers up to R 12,000 (about $1,700) per worker for company-specific training and skills acquisition. Under its ICT Capacity Building Program, Module 4 Sri Lanka funds a portion of the training costs of IT and ITES companies. It also offers grants of up to $10,000 to bring in a special- ized trainer from abroad under a “train the trainer” program. To address the significant shortage of skills, many large IT and ITES companies have launched skills development initiatives, building training centers, employing hundreds of training staff, and collaborating with academic institutions. Infosys’s new Global Education Center in Mysore, India, for example, has more than 300 full-time faculty and is able to train 13,500 employees at a time. The company invested more than $120 million in this 335-acre, 2-million-square-foot facility. Some governments and universities have used public funding and public-private partnerships to nurture and expand the talent pool. These initiatives have been designed to expand existing university infrastructure and faculty, develop competencies that are benchmarked globally, and forge linkages for skills development with private sector and best-in-class institutions. Source: Sudan and others 2010. Module 4: Policy Options for Increasing Competitiveness and Trade in the Services Sector 119 Figure 4.11. Human Capital versus Services Exports in Selected Countries, 2010 0.5 IRL 0.4 LBN Trade in services/GDP 0.3 PAN 0.2 KHM BGR MYS MDA 0.1 EGY MKD PRY IND DEU MOZ UGA PHL LSO BTN CHL PAK OMN RUS ARE JPN USA BRA SAU 0 GIN GHA MUS ETH MEX CHN CRI −2 −1 0 1 2 Human capital Fitted values High-income country Upper-middle-income country Low-income country Lower-middle-income country Sources: World Development Indicators and World Economic Forum (WEF 2010). Note: The horizontal axis shows standard deviations from the mean on the human capital scale of the World Economic Forum 2010. For the International Organization for Standardization (ISO) abbreviations, see appendix B. to be produced together in order for any of them to be example, is attributed to entrepreneurs of Indian origin of high value. They emphasize the quality of task per- who received world-class management training or worked formance in the home and host countries for the suc- abroad in well-established multinational firms (Gregory, cessful delivery of services. In contrast, for goods, only Nollen, and Tenev 2009). Lall (1999) illustrates the criti- the exporting country’s characteristics seem to affect cal role managers played in changing traditional human exports. This model is confirmed by the authors’ empiri- resource attitudes and policies, allowing these economies cal results, which show that human capital in both the to develop and export the products that were in demand in importing and exporting countries are important for the advanced economies. services trade. These results suggest that developing Few data are available on the relationship between countries would do well to focus their services exports services exports and management practices. Figure 4.12 on higher-income countries. shows a positive association between overall management and exports of commercial services. The overall manage- ment measure, taken from Bloom and van Reenen (2010), Management and Entrepreneurial Skills Module 4 reflects practices such as monitoring management, target Micro-level studies in developing countries suggest that management, and incentive management based on firm- managerial skills play a key role in the adoption of mod- level surveys. ern technologies, which are critical for producing high- Results on the link between management practices quality goods and services for export. Saxenian (1999, and productivity in services are mixed. Wolf (1999) 2000, 2004) and others suggest that transnational entre- finds that the introduction of ICT or “computerization” preneurial networks have played an essential role in the decreased TFP in the services sector. O’Mahony and van development of exports from several developing coun- Ark (2005) provide evidence that productivity in the tries. Development of the Indian software industry, for retail sector in the United Kingdom increased thanks 120 Valuing Services in Trade Box 4.14. Assessing the Knowledge Economy The World Bank created the Knowledge Assessment Methodology (KAM) to assess a country’s preparedness to compete in the knowledge economy. It measures 148 structural and qualitative variables of 146 countries, including most OECD and more than 90 developing countries. To allow for cross-country comparisons, KAM provides both absolute and normalized (relative to other countries in the comparison group) values. The normalization procedure used is as follows: • Actual data (u) for all variables and countries are collected from World Bank data sets and the literature. • Countries are ranked based on the absolute values. Countries that perform at the same level are assigned the same rank. • The number of countries with a higher rank (Nh) is calculated for each country. • The formula 10 * (1 – Nh/Nc) is used to normalize the scores for every country on every variable according to their ranking and in relation to the total number of countries in the sample (Nc) with available data. It yields a normalized score of 0–10 for each country. The strength of the KAM methodology is its cross-sectoral approach, which allows users to take a holistic view of a wide range of relevant factors. The variables serve as proxies for the four pillars of the knowledge economy framework: • an economic and institutional regime to provide incentives for the efficient use of existing and new knowledge and the flourishing of entrepreneurship • an educated and skilled population to create, share, and use knowledge well • an efficient innovation system of firms, research centers, universities, consultants, and other organizations to tap into the growing stock of global knowledge, assimilate and adapt it to local needs, and create new technology • ICT to facilitate the effective creation, dissemination, and processing of information. The methodology includes several variables that track the performance of the economy. These variables indicate how well an economy is using knowledge for its overall economic and social development. KAM offers several preset display modes for simple visual representations of a country’s knowledge economy readiness. A country can be assessed and compared with others on the aggregate performance on each of the knowledge economy pillars or the overall Knowledge Economy and Knowledge Indexes for 1995–2000 and the most recent available year. KAM also makes possible customized country analysis and cross-country comparison on the indicators chosen, allowing users to focus on specific aspects of a country’s ability to generate, diffuse, and apply knowledge for economic development. Source: World Bank (http://go.worldbank.org/39Z6SV9C80). Figure 4.12. Commercial Services Exports versus to the adoption and diffusion of ICT. As for policy, the Management Score in Selected Countries, 2006–09 literature suggests that one-size-fits-all management practices are not the solution. For each services sector, 27 Log of commercial services exports policy makers need to carefully assess which manage- ment or entrepreneurial skills are necessary to increase output and productivity. 26 Trade-Related Infrastructure 25 Trade-related infrastructure encompasses facilities that enable exporting firms to reduce costs related to the delivery of goods and services. This infrastructure includes systems that reduce transportation, transaction, and search 24 costs for firms and consumers. Modernizing customs, for Module 4 2.6 2.8 3.0 3.2 3.4 example, makes it easier to trade goods. Documentation Overall management score procedures, trade facilitation initiatives, and public infra- Sources: Trade in Services Database and Bloom and van Reenen 2010. structure also facilitate trade. Many of these factors help Note: The overall management score is based on the responses to 18 questions on management practice. deliver goods efficiently. Cross-country evidence indicates that economies with large shares of services in value added tend to have better trade-related infrastructure than countries with smaller shares (figure 4.13). Wide variations exist, however, including among countries in the same income group. Module 4: Policy Options for Increasing Competitiveness and Trade in the Services Sector 121 Figure 4.13. Value Added of Services as Percent of GDP versus Logistics Performance Index in Selected Countries, 2010 100 LUX Value added of services as percent of GDP 80 PAN USA MDA LBN SGP JPN DEU MUS BRA CRI 60 MKD RUS MEX PRY CHL IND GHA UGA MOZ MYS CHN ARE KHM 40 GIN ETH SAU 20 1.5 2.0 2.5 3.0 3.5 4.0 Logistics Performance Index (LPI) Fitted values High-income country Upper-middle-income country Low-income country Lower-middle-income country Sources: World Bank Logistics Performance Index and World Development Indicators. Note: The LPI is a quantitative and qualitative measure assessing the logistics friendliness and the logistics supply chain. For the International Organization for Standardization (ISO) abbreviations, see appendix B. The increased use of telecommunications has vastly Freund and Weinhold (2000) find that a 10 percent increased the scope of services trade. It is the most pow- increase in Internet penetration in the partner country is erful symbol of vitality in the services sector. It is also an associated with a 1.2 percent increase in its imports of busi- important source for development of other services. ICT ness and professional services. Using different measures has reduced the costs of delivering many cross-border of telecom infrastructure (such as the number of fixed services to virtually zero. and mobile phone subscribers per 1,000 people), Shingal Empirical research finds that electronic infrastruc- (2010) concludes that the quality of infrastructure in the ture has a positive effect on services exports (Freund importing country is more important than the quality of and Weinhold 2000). Figure 4.14 illustrates the cross- infrastructure in the exporting country. Using the OECD’s country correlation between Internet penetration and telecommunication index, Lennon, Mirza, and Nicoletti trade in services. It shows that countries with higher (2009) conclude that telecom infrastructure in both the Module 4 densities of Internet connections have higher levels exporting and importing countries are important. These of services trade through Modes 1 and 2. It also shows authors show that telecom and ICT are particularly impor- that developed countries have the highest penetration of tant for commercial services. One lesson that can be drawn Internet users and levels of trade, whereas low-income from these studies is that policy makers should target countries are at the bottom line for both indicators (an export destinations that are well equipped with ICT and exception is India). have a high density of telecom or Internet connections. 122 Valuing Services in Trade Figure 4.14. Internet Penetration versus Trade in Services in Selected Countries, 2010 0.5 IRL 0.4 LBN Trade in services/GDP 0.3 PAN 0.2 KHM MYS MDA GBR 0.1 EGY MKD MOZIND UGA PRY PHL DEU LSO PAK BTN CHL BOL RUS USA JPN BRA 0 ETH GINCMR GHA DZA MUS ECU CHN MEX VEN CRI 0 20 40 60 80 100 Percentage of internet users Fitted values High-income country Upper-middle-income country Low income country Lower-middle-income country Sources: World Development Indicators and Trade in Services Database. Note: For the International Organization for Standardization (ISO) abbreviations, see appendix B. Notes Arnold, Jens M., Beata S. Javorcik, and Aaditya Mattoo. 2011. “Does Services Liberalization Benefit Manufacturing Firms? Evidence from 1. For more on this issue, see Copeland and Mattoo (2008), Deardoff the Czech Republic.” Journal of International Economics 85 (1): 136–46. and Stern (2008), and Hoekman and Kostecki (2009). Arnold, Jens M., Aaditya Mattoo, and Gaia Narciso. 2008. “Services 2. For a discussion of and policy framework on regulations and Inputs and Firm Productivity in Sub-Saharan Africa: Evidence from services trade, see Molinuevo and Sáez (2014). Firm-Level Data.” Journal of African Economies 17 (4): 578–99. 3. The criteria have not been tested under the WTO dispute settle- Australian Productivity Commission. 2011. “Planning and Zoning ment system. Although there is an ongoing work program to further Regulations.” In Economic Structure and Performance of the Australian develop these criteria, progress has been elusive. Retail Industry, chapter 8. Melbourne. 4. Park (2002) assumes an elasticity of substitution of 5.6 for all Baldwin, R. 2011. “Trade and Industrialisation after Globalisation’s 2nd services sectors; Fontagné, Guillin, and Mitaritonna (2011) use the same Unbundling: How Building and Joining a Supply Chain Are Different value. Walsh (2006) uses a trade-weighted elasticity of substitution for and Why It Matters.” NBER Working Paper 17716, National Bureau the services sector as a whole of 1.95, calculated from the Global Trade of Economic Research, Cambridge, MA. http://www.nber.org/papers Analysis Project. Francois (2001) uses an elasticity of 4 for overall services /w17716. trade. Francois, van Meijl, and van Tongeren (2003) use elasticities of Balistreri, Edward J., Thomas F. Rutherford, and David G. Tarr. 2009. 1.26–1.68 for different services sectors. “Modeling Services Liberalization: The Case of Kenya.” Economic Modelling 26 (3): 668–79. Bennet, D. 2002. “Innovative Technology Transfer Framework Linked Module 4 References to Trade for UNIDO Action.” World Summit on Sustainable Development, United Nations Industrial Development Organization, Acemoglu, Daron, and Simon Johnson. 2005. “Unbundling Institutions.” Johannesburg, August 26–September 4. Journal of Political Economy 113 (5): 949–95. Blomström, Magnus, and Ari Kokko. 1998. “Multinational Corporations Amin, Mohammad, and Aaditya Mattoo. 2006. “Do Institutions Matter and Spillovers.” Journal of Economic Surveys 12 (3): 247–77. More for Services?” Policy Research Working Paper Series 4032, Bloom, Nicholas, and John Van Reenen. 2010. “Why Do Management World Bank, Washington, DC. Practices Differ across Firms and Countries?” Journal of Economic Arnold, Jens M., Beata Javorcik, Molly Lipscomb, and Aaditya Mattoo. Perspectives 24 (1): 203–24. 2010. “Services Reform and Manufacturing Performance: Evidence Borchert, Ingo, Batshur Gootiiz, and Aaditya Mattoo. 2012a. “Guide from India.” CEPR Discussion Paper 8011, Centre for Economic to the Services Trade Restrictions Database.” World Bank Policy Policy Research, London. Research Working Paper 6108, Washington, DC. Module 4: Policy Options for Increasing Competitiveness and Trade in the Services Sector 123 ———. 2012b. “Policy Barriers to International Trade in Services: New Gwartney, James, Robert Lawson, and Joshua Hall. 2012. Economic Empirical Evidence.” World Bank Policy Research Working Paper Freedom of the World: Annual Report. Vancouver: Fraser Institute. 6109, Washington, DC. http://www.freetheworld.com/2012/EFW2012-complete.pdf. Borzenstein, Eduardo, José De Gregorio, and Jong-Wha Lee. 1998. “How Hall, Robert E., and Charles I. Jones. 1999. “Why Do Some Countries Does Foreign Direct Investment Affect Economic Growth?” Journal of Produce So Much More Output per Worker Than Others?” Quarterly International Economics 45 (1): 115–35. Journal of Economics 114 (1): 83–116. Cadot, Olivier, Maryla Maliszewska, and Sebastián Sáez. 2011. “Nontariff Helpman, Elhanan. 2011. Understanding Global Trade. Cambridge, MA: Measures: Impact, Regulation, and Trade Facilitation.” In Border Harvard University Press. Management Modernization, edited by Gerard McLinden, Enrique Hoekman, Bernard, and Michel Kostecki. 2009. The Political Fanta, David Widdowson, and Tom Doyle, 215–30. Washington, DC: Economy of the World Trading System, 3rd. edited by Oxford: Oxford World Bank. University Press. CEPII (Centre d’Etudes Prospectives et d’Informations Internationales). Inklaar, R., M. P. Timmer, and B. van Ark. 2008. “Market Services GeoDist database, Paris. http://www.cepii.fr/anglaisgraph/bdd Productivity Across Europe and the US.” Economic Policy 23: 141–94. /distances.htm. Jensen, J. Bradford. 2008. “Trade in High-Tech Services.” Journal of Copeland, Brian, and Aaditya Mattoo. 2008. “The Basic Economics of Industry, Competition, and Trade 8 (3–4): 181–97. Services Trade.” In A Handbook of International Trade in Services, Jensen, Jesper, Thomas F. Rutherford, and David G. Tarr. 2008. edited by Aaditya Mattoo, Robert M. Stern, and Gianni Zanini, “Modeling Services Liberalization: The Case of Tanzania.” Policy 84–129. Oxford: Oxford University Press. Research Working Paper Series 4801, Washington, DC. Costinot, Arnaud. 2009. “On the Origins of Comparative Advantage.” Jensen, Jesper, and David G. Tarr. 2012. “Deep Trade Policy Options for Journal of International Economics 77 (2): 255–64. Armenia: The Importance of Trade Facilitation, Services and Standards Deardoff, Alan V., and Robert M. Stern. 2008. “Empirical Analysis of Liberalization.” Economics, The Open-Access, Open-Assessment Barriers to International Services Transactions and the Consequences E-Journal 6 (1): 1–54, Kiel Institute for the World Economy. of Liberalization.” In A Handbook of International Trade in Services, Jones, Ronald W. 2000. Globalization and the Theory of Input Trade. edited by Aaditya Mattoo, Robert M. Stern, and Gianni Zanini, Cambridge, MA: MIT Press. 169–220. Oxford: Oxford University Press. Kaufmann, Daniel, Aart Kraay, and Massimo Mastruzzi. 2009. Dihel, Nora, Ana Margarida Fernandes, Richard Gicho, John “Governance Matters VIII: Aggregate and Individual Governance Kashangaki, and Nicholas Strychcz. 2012. “Becoming a Global Indicators, 1996–2008.” World Bank Policy Research Paper 4978, Exporter of Business Services? The Case of Kenya.” In Exporting Washington, DC. Services: A Developing Country Perspective , edited by Arti Kelle, Markus, J. Kleinert, Horst Raff, and Farid Toubal. 2013. “Cross- Grover Goswami, Aaditya Mattoo, and Sebastián Sáez, 237–69. Border and Affiliate Sales of Services: Evidence from German Washington, DC: World Bank. Microdata.” World Economy 36 (11): 1373–92. Feenstra, Robert C. 2010. Offshoring in the Global Economy: Microeconomic Lall, Sanjaya. 1999. “Competing with Labour: Skills and Technology in Structure and Macroeconomic Implications . Cambridge, MA: Developing Countries.” Issues in Development Working Paper 31, MIT Press. International Labour Organization, Geneva. Findlay, Ronald, and Henryk Kierzkowski. 1983. “Human Capital and Lejour, Arjan, and Jan-Willem de Paiva Verheijden. 2004. “Services International Trade: A Simple General Equilibrium Model.” Journal Trade within Canada and the European Union: What Do They Have of Political Economy 91: 957–78. in Common?” CPB Discussion Paper 42, Netherlands Bureau for Fontagné, L., A. Guillin, and C. Mitaritonna. 2011. “Estimations of Economic Policy Analysis, The Hague. Tariff Equivalents for the Services Sectors.” Working Paper 2011–24, Lennon, Carolina. 2009. “Trade in Services and Trade in Goods: Centre d’Etudes Prospectives et d’Informations Internationales Differences and Complementarities.” WIIW Working Paper 53, (CEPII), Paris. Vienna Institute of International Economics Studies (WIIW), Vienna. Francois, J. 2001. The Next WTO Round: North-South Stakes in New Lennon, Carolina, Daniel Mirza, and Giuseppe Nicoletti. 2009. Market Access Negotiations. Adelaide: Centre for International “Complementarity of Inputs in Services Trade.” Annales d’Economie Economic Studies. et de Statistiques 93/94 (April/June): 183–205. Francois, J., H. van Meijl, and F. van Tongeren. 2003. Economic Benefits of Levchenko, Andrei A. 2007. “Institutional Quality and International the Doha Round for the Netherlands. Agricultural Economics Institute, Trade.” Review of Economic Studies 74 (3): 791–19. The Hague. Miroudot, Sébastien, Rainer Lanz, and Alexandros Ragoussis. 2009. Freund, Caroline, and Diana Weinhold. 2002. “The Internet and “Trade in Intermediate Goods and Services.” OECD Trade Policy International Trade in Services.” American Economic Review 92 (2): Working Papers 93, Paris. http://dx.doi.org/10.1787/5kmlcxtdlk8r-en. 236–40. Miroudot, Sébastien, Jehan Sauvage, and Ben Shepherd. 2010. “Measuring Gibbs, J. Murray. 1986. “Services, Development, and TNCs.” CTC the Cost of International Trade in Services.” MPRA Paper 27655, Reporter 21 (Spring): 51–53. Munich Personal RePEc Archive, University Library of Munich. Gregory, Neil, Stanley Nollen, and Stoyan Tenev. 2009. New Industries Molinuevo, Martin, and Sebastián Sáez. 2014. Regulatory Assessment from New Places: The Emergence of the Hardware and Software Toolkit: A Practical Methodology for Assessing Regulation on Trade and Industries in China and India. Redwood City, CA: Stanford University Investment in Services. Washington, DC: World Bank. Module 4 Press and Washington, DC: World Bank. Ng, Francis, and Alexander Yeats. 2000. “Production Sharing in East Asia: Grossman, Gene M., and Esteban Rossi-Hansberg, 2008. “Trading Tasks: Who Does What for Whom, and Why?” World Bank Policy Research A Simple Theory of Offshoring.” American Economic Review 98 (5): Working Paper 2197, Washington, DC. 1978–97. Nunn, Nathan. 2007. “Relationship-Specificity, Incomplete Contracts, Grünfeld, L., and A. Moxnes. 2003. “The Intangible Globalisation: and the Patterns of Trade.” Quarterly Journal of Economics 122 (2): Explaining Patterns of International Trade in Services.” Norwegian 569–600. Institute of International Affairs Paper 657, Oslo. Nusbaumer, Jacques. 1987. Services in the Global Market. Amsterdam: Guislain, Pierre, edited by 2004. “The Postal Sector in Developing and Kluwer. Transition Countries: Contributions to a Reform Agenda.” World O’Mahony, Mary, and Bart van Ark. 2005. “Assessing the Productivity of Bank, Global Information and Communication Technologies the UK Retail Trade Sector: The Role of ICT.” International Review of Department, Policy Division, Washington, DC. Retail, Distribution and Consumer Research 15 (3): 297–303. 124 Valuing Services in Trade Park, S.-C. 2002. “Measuring Tariff Equivalents in Cross-Border Trade Triplett, Jack E., and Barry P. Bosworth. 2004. Productivity in the US in Services.” Working Paper 02-15, Korea Institute for International Services Sector: New Sources of Economic Growth. Washington, DC: Economic Policy, Seoul. Brookings Institution Press. PEZA (Philippine Economic Zone Authority). 2000. “Guidelines on the Trade in Services Database. World Bank, Washington, DC. Registration of Information Technology (IT) Enterprises and the UNCTAD (United Nations Conference on Trade and Development). Establishment and Operation of IT Parks/Buildings.” Certificate of 1999. World Investment Report 1999: Foreign Direct Investment and the Board Resolution 00-411, December 29, Pasay City. http://www.peza Challenge of Development. Geneva: UNCTAD. .gov.ph/issuances/guidelines/Guidelines_IT.pdf. van der Marel, Erik. 2011. “Determinants of Comparative Advantage in Reis, José Guilherme, and Thomas Farole. 2012. Trade Competitiveness Services.” GEM Working Paper, Groupe d’Economie Mondiale, Paris. Diagnostic Toolkit. Washington, DC: World Bank. ———. 2012. Trade in Services and TFP: The Role of Regulation.” World Rossotto, Carlo Maria, Khalid Sekkat, and Aristomene Varoudakis. Economy 35 (11): 1387–429. 2003. “Opening up Telecommunications to Competition and MENA Walsh, Keith. 2006. “Trade in Services: Does Gravity Hold? A Gravity Integration in the World Economy.” Middle East and North Africa Model Approach to Estimating Barriers to Services Trade.” IIIS Working Paper 33, World Bank, Washington, DC. Discussion Paper 183, Institute for International Integration Studies, Saxenian, Anna Lee. 1999. Silicon Valley’s New Immigrant Entrepreneurs. Trinity College, Dublin. San Francisco, CA: Public Policy Institute of California. WEF (World Economic Forum). 2010. Human Capital Report. Geneva: ———. 2000. “Back to India.” Wall Street Journal: Technology Journal World Economic Forum. Asia, January 24. Wolf, E. N. 1999. “The Productivity Paradox: Evidence from Indirect ———. 2004. “The Bangalore Boom: From Brain Drain to Brain Indicators of Services Sector Productivity Growth.” Canadian Journal Circulation.” In IT Experience in India: Bridging the Digital Divide, of Economics 32 (2): 281–308. edited by. Kenneth Keniston and Deepak Kumar, 169–81. New Delhi: World Bank Governance Indicators. Washington, DC. Sage. www.govindicators.org. Shepherd, Ben, and Erik van der Marel. 2013. “Services Trade, Regulation World Development Indicators (database). World Bank, Washington, and Regional Integration: Evidence from Sectoral Data.” World DC. http://data.worldbank.org/data-catalog/world-development Economy 36 (11): 1393–405. -indicators. Shingal, Anirudh. 2010. “How Much Do Agreements Matter for Services WTO (World Trade Organization). 2000. “Korea: Measures Affecting Trade?” Paper presented at the World Trade Organization Public Imports of Fresh, Chilled and Frozen Beef.” WT/DS161/AB/R, WT Forum, Geneva, September 17. /DS169/AB/R, December 11. Sudan, Randeep, Seth Ayers, Philippe Dongier, Arturo Muente-Kunigami, ———. 2003. “The Necessity Tests in the WTO.” Working Party on and Chirstine Zhen-Wei Qiang. 2010. The Global Opportunity in Domestic Regulation, S/WPDR/W/27. December 2. IT-Based Services: Assessing and Enhancing Country Competitiveness. ———. 2005. “United States: Measures Affecting the Cross-Border Washington, DC: World Bank. Supply of Gambling and Betting Services.” WT/DS285/AB/R, April 7. Teerawattananon, Yot, Viroj Tangcharoensathien, Sripen Tantivess, and ———. 2011. “The Necessity Tests in the WTO.” Working Party on Anne Mills. 2003. “Health Sector Regulation in Thailand: Recent Domestic Regulation, S/WPDR/W/27/Add.1. January 18. Progress and the Future Agenda.” Health Policy 63 (3): 323–38. Yi, Soonhwa. 2012. “Reaching the World through Private Sector Thomson, R. and J. Nielson. 2001. “Consolidated List of Cross-Sectoral Initiative: Service Exports from the Philippines.” In Exporting Services: and Sector-Specific Barriers.” In Trade in Services: Negotiating A Developing-Country Perspective, edited by Arti Goswami, Additya Issues and Approaches, 11–26, Paris: Organisation for Economic Mattoo, and Sebastián Sáez, 121–60. Washington, DC: World Bank. Co-operation and Development (OECD). Module 4 Appendix A: EXPORT OF Value Added Database Trade data is usually measured at transaction values, which backward linkages—the contribution of all other sectors is the price actually paid or payable for goods and services. to a particular sector’s exports. These value chain linkages Transaction values of goods and services are gross val- not only capture the full (direct and indirect) contribu- ues, or value added plus domestic and foreign intermedi- tion of the services sector to a country’s exports, but also ate inputs. The measure of gross exports may undervalue inform a comprehensive strategy to improve export com- the real contribution of a sector to trade if value added petitiveness of both goods and services. from this sector is embedded as inputs in other sectors’ exports. This is particularly true for services exports. Why This Database? Alternatively, the measure of gross exports may overvalue the real contribution to trade if the sector’s exports embed The Export of Value Added Database provides a full inputs from other sectors. This is particularly true for picture of the economic relationships or value-added manufacturing exports. Measuring trade on a value-added structure between sectors that are accounted for in basis, as achieved in the Export of Value Added Database, national income, or gross domestic product (GDP). The overcomes this shortcoming. underlying construction of the database stems from eco- This database follows up on pioneering work of nomics and is based on a general equilibrium principle, Christen, Francois, and Hoekman (2012) and Francois, which means that every income of an economy also has Manchin, and Tomberger (2013). It provides information a counterpart in terms of expenditure so that all receipts on the domestic value-added content of domestic output and outlays correspond with each other. The strength of and exports across 27 sectors of the economy, includ- this database, therefore, is its comprehensive setting out of ing 9 commercial services sectors, 3 primary sectors, and the interrelationships within an economy of linkages that 14 manufacturing sectors. record intermediate input use and final demand. Thus this alternative measure to trade makes explicit This value-added structure can be broken down into the direct value-added contribution of services sectors direct and indirect value added. Direct value added cap- to domestic production as well as exports, as well as the tures the true sector-specific value added generated within linkages that services provide to all other sectors of the an economy and nets out domestic and foreign inputs. economy in terms of value added. In addition to direct The direct contribution of each services sector to value potential for export, services have indirect potential as added, either of the domestic economy as a whole or only important inputs for other economic activities’ exports. of exports, is the value added sold directly to final con- The services sector is intrinsically linked with the overall sumers. For example, in Bangladesh, trade and transport activity of an economy through value chains. By measur- services represent 10.0 percent of the direct value added of ing exports on a value-added basis, it is possible to mea- the domestic economy and 1.0 percent of the direct value sure these value chain linkages between the services sector added of exports. However, the contribution of this sec- and other export sectors of the economy. This includes tor to overall value added and exports of the economy is both forward linkages—the contribution of a particu- much higher, since transport services are used as inputs lar sector as an input to others sectors’ exports—and by other sectors but also use inputs from other sectors. 125 126 Valuing Services in Trade Indirect value added adds to the direct measure the por- dataset measures and describes domestic consumption, tion of value added of the inputs that are produced domes- production, intermediate input use, and trade patterns of tically, expressed in terms of forward or backward linkages. goods and services across sectors of an economy. It covers Forward linkages represent the contribution of a partic- not only key OECD countries but also a wide range of ular sector as an input to other sectors’ value added. More developing countries. Hence, the dataset represents an specifically, they are the supply response of a particular sector advanced input-output panel—also known as a social to all other (downstream) sectors’ demand for more inputs. accounting matrix—of incomes and expenditures for the For example, forward linkages will indicate how much other domestic economy and their link to trade over intermittent sectors are using financial services as inputs. In Bangladesh, years between 1992 and 2011. trade and transport services represent 28.9 percent of value This database is developed for various years, so one can added of the domestic economy and 24.2 percent of value obtain data for the linkages in recent years and observe how added of exports after considering this sector’s contribution the linkages have changed over time. It should be noted, as inputs in all other sectors’ value added. however, that changes such as a rebasing of the national Analogously, backward linkages represent the contribu- accounts, for example, may make cross-year comparisons tion of all other sectors to a particular sector’s value added. unreliable in some countries. The full set of economies More specifically, they represent how much the demand and years in the Export of Value Added Database is listed of a particular sector will pull the supply of all other in table A.2 at the end of this appendix. Table A.3 lists (upstream) sectors. For example, backward linkages will the 27 sectors in the database; table A.4 provides details indicate how much financial services are using other sec- regarding their classifications. tors as inputs. In Bangladesh, trade and transport services only demanded 11.5 percent of value added of the domes- tic economy and 4.9 percent of value added of exports Methodology to Construct the Export of across all sectors. Value Added Database These linkages represent the interdependence of The following is a formal representation of how to obtain sectors in the economy. Industries with strong backward the direct and indirect (in terms of forward and backward and forward linkages play an important role in the devel- linkages) value-added content of domestic output and opment strategy of a country. A sector with strong back- exports across sectors of an economy. The variables in the ward linkages means that an increase in the final demand database are explained in box A.1. Equation A.1 provides a of these industries’ output will have a large impact on simple representation of a domestic economy in terms of industries that supply inputs in the production of these gross output and intermediate input requirements: industries’ output, while a sector with strong forward linkages means that an increase in the fi nal demand Y = Z − AZ. (A.1) of other industries’ output will have a large impact on where Y denotes a vector of final demand of each sector, the industry. Naturally, strong linkages to value added Z denotes a vector of gross outputs of each sector, and A exports suggest an important role in the export strategy denotes a matrix of intermediate use coefficients. of a country. First, it is necessary to calculate how much input Manufacturing and services contribute differently to from each sector is contained by one unit of final out- exports due to a different structure of forward and back- put of all other sectors. This is the so-called intermediate ward linkages. From a policy point of view, when looking multiplier matrix, also commonly known as the Leontief at services, the primary interest is forward linkages, as ser- matrix or M-matrix. Equation A.2 is used to construct the vices tend to be used as inputs to other sectors of the econ- M-matrix: omy. On the other hand, backward linkages are dominant for manufacturing sectors because manufacturing tends to Z = (1−A)−1Y = MY. (A.2) use inputs from other sectors of the economy (or imports from other countries). This M-matrix hence holds information on direct and indirect input use in each sector. Each sector uses a differ- ent amount of intermediate input from all other sectors, Data Source and hence each linkage or intermediate use coefficient The Export of Value Added Database is constructed using between sectors will vary. input-output tables. The underlying data come from the Second, it is necessary to have information about Global Trade Analysis Project (GTAP) dataset.1 This global the shares of each sector’s domestic value added in total Appendix A: Export of Value Added Database 127 Box A.1. Export of Value Added Database Variables Sector_GMatrix: This matrix contains the direct and indirect value-added content of domestic output. Depending on whether rows or columns are considered, its sum corresponds to forward (row) or backward (column) linkages. Thus reading a row for a given sector (sector presented on the y-axis) provides information about how much this sector went into each sector (on the x-axis) as inputs. The diagonal corresponds to the direct value added contribution. The matrix represents to the G-matrix as described above. Sector_HMatrix: This matrix contains the direct and indirect value-added content of exports. Depending on whether rows or columns are consid- ered, its sum corresponds to forward (row) or backward (column) linkages. Thus reading a row for a given sector (sector presented on the y-axis) provides information about how much this sector went into each sector (on the x-axis) as inputs. The diagonal corresponds to the direct value added contribution. The matrix represents to the H-matrix as described above. DomVAshare: This vector contains the share of each sectors’ domestic value added in gross output value. It is the diagonal of the B-matrix as described above. GXshare: This vector contains the share of each sector’s gross exports in total exports based on gross value of exports. DXshare: This vector contains the share of each sector’s exports of value added in total exports of value added based on direct value added, ignoring linkages. VXsharefwd: This vector contains the share of each sector’s exports of total value added in total exports of value added based on forward linkages. This vector corresponds to the share of the row-sums in the total sum of the H-matrix as described above. VXsharebwd: This vector contains the share of each sector’s exports of total value added in total exports of value added based on backward linkages. This vector corresponds to the share of the column-sums in the total sum of the H-matrix as described above. output, which is then of course consumed domestically or This obtains the direct and indirect value-added content of otherwise exported, and is called the B-matrix. The output output (G-matrix) and exports (H-matrix) across sectors vector (Z) is measured in gross terms whereas the goal of of an economy: this database is to obtain a cleaner “net” measure of output in value-added terms. As some share of the sector’s gross G = VY (A.4) output measure also involves value added created within that sector, the next step is to disentangle the flow of gross activities into the value added activities and intermediate H = V X. (A.5) use. The B-matrix is a diagonal matrix where the diagonal elements are the value added shares of gross output Z of Note that each of the matrixes contains each sector’s each sector. direct value-added contribution on the diagonal and the Third, the M-matrix (containing the intermediate input indirect value-added contribution through forward and use shares) is then multiplied with the B-matrix (a diagonal backward linkages off the diagonal. Summing across col- matrix containing the value added shares of gross output) umns within a row provides the forward linkages, while to obtain a V-matrix, which identifies the inputs of value summing across rows within a column provides the back- added from each sector related to a unit of final demand of ward linkages. all other sectors: V = BM. (A.3) Methodology Example Finally, in order to obtain flows of value added broken Table A.1 shows how this methodology is applied for down across sectoral activities, the V-matrix is then mul- Turkey. It provides some measures based on how value tiplied by a diagonal matrix whose non-zero elements are added is measured in the database, both for the domestic the vector of final outputs (Y ) or exports ( X ) of each sector. economy as well as for exports. Column 1 shows measures 128 Valuing Services in Trade Table A.1. Value Added in Final Output in Turkey, by Sectors, 2007 Domestic Direct Total value added Total value added value added Gross export value added export share considering export share considering share share export share forward linkages backward linkages Sectora (1) (2) (3) (4) (5) 1 primaryagr 7% 3% 5% 4% 4% 2 otherprim 2% 2% 3% 3% 2% 3 energy 2% 2% 1% 4% 1% 4 procfoods 6% 3% 3% 3% 4% 5 bevtobacco 1% 0% 0% 0% 0% 6 textiles 2% 10% 9% 8% 10% 7 clothing 1% 7% 5% 3% 7% 8 leather 0% 0% 0% 0% 0% 9 lumber 0% 1% 1% 1% 1% 10 paperpub 1% 1% 1% 1% 1% 11 chemicals 2% 6% 4% 4% 5% 12 nonmetmin 2% 3% 3% 2% 3% 13 metals 1% 9% 5% 5% 7% 14 fabmetals 1% 3% 3% 4% 3% 15 transpequip 2% 14% 13% 8% 14% 16 machinery 3% 12% 11% 8% 12% 17 othermanuf 1% 2% 1% 1% 1% 18 water 1% 0% 0% 0% 0% 19 construction 7% 1% 1% 0% 1% 20 distribution 17% 1% 3% 10% 2% 21 transport 11% 15% 19% 16% 16% 22 communication 2% 0% 1% 2% 1% 23 finance 7% 1% 1% 5% 1% 24 insurance 0% 1% 1% 1% 1% 25 obsict 4% 1% 1% 3% 1% 26 oconsumer 1% 1% 2% 1% 1% 27 oservices 18% 1% 3% 2% 2% Total 100% 100% 100% 100% 100% Services categories 22–27 31% 5% 8% 14% 6% Source: Calculations based on the Export of Value Added Database. a. See table A.3 for sector descriptions. of value added for the total economy, i.e. the allocation However, as previously explained, with this database, of value-added across sectors. It shows how much value the value-added embodied in exports for Turkey’s services added (including direct and forward linkages) is pro- sector can be estimated. Because all linkages across sectors duced in each sector whereas column 2 shows the share of are accounted for, it is possible to estimate the total con- exports at gross terms in total gross exports. It shows that tribution of services to total exports in two ways, namely services—including communication, finance, insurance, forward and backward. Analogously to the situation for and other business and information and communication domestic production, forward linkages tell how much each technology services—are an important sector because it sector contributes value added, which is exported directly generates 31 percent of value added in Turkey’s economy or indirectly in the production of other goods and services. although it only occupies 5 percent of gross exports. In Hence, it represents the contribution of a particular sec- contrast, in the case of machinery, for instance, these tor as an input to other sectors’ value added. Backward figures are reversed. Column 3 shows the direct contribu- linkages represent the contribution of all other sectors to a tion of services to total exports, that is, services exported particular sector’s value added. The shares once consider- for final consumption. For example, this is the case of ing direct plus forward linkages are given in column 4 and software that is ready to be installed in a particular firm backward linkages in column 5. For instance, in Turkey information system. commercial services account for 8 percent of total value Appendix A: Export of Value Added Database 129 Table A.2. Economies in the Export of Value Added Database Economy Income level Years covered in the database Albania low income 2011 2007 2004 2001 1997 Argentina upper middle income 2011 2007 2004 2001 1997 1995 1992 Armenia lower middle income 2011 2007 2004 Australia high income 2011 2007 2004 2001 1997 1995 1992 Austria high income 2011 2007 2004 2001 1997 Azerbaijan upper middle income 2011 2007 2004 Bahrain high income 2011 2007 2004 Bangladesh low income 2011 2007 2004 2001 1997 Belarus upper middle income 2011 2007 2004 Belgium high income 2011 2007 2004 2001 1997 Benin low income 2011 Bolivia lower middle income 2011 2007 2004 2001 Botswana upper middle income 2011 2007 2004 2001 1997 Brazil upper middle income 2011 2007 2004 2001 1997 1995 1992 Brunei Darassalam high income 2011 Bulgaria upper middle income 2011 2007 2004 2001 1997 Burkina Faso low income 2011 2007 2007 Cambodia low income 2011 2007 2004 Cameroon lower middle income 2011 2007 2004 Canada high income 2011 2007 2004 2001 1997 1995 1992 Caribbean Region – 2007 2004 Central Africa Region – 2011 2007 2004 Chile upper middle income 2011 2007 2004 2001 1997 1995 1992 China lower middle income 2011 2007 2004 2001 1997 1995 1992 Colombia upper middle income 2011 2007 2004 2001 1997 1995 Costa Rica upper middle income 2011 2007 2004 Côte d’Ivoire low income 2011 2007 2004 Croatia upper middle income 2011 2007 2004 2001 1997 Cyprus high income 2011 2007 2004 2001 1997 Czech Republic high income 2011 2007 2004 2001 1997 Denmark high income 2011 2007 2004 2001 1997 1995 Dominican Republic upper middle income 2011 Ecuador upper middle income 2011 2007 2004 2001 Egypt, Arab Rep. lower middle income 2011 2007 2004 El Salvador lower middle income 2011 2007 2004 Estonia high income 2011 2007 2004 2001 1997 Ethiopia low income 2011 2007 2004 Finland high income 2011 2007 2004 2001 1997 1995 France high income 2011 2007 2004 2001 1997 Georgia lower middle income 2011 2007 2004 1995 1992 Germany high income 2011 2007 2004 2001 1997 1995 Ghana low income 2011 2007 2004 Greece high income 2011 2007 2004 2001 1997 Guatemala lower middle income 2011 2007 2004 Guinea low income 2011 Honduras lower middle income 2011 2007 2004 Hong Kong SAR, China high income 2011 2007 2004 2001 1997 1995 1992 Hungary high income 2011 2007 2004 2001 1997 India lower middle income 2011 2007 2004 2001 1997 1995 1992 Indonesia lower middle income 2011 2007 2004 2001 1997 1995 1992 Iran, Islamic Rep. lower middle income 2011 2007 2004 2001 Ireland high income 2011 2007 2004 2001 1997 Israel high income 2011 2007 2004 Italy high income 2011 2007 2004 2001 1997 Jamaica upper middle income 2011 Japan high income 2011 2007 2004 2001 1997 1995 1992 (continued on next page) 130 Valuing Services in Trade Table A.2. (continued) Economy Income level Years covered in the database Jordan upper middle income 2011 Kazakhstan upper middle income 2011 2007 2004 Kenya low income 2011 2007 2004 Korea, Rep. high income 2011 2007 2004 2001 1997 1995 1992 Kuwait high income 2011 2007 2004 Kyrgyz Republic low income 2011 2007 2004 Lao PDR lower middle income 2011 2007 2004 Latvia upper middle income 2011 2007 2004 2001 1997 Lithuania upper middle income 2011 2007 2004 2001 1997 Luxembourg high income 2011 2007 2004 2001 1997 Madagascar low income 2011 2007 2004 2001 Malawi low income 2011 2007 2004 2001 1997 Malaysia upper middle income 2011 2007 2004 2001 1997 1995 1992 Malta high income 2011 2007 2004 2001 1997 Mauritius low income 2011 2007 2004 2001 Mexico upper middle income 2011 2007 2004 2001 1997 1995 1992 Mongolia lower middle income 2011 2007 2004 Morocco lower middle income 2011 2007 2004 2001 1997 1995 Mozambique low income 2011 2007 2004 2001 1997 Namibia low income 2011 2007 2004 Nepal low income 2011 2007 2004 Netherlands high income 2011 2007 2004 2001 1997 New Zealand high income 2011 2007 2004 2001 1997 1995 1992 Nicaragua lower middle income 2011 2007 2004 Nigeria low income 2011 2007 2004 2001 Norway high income 2011 2007 2004 Oman high income 2011 2007 2004 Pakistan low income 2011 2007 2004 2001 Panama upper middle income 2011 2007 2004 Paraguay lower middle income 2011 2007 2004 Peru upper middle income 2011 2007 2004 2001 1997 Philippines lower middle income 2011 2007 2004 2001 1997 1995 1992 Poland high income 2011 2007 2004 2001 Portugal high income 2011 2007 2004 2001 1997 Qatar high income 2011 2007 2004 Romania upper middle income 2011 2007 2004 2001 1997 Russian Federation upper middle income 2011 2007 2004 2001 1997 Rwanda low income 2011 Saudi Arabia high income 2011 2007 2004 Senegal low income 2011 2007 2004 Singapore high income 2011 2007 2004 2001 1997 1995 1992 Slovak Republic high income 2011 2007 2004 2001 1997 Slovenia high income 2011 2007 2004 2001 1997 South Africa upper middle income 2011 2007 2004 2001 Spain high income 2011 2007 2004 2001 1997 Sri Lanka lower middle income 2011 2007 2004 2001 1997 1995 Sweden high income 2011 2007 2004 2001 1997 1995 Switzerland high income 2011 2007 2004 2001 1997 Taiwan, China high income 2011 2007 2004 2001 1997 1995 1992 Tanzania low income 2011 2007 2004 2001 1997 Thailand lower middle income 2011 2007 2004 2001 1997 1995 1992 Togo low income 2011 Tunisia lower middle income 2011 2007 2004 2001 Turkey upper middle income 2011 2007 2004 2001 1997 1995 Uganda low income 2011 2007 2004 2001 1997 (continued on next page) Appendix A: Export of Value Added Database 131 Table A.2. (continued) Economy Income level Years covered in the database Ukraine lower middle income 2011 2007 2004 United Arab Emirates high income 2011 2007 2004 United Kingdom high income 2011 2007 2004 2001 1997 1995 United States high income 2011 2007 2004 2001 1997 1995 1992 Uruguay upper middle income 2011 2007 2004 2001 1997 1995 Venezuela, RB upper middle income 2011 2007 2004 2001 1997 1995 Vietnam low income 2011 2007 2004 2001 1997 1995 Zambia low income 2011 2007 2004 2001 1997 Zimbabwe low income 2011 2007 2004 2001 1997 Table A.3. Sectors in the Export of Value Added Database Sector name Sector description primaryagr Agriculture, forestry, and fisheries otherprim Minerals n.e.c. energy Energy extraction procfoods Processed foods bevtobacco Beverages and tobacco products textiles Textiles clothing Wearing apparel leather Leather products lumber Wood products paperpub Paper products and publishing chemicals Chemicals, rubber, and plastic products nonmetmin Mineral products n.e.c. metals Ferrous metals and metals n.e.c. fabmetals Metal products transpequip Transport equipment machinery Machinery and equipment n.e.c. othermanuf Manufactures n.e.c. water Water and utilities construction Construction distribution Distribution and trade transport Transport communcation Communications finance Finance insurance Insurance obsict Other business and information and communication technology oconsumer Other consumer services oservices Other services Note: n.e.c. = not elsewhere classified. 132 Valuing Services in Trade Table A.4. Service Sectors Mapping to International Standard Industrial Classification (ISIC) ISIC number Service sector Water: water and other utility service 401 Production, collection, and distribution of electricity 402 Manufacture of gas; distribution of gaseous fuels through mains 403 Steam and hot water supply 41 Collection, purification, and distribution of water Construction: construction 45 Construction Distribution: trade and distribution services 50 Sales, maintenance, and repair of motor vehicles and motorcycles; retail sale of automotive fuel 51 Wholesale trade and commission trade, except of motor vehicles and motorcycles 521 Nonspecialized retail trade in stores 522 Retail sale of food, beverages, and tobacco in specialized stores 523 Other retail trade of new goods in specialized stores 524 Retail sale of second-hand goods in stores 525 Retail trade not in stores 526 Repair of personal and household goods 55 Hotels and restaurants Transport: transport services 60 Land transport; transport via pipelines 61 Water transport 62 Air transport 63 Supporting and auxiliary transport activities; activities of travel agencies Communication: post and communications services 64 Post and telecommunications Finance: financial services 65 Financial intermediation, except insurance and pension funding 67 Activities auxiliary to financial intermediation Insurance: insurance services 66 Insurance and pension funding, except compulsory social security OBSICT: Other business and ICT services 70 Real estate activities 711 Renting of transport equipment 712 Renting of other machinery and equipment 713 Renting of personal and household goods n.e.c. 72 Computer and related activities 73 Research and development 74 Other business activities OConsumer: other consumer services 92 Recreational, cultural, and sporting activities 93 Other service activities 95 Private households with employed persons OServices: public services, dwellings 75 Public administration and defense; compulsory social security 80 Education 85 Health and social work 90 Sewage and refuse disposal, sanitation, and similar activities 91 Activities of membership organizations n.e.c. 99 Extra-territorial organizations and bodies dwellings (no ISIC mapping) Note: n.e.c. = not elsewhere classified. Appendix A: Export of Value Added Database 133 added exported directly and 14 percent when including References forward linkages; these are the services used as inputs to Christen, E., J. Francois, and B. Hoekman. 2012. “CGE Modeling of other sectors’ exports. When including backward linkages, Market Access in Services.” In Handbook of Computable General on the other hand, commercial services represent only Equilibrium Modeling, edited by P.B. Dixon and D.W. Jorgensen. Waltham, MA: Elsevier. 6 percent of total exports. Francois, J., M. Manchin, and P. Tomberger. 2013. “Services Linkages and the Value Added Content of Trade.” Policy Research Working Paper 6432, World Bank, Washington, DC. Note McDougall, R. 2001. The GTAP Database, Version 5. Centre for Global Trade Analysis. West Lafayette, IN: Purdue University. 1. Available at https://www.gtap.agecon.purdue.edu/. The basic McDougall, R., and J. Hagemeijer. 2005. “Services Trade Data.” In Global structure of GTAP is explained by McDougall (2001) and McDougall and Trade, Assistance and Production: The GTAP 6 Database, edited Hagemeijer (2005). It is produced by a consortium of institutions that by B. V. Dimaranan and R. McDougall. Centre for Global Trade include the World Bank, US International Trade Commission, World Analysis. West Lafayette, IN: Purdue University. Trade Organization, OECD, UNCTAD, UNFAO and a number of univer- sities and independent research institutes. Appendix B: Trade in Services Database The collection of data on cross-border trade in services consumption abroad. Modes 3 and 4, which are not cov- is notoriously difficult. This is of course due, in part, ered here, are commercial presence of foreign firms that to the intangible nature of services. In addition, some- deliver a service and service provided through the presence times national statistical agencies simply do not have the of natural persons, respectively. resource capacity to collect services trade data. Quite often This database has collected data of services trade based this is the case for developing countries. Although some on the official information reported on a country’s bal- international institutions collect data on services trade, ance of payment (BOP) statistics. Therefore, only Modes 1 cross-border trade in services is never complete. This data- and 2 could be included. Although these two categories are base tries to fill that gap by consolidating multiple sources also used when constructing (inter)national input/output of bilateral trade data in services to provide a broader tables—as used for the Export of Value Added Database— coverage of developed and developing countries over time. one should note that Mode 3 is also a very important chan- As such, the Trade in Services Database should be seen as nel of foreign services delivery. Mode 3 trade is often proxied the best approximation currently available to capture a by foreign direct investment (FDI) because the official trade comprehensive picture of global trade flows in services. statistics for this mode, namely Foreign Affiliates Trade in To combine all these different data sources and to Services (FATS), are hard to obtain. Unfortunately, cur- make each bilateral trade relationship between countries rently only a few countries in the world report FATS statis- consistent, the creators of this database applied a mirror tics. Nonetheless, from FDI data it is known that about 60 technique and performed many checkups to make data as percent of global FDI stock is undertaken in the services sec- comprehensive as possible. Mirror technique is a method tor, with finance and trade being the most important sectors. to retrieve export trade flows of a partner by using infor- mation on imports of the reporter country. Yet, although Data Sources data on many bilateral trade relationships between North- South countries could be recovered, one weakness of this The underlying data sources are taken from Eurostat, dataset is that a substantial share of South-South trade International Monetary Fund (IMF), Organisation for remains largely unreported. The magnitude of this gap Economic Co-operation and Development (OECD), and can be verified by comparing trade with the world with all United Nations (UN). Most of these databases report existing bilateral flows in the database. bilateral trade relationships between individual countries; One should bear in mind that this database measures only the IMF provides solely trade with the world as a mainly cross-border trade flows in services together partner. All sources also give a breakdown of sectors, which with consumption abroad. As such, other modes of sup- allowed the creators to construct a database by exporter, ply within the services framework are left unmeasured. partner, sector, and year. Services sectors are divided As explained in the introduction of this toolkit, there are into individual BOP codes at aggregate and disaggregate four modes of supply. Mode 1 covers cross-border trade, levels. Table B.2 provides a complete list of the sectors which is mostly services traded over the internet, or more covered, including each BOP code and chapter division formally, services supplied from the territory of one and table B.3 lists the economies covered, including each country into the territory of another. Mode 2 represents ISO 3-digit country code. 135 136 Valuing Services in Trade The most comprehensive database of reporting coun- Table B.1. Variable Descriptions tries for bilateral services trade among the sources used Variable name Description is the UN, which provides data on 190 economies. The REP ISO 3-digit code for reporting country (importer) other two databases—from Eurostat and OECD—release PAR ISO 3-digit code for partner country (exporter) YEAR Year data for only a limited number of economies. Specifically, BOP BOP Manual 5 3-digit code for services activity the OECD covers 34 country markets, which include VALUE Services imports value (current US$ millions) its members and various emerging economies. Eurostat DESCRIPTION Sector description covers 28 EU member countries (including Croatia) plus NAME Services activity (sector) name Iceland, Japan, Norway, Switzerland, Turkey, and the United States. was found, the authors made adjustments concordantly. Such checks were also performed for trade flows of each reporter-partner with respect to overall trade with the Compilation world. If differences were found between the aggregate Not only is the collection of services trade data difficult, figure and the sum of its parts—either by sector or part- the quality of the data is also often poor compared to ner country—the difference between its initial value and merchandise trade statistics. One reason is that services the sum of its components was changed to correct for the cannot easily be taxed when imported due to their intan- aggregate value. gible nature. To give the best quality possible, the creators of the Trade in Services Database have adjusted the data Database Description extensively to assure consistency. This assurance involved cross-checking across the four Together, the dataset contains bilateral services trade flows data sources to identify inconsistencies where identi- for 195 economies as reporters and partners, plus the cat- cal bilateral trade relations are reported more than once. egory “rest of the world.”1 In addition, for each economy, For example, sometimes cases were discovered where one the database reports total trade with the world. Trade data source reported figures with a different order of mag- flows are reported in US$ millions for 1981–2010, with nitude compared to the other sources. Through exten- later years having greater accuracy. Table B.1 describes the sive examination of the data these cases have been traced variables. The data includes more than 20 economic activ- and corrected. In many cases mistakes had already been ities according to BOP classification, as listed in table B.2. cleaned up by the UN. Note that not all sectors have the same coverage regard- A further verification has been made across sectors. ing time and economies. Generally, a more disaggregated Each individual subsector has been scanned and com- level of trade flows will have fewer observations avail- pared with the reported aggregate flow within that sector. able. Despite all these corrections, some bilateral data on For example, when a higher value of the aggregate figure flows remain unallocated. This unmatched information is Table B.2. Services Sector Components, by Balance of Payments (BOP) Code and Chapter BOP code Services sector BOP chapter 200 Total services – 205 Transportation services 1 206 Sea transport 1.1 207 Passenger 1.1.1 208 Freight 1.1.2 209 Other 1.1.3 210 Air transport 1.2 211 Passenger 1.2.1 212 Freight 1.2.2 213 Other 1.2.3 (continued on next page) Appendix B: Trade in Services Database 137 Table B.2. (continued) BOP code Services sector BOP chapter 214 Other transport 1.3 215 Passenger 1.3.1 216 Freight 1.3.2 217 Other 1.3.3 218 Space transport 1.4 219 Rail transport 1.5 220 Passenger 1.5.1 221 Freight 1.5.2 222 Other 1.5.3 223 Road transport 1.6 224 Passenger 1.6.1 225 Freight 1.6.2 226 Other 1.6.3 227 Inland waterway transport 1.7 228 Passenger 1.7.1 229 Freight 1.7.2 230 Other 1.7.3 Pipeline transport and electricity 231 transmission 1.8 Other supporting and auxiliary 232 transport services 1.9 236 Travel 2 237 Business travel 2.1 Expenditure by seasonal and 238 border workers 2.1.1 239 Other 2.1.2 240 Personal travel 2.2 241 Health-related expenditures 2.2.1 242 Educated-related expenditures 2.2.2 243 Other 2.2.3 245 Communication services 3 246 Postal and courier services 3.1 958 Postal services 3.1.1 959 Courier services 3.1.2 247 Telecommunications services 3.2 249 Construction services 4 250 Construction abroad 4.1 Construction in the compiling 251 economy 4.2 253 Insurance services 5 Life insurance services and 254 pension funding 5.1 255 Freight insurance 5.2 256 Other direct insurance 5.3 257 Reinsurance 5.4 258 Auxiliary services 5.5 260 Financial services 6 Computer and information 262 services 7 (continued on next page) 138 Valuing Services in Trade Table B.2. (continued) BOP code Services sector BOP chapter 263 Computer services 7.1 264 Information services 7.2 889 New agency services 7.2.1 890 Other information services 7.2.2 266 Royalties and license fees 8 891 Franchises and similar rights 8.1 892 Other royalties and license fees 8.2 268 Other business services 9 Merchanting and other trade- 269 related services 9.1 270 Merchanting 9.1.1 271 Other trade-related services 9.1.2 272 Operational leasing services 9.2 Misc. business, professional, and 273 technical services 9.3 Legal, accounting, management 274 consulting, and public relations 9.3.1 275 Legal services 9.3.1.1 Accounting, auditing, bookkeeping, and tax consulting 276 services 9.3.1.2 277 Business and management 9.3.1.3 278 Advertising and market research 9.3.2 279 Research and development 9.3.3 Architectural, engineering, and 280 other technical services 9.3.4 Agricultural, mining, and on-site 281 processing services 9.3.5 282 Waste treatment and depollution 9.3.5.1 Agricultural, mining, and on-site 283 processing services 9.3.5.2 284 Other business services 9.3.6 Services between related 285 enterprises, n.i.e. 9.3.7 Personal, cultural and recreational 287 services 10 288 Audiovisual and related services 10.1 Other personal, cultural, and 289 recreational services 10.2 895 Educational services 10.2.1 896 Health services 10.2.2 897 Other 10.2.3 291 Government services, n.i.e. 11 292 Embassies and consulates 11.1 293 Military units and agencies 11.2 294 Other government services 11.3 Other services, total (category 981 3–11) – Commercial services, total 982 (category 3–10) – 983 Services not allocated – Source: IMF, www.imf.org/external/np/sta/bopcode/topical.htm. Note: n.i.e = not included elsewhere. Appendix B: Trade in Services Database 139 Table B.3. Economies Covered in the Trade in Services Database Economy ISO3 Economy ISO3 Economy ISO3 Economy ISO3 Afghanistan AFG Denmark DNK Kuwait KWT Qatar QAT Albania ALB Djibouti DJI Kyrgyz Republic KGZ Romania ROU Algeria DZA Dominica DMA Lao PDR LAO Russian Federation RUS Andorra AND Dominican Republic DOM Latvia LVA Rwanda RWA Angola AGO Ecuador ECU Lebanon LBN Samoa WSM Antigua and Barbuda ATG Egypt, Arab Republic of EGY Lesotho LSO San Marino SMR Argentina ARG El Salvador SLV Liberia LBR São Tomé and Príncipe STP Armenia ARM Equatorial Guinea GNQ Libya LBY Saudi Arabia SAU Aruba ABW Eritrea ERI Liechtenstein LIE Senegal SEN Australia AUS Estonia EST Lithuania LTU Serbia YUG Austria AUT Ethiopia ETH Luxembourg LUX Seychelles SYC Azerbaijan AZE Fiji FJI Macao SAR, China MAC Sierra Leone SLE Bahamas, The BHS Finland FIN Macedonia, FYR MKD Singapore SGP Bahrain BHR France FRA Madagascar MDG Slovak Republic SVK Bangladesh BGD French Polynesia PYF Malawi MWI Slovenia SVN Barbados BRB Gabon GAB Malaysia MYS South Africa ZAF Belarus BLR Gambia, The GMB Maldives MDV Spain ESP Belgium BEL Georgia GEO Mali MLI Sri Lanka LKA Belize BLZ Germany DEU Malta MLT Sudan SDN Benelux BLX Ghana GHA Mauritania MRT Suriname SUR Benin BEN Gibraltar GIB Mauritius MUS Swaziland SWZ Bhutan BTN Greece GRC Mexico MEX Sweden SWE Bolivia BOL Greenland GRL Micronesia, Fed. Sts. FSM Switzerland CHE Bosnia and Herzegovina BIH Grenada GRD Moldova MDA Syrian Arab Republic SYR Botswana BWA Guam GUM Mongolia MNG Taiwan, China TWN Brazil BRA Guatemala GTM Montenegro MNE Tajikistan TJK Brunei Darussalam BRN Guinea GIN Morocco MAR Tanzania TZA Bulgaria BGR Guinea-Bissau GNB Mozambique MOZ Thailand THA Burkina Faso BFA Guyana GUY Myanmar MMR Timor-Leste TLS Burundi BDI Haiti HTI Namibia NAM Togo TGO Cabo Verde CPV Honduras HND Nepal NPL Tonga TON Cambodia KHM Hong Kong SAR, China HKG Netherlands NLD Trinidad and Tobago TTO Cameroon CMR Hungary HUN New Caledonia NCL Tunisia TUN Canada CAN Iceland ISL New Zealand NZL Turkey TUR Cayman Islands CYM India IND Nicaragua NIC Turkmenistan TKM Central African Republic CAF Indonesia IDN Niger NER Uganda UGA Chad TCD Iran, Islamic Republic of IRN Nigeria NGA Ukraine UKR Chile CHL Iraq IRQ Norway NOR United Arab Emirates ARE China CHN Ireland IRL Oman OMN United Kingdom GBR Colombia COL Isle of Man IMN Pakistan PAK United States of America USA Comoros COM Israel ISR Palau PLW Uruguay URY Congo, Dem. Rep. COD Italy ITA Panama PAN Uzbekistan UZB Congo, Rep. COG Jamaica JAM Papua New Guinea PNG Vanuatu VUT Costa Rica CRI Japan JPN Paraguay PRY Venezuela, RB VEN Côte d’Ivoire CIV Jordan JOR Peru PER Vietnam VNM Croatia HRV Kazakhstan KAZ Philippines PHL World WLD Cuba CUB Kenya KEN Poland POL Yemen, Republic of YEM Cyprus CYP Kiribati KIR Portugal PRT Zambia ZMB Czech Republic CZE Korea, Republic of KOR Puerto Rico PRI Zimbabwe ZWE obtained after carefully mapping all bilateral trade flows addition of XWD, bilateral and aggregate trade flows are against total flows. After comparing total identified bilat- internally consistent. Note that where the bilateral sums eral flows with respect to reported total flows, an extra exceeded reported totals, no XWD residual was added. reporter category was created to cover this type of undes- This is usually the case if no reported totals were avail- ignated trade flow: XWD or rest of the world. With the able, but only bilateral flows (including the ones obtained 140 Valuing Services in Trade through mirror techniques). XWD is created on both the OECD, or UN) reported zero flows rather than missing. Some bilateral combinations are unreported, which means that they do not appear in importer as well on the exporter sides. the dataset; these observations can be safely assumed to be missing and a reported source of these flows could not be found in any of the underlying databases. Note 1. This database contains 7 variables and a total of 7,117,775 observa- tions. Missing values mean that no source for an observation was avail- able. Zero value means that at least one primary source (such as Eurostat, ECO-AUDIT Environmental Benefits Statement The World Bank is committed to preserv- Saved: ing endangered forests and natural resources. • 9 trees Valuing Services in Trade: A Toolkit for • 4 million British thermal Competitiveness Diagnostics was printed on units of total energy recycled paper with 100 percent postconsumer • 809 pounds of net fiber in accordance with the recommended greenhouse gases standards for paper usage set by the Green Press (CO2 equivalent) Initiative, a nonprofit program supporting pub- • 4,387 gallons of waste lishers in using fiber that is not sourced from water endangered forests. For more information, visit • 293 pounds of solid waste www.greenpressinitiative.org.