C H I N A A N D T H E WTO ACCESSION, POLICY REFORM, AND POVERTY REDUCTION STRATEGIES Editors Deepak Bhattasali · Shantong Li · Will Martin China and the WTO Accession, Policy Reform, and Poverty Reduction Strategies China and the WTO Accession, Policy Reform, and Poverty Reduction Strategies Editors Deepak Bhattasali, Shantong Li, and Will Martin A copublication of the World Bank and Oxford University Press © 2004 The International Bank for Reconstruction and Development / The World Bank 1818 H Street, NW Washington, DC 20433 Telephone 202-473-1000 Internet www.worldbank.org E-mail feedback@worldbank.org All rights reserved. A copublication of the World Bank and Oxford University Press. 1 2 3 4 07 06 05 04 The findings, interpretations, and conclusions expressed herein are those of the author(s) and do not necessarily reflect the views of the Board of Executive Directors of the World Bank or the governments they represent. The World Bank 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. Rights and Permissions The material in this work is copyrighted. Copying and/or transmitting portions or all of this work without permission may be a violation of applicable law. The World Bank encourages dissemination of its work and will normally grant permission promptly. For permission to photocopy or reprint any part of this work, please send a request with complete information to the Copyright Clearance Center, Inc., 222 Rosewood Drive, Danvers, MA 01923, USA, telephone 978-750-8400, fax 978-750-4470, www.copyright.com. All other queries on rights and licenses, including subsidiary rights, should be addressed to the Office of the Publisher, World Bank, 1818 H Street NW, Washington, DC 20433, USA, fax 202-522-2422, e-mail pubrights@worldbank.org. Cover collage created by Tomoko Hirata using the following photographs: © Michael S. Yamashita/CORBIS (farmer); © Enzo & Paolo Ragazzini/CORBIS (woman); © Yang Liu/CORBIS (automobile factory); © How-Man Wong/CORBIS (man). Library of Congress Cataloging-in-Publication Data China and the WTO: accession, policy reform, and poverty reduction strategies / edited by Deepak Bhattasali, Shantong Li, William J. Martin. p. cm. ­ (Trade and development series) Includes bibliographical references and index. ISBN 0-8213-5667-4 1. China--Commercial policy. 2. World Trade Organization--China. 3. Foreign trade Regulation--China. 4. Agriculture and state--China. 5. Poverty--Government policy--China. 6. China--Economic policy--2000-. I. Bhattasali, Deepak. II. Li, Shantong. III. Martin, Will, 1953- IV. Series. HF1604.C4426 2004 382'.3'0951--dc22 2004044097 Contents Foreword xi Acknowledgments xii About the Editors and Authors xiii Acronyms and Abbreviations xiv 1. IMPACTS AND POLICY IMPLICATIONS OF WTO ACCESSION FOR CHINA 1 Deepak Bhattasali, Shantong Li, and Will Martin Part I. Policy Reforms Associated with Accession 19 2. WHAT CHINA'S WTO ACCESSION IS ALL ABOUT 21 Jeffrey L. Gertler 3. CHINA IN THE WTO: ANTIDUMPING AND SAFEGUARDS 29 Patrick A. Messerlin 4. INTELLECTUAL PROPERTY RIGHTS IN THE WTO ACCESSION PACKAGE: ASSESSING CHINA'S REFORMS 49 Keith E. Maskus 5. CHINA'S AGRICULTURAL DEVELOPMENT AND POLICY READJUSTMENT AFTER ITS WTO ACCESSION 69 Chen Xiwen 6. THE NATURE OF DISTORTIONS TO AGRICULTURAL INCENTIVES IN CHINA AND IMPLICATIONS OF WTO ACCESSION 81 Jikun Huang, Scott Rozelle, and Min Chang Part II. Economic Impacts of Accession 99 7. THE IMPACTS OF WTO ACCESSION ON CHINESE AGRICULTURE AND RURAL POVERTY 101 Kym Anderson, Jikun Huang, and Elena Ianchovichina 8. THE SERVICES DIMENSION OF CHINA'S ACCESSION TO THE WTO 117 Aaditya Mattoo 9. LOGISTICS IN CHINA: IMPLICATIONS OF ACCESSION TO THE WTO 141 Wenping Luo and Christopher Findlay v vi Contents 10. TELECOMMUNICATIONS SERVICES IN CHINA: FACING THE CHALLENGES OF WTO ACCESSION 157 Mari Pangestu and Debbie Mrongowius 11. ACCELERATING FINANCIAL MARKET RESTRUCTURING IN CHINA 181 Deepak Bhattasali 12. WTO ACCESSION AND THE STRUCTURE OF CHINA'S MOTOR VEHICLE SECTOR 191 Joseph F. Francois and Dean Spinanger 13. ECONOMIC IMPACTS OF CHINA'S ACCESSION TO THE WTO 211 Elena Ianchovichina and Will Martin Part III. Impacts on Households and on Poverty 237 14. EARNINGS AND LABOR MOBILITY IN RURAL CHINA: IMPLICATIONS FOR CHINA'S ACCESSION TO THE WTO 239 Terry Sicular and Yaohui Zhao 15. WELFARE IMPACTS OF CHINA'S ACCESSION TO THE WTO 261 Shaohua Chen and Martin Ravallion 16. IMPLICATIONS OF WTO ACCESSION FOR POVERTY IN CHINA 283 Thomas W. Hertel, Fan Zhai, and Zhi Wang 17. COPING WITH AND ADAPTING TO JOB LOSSES AND DECLINES IN FARM EARNINGS IN CHINA 305 Athar Hussain Index 325 Boxes 9.1 Local Policies and Their Implications for Logistics Operations 148 10.1 Summary of China's WTO Commitments 166 10.2 Coverage and Definition of China's Telecommunications Sector 167 10.3 Experience of Chile, Peru, South Africa, and Hong Kong (China) with Universal Access 175 11.1 Modeling State Bank Performance: Can the State Banks Survive Liberalization? 188 12.1 Summary of Developments in China's Automotive Sector 193 17.1 Advantages and Disadvantages of Decentralization 312 Figures 8.1 Country Services Commitments by Mode 119 8.2 Sequence of Telecommunications Reform in Nine Asian Countries, 1989­99 135 9.1 Choice of Transport Modes: China, 1970­2000 143 Contents vii 10.1 Mobile and Fixed-Line Use in China, 1996­2003 163 10.2 Internet Users in China, 1996­2002 164 10.3 Restructuring of China's Telecommunications Industry, 1994­2003 170 10.4 Share of Revenue by Operator, 2001 and 2002 172 12.1 China's Production of Motor Vehicles since Adoption of the Open-Door Policy 194 12.2 Scale and Average Cost (with a CDR of 0.125) 201 12.3 Motor Vehicle Production in China and United States 201 12.4 Import Share of Total Auto Parts 203 12.5 Location of Automobile Production in China 208 15.1 Poverty Incidence Curves, Rural 268 15.2 Poverty Incidence Curves, Urban 269 15.3 Mean Gains by Provinces: Absolute Gain in Yuan Per Capita 270 15.4 Mean Gains by Provinces: Proportionate Gains in Percent 270 15.5 Mean Gains by Provinces: Percentage of Gainers by Provinces 271 15.6 Mean Gains in Yuan by Income Percentile 271 15.7 Mean Percentile Gain by Income Percentile 272 15.8 Percentage of Gainers by Income Percentile 272 16.1 Change in Sector Output due to WTO Accession, China 298 16.2 Impact of WTO Accession on Rural Households 300 16.3 Impact of WTO Accession on Urban Households 300 Tables 1.1 Some Measures of Import Protection in China's Agricultural Sector 6 3.1 Top 10 Antidumping Users, Measures in Force, 1995­2001 30 3.2 Major Antidumping Targets, Measures in Force, 1995­2001 31 3.3 Shares of Antidumping Measures in Force against Imports from China, 1995­2002 32 3.4 Antidumping Measures in Force in the World by Sector, 1995­2000 34 3.5 EC and U.S. Echoing Antidumping Cases against Chinese Producers, 1980­99 37 3.6 China's Antidumping Measures in Force and Investigations, as of 30 June 2003 40 3.7 Do Antidumping Investigations Really Look at Dumping? 42 4.1 Substantive Requirements of the TRIPS Agreement in the WTO 51 4.2 An Assessment of China's IPRs System for Development Purposes 58 4.3 Patent Applications by Type and Nationality, 1994­2000 62 4.4 Patent Grants by Type and Nationality, 1994­2000 63 4.5 Bilateral Invention Patent Applications, 1994­98 63 4.6 Bilateral Invention Patent Grants, 1994­98 64 4.7 Patenting Indicators for Top Patenting Regions, 1985­96 and 2000 64 5.1 Changes in Farm Output Value Shares 70 5.2 Sources of Farmers' Incomes, 1997­2000 71 5.3 Declines in Per Capita Farm Incomes in Six Key Provinces 72 5.4 Changes in Farm Household Spending, 1997­2000 72 5.5 Share of Farmers in Consumer Spending at County and Lower Levels 73 5.6 Changes in Urban and Rural Resident Income and Spending, 1996­2001 75 6.1 Changes in Nominal Rates of Protection Over Time of China's Major Agricultural Commodities, 1978­2000 83 6.2 China's Agricultural Trade (US$ millions) by Factor Intensity, 1985­97 83 6.3 Subsidies and Tax Rebates for Exports of Selected Agricultural Commodities in China, 2001 85 viii Contents 6.4 Disaggregated Nominal Rates of Protection for Selected Grains in China, October 2001 87 6.5 Average Nominal Protection Rates for Major Imports and Exports in China, October 2001 88 6.6 Percentage of Market Pairs that Test Positive for Being Integrated Based on Dickey-Fuller Test in Rural China, 1988 to 2000 89 6.7 Soybean, Corn, and Wheat Village Price Regression, 2000 93 6.8 Transmission Coefficients for Rice and Maize Measuring the Percentage of Price Shock at Port that Is Transmitted to Inland Markets in China, 1996­2000 93 7.1 Growth Rates of China's Economy, 1970­2000 103 7.2 Farm and Rural Enterprise (RE) Developments in China, 1980­99 103 7.3 Changing Structure of China's Economy, 1970­2000 104 7.4 China's Grain Production, Trade, and Consumption, 1980­2000 104 7.5 Structure of China's Food and Feed Trade, 1980­99 105 7.6 Nominal Rates of Protection (NRP, Applied Tariff or Tariff Equivalent), Agricultural Products: China, 1995­2007 106 7.7 Rural Income, Employment, and Food Output by Region: China, 2000 108 7.8 Income and Its Distribution: Rural and Urban China, 1980­99 108 7.9 Sectoral Volume Effects of China's WTO Accession Reforms (Core Case), 2002­07 110 7.10 Change in China's Real Factor Prices and National Economic Welfare Attributable to WTO Accession, 2001­07 111 7.11 Changes in China's Farm Output, Employment, and Trade Volumes: Without WTO Accession, 1995­2007, and Resulting from China's WTO Accession Reforms, 2002­07 113 7.12 Effect on Agricultural Self-Sufficiency of China's WTO Accession without and with Cuts in Agricultural Protection in OECD Countries 114 8.1 Coverage of Specific Commitments by China to Services in Trade 120 8.2 China's Commitments: Past, Present, and Future 122 9.1 Starting Position of China's Logistics Sector 145 9.2 Position of China's Logistics Sector One Year after WTO Accession 146 9.3 Position of China's Logistics Sector Three Years after WTO Accession 146 9.4 Summary of Indirect Effects of WTO Accession on China's Logistics Sector 147 9.5 Logistics Costs as a Share of Total Revenue, Europe and Asia, 2001 149 10.1 Network Operators and Licensing, China 159 10.2 Network Operators in China: Description of Network and Coverage 160 10.3 Impacts of Reforms on China's Telecommunications Industry, 1998­2003 165 11.1 China's Financial Services Industry, 2001 183 11.2 China's WTO Commitments in the Financial Services Sector 186 12.1 GNP, Population, and Stocks of Cars: Selected Countries, 2000 192 12.2 Auto Production by Plant in China, 1995­2002 195 12.3 Import and Domestic Shipment Values of China's Motor Vehicle Industry, 1997 196 12.4 Tariffs on Chinese Motor Vehicles 196 12.5 Regional and Sectoral Breakdown of the Model 198 12.6 Mainland China's Pre- and Post-WTO Accession Tariff Rates 199 12.7 Impact of WTO Accession of Greater China on Output 202 12.8 Mainland China Motor Vehicle Market 203 12.9 Impact of Greater China Accession on Value Added 204 12.10 Impact of Greater China Accession on Value of Exports 205 12.11 Impact of Greater China Accession on Value of Exports 206 Contents ix 12.12 China Export Shares: Baseline and Scenario 207 13.1 Import Coverage of Nontariff Barriers 214 13.2 China's Average Statutory Rates 215 13.3 China's Import Protection before and after WTO Accession 216 13.4 Projected Growth in Factor Inputs and Total Factor Productivity 1997­2007 218 13.5 Changes in China's Key Economic Indicators after 2001 as a Result of WTO Accession 220 13.6 Change in Real Factor Prices in China as a Result of Accession, 2001­07 221 13.7 Welfare Change and Sources of Welfare Change as a Result of China's WTO Accession 222 13.8 Change in Output and Employment in China as a Result of WTO Accession and Other Reforms 225 13.A.1 Tariff Lines Subject to Import NTBs, China, 2001 227 13.A.2 The Import Coverage of Nontariff Barriers in China, 2001 228 13.A.3 Commodity Import Shares by NTB Measure 229 13.A.4 Elasticity of Substitution and Changes in Real Consumer Prices due to China's WTO Accession 230 13.A.5 Changes in China's Key Economic Indicators in the Baseline, without WTO Accession, 1995­2007 231 13.A.6 Changes in China's Key Economic Indicators due to WTO Accession for the Period before 2001 232 13.A.7 Sensitivity Analysis with Respect to the Elasticity of Transformation (2.67) 233 14.1 Descriptive Statistics, CHNS, 1997 244 14.2 Income Generation Functions 245 14.3 Descriptive Statistics for Variables Used in the Income Generation Functions, CHNS, 1997 246 14.4 Decomposition of Household Net Earnings Differentials between Households Belonging to the Top and Bottom 20 Percent 248 14.5 Agricultural Production Function 251 14.6 Nonagricultural Production Function 252 14.7 Shadow Wage Equations for Household Agricultural and Nonagricultural Production 253 14.8 Employed Wage Equation 254 14.9 Summary Statistics for Wage Rates and Shadow Wages 255 14.10 Household Labor Supply Equations 257 15.1 Predicted Price Changes from GTAP Model and Per Capita Net Gain or Loss for Rural and Urban Households, 1995­2001 266 15.2 Predicted Price Changes from GTAP Model and Per Capita Net Gain or Loss for Rural and Urban Households, 2001­07 267 15.3 Summary Statistics on Aggregate Welfare Impacts, 1995­2001 and 2001­07 268 15.4 Regression Results for Level of Gain (Yuan) in Rural Areas of Three Provinces, 2001­07 274 15.5 Regression Results for Percentage Gains in Rural Areas of Three Provinces, 2001­07 275 15.6 Regression Results for Level of Gain (Yuan) in Urban Areas of Three Provinces, 2001­07 276 15.7 Regression Results for Percentage Gains in Urban Areas of Three Provinces, 2001­07 277 15.8 Average Impacts for Agriculture Households in Selected Countries, 2001­07 278 15.A.1 Rural Gains and Losses by Province, 2001­07 280 15.A.2 Urban Gains and Losses by Province, 2001­07 281 16.1 Per Capita Income by Location, Stratum, and Vingtile 285 16.2 Economic Structure and Market Openness in China, 1997 287 16.3 Summary of Baseline Calibration 293 16.4 Summary of Experiments 294 16.5 Implications of China's WTO Accession, 2007 296 x Contents 16.6 Sector Impacts of China's WTO Accession in 2007: Percentage Deviation from Baseline 297 16.7 Household Impacts of China's WTO Accession, 2007 299 17.1 Classification of Schemes and Measures 308 17.2 Differences between Social Insurance and Social Assistance 310 17.3 Income Maintenance Schemes in China 311 Foreword China's era of reform has seen enormous progress number of training courses undertaken by the in reducing poverty at the same time as increasing World Bank Institute and partner organizations in integration with the world economy. The process of different parts of China. This book will provide a accession to the WTO contributed to a decisive lasting reference to this major step in the evolution deepening of the process of integration with the of the Chinese economy. world and a move to a much stronger emphasis on This book is the result of a number of years of rules-based reform. However, in recent years, there work by a team of Chinese and foreign scholars. has been increasing concern about an increase in Just as China and her trading partners have bene- inequality within China and a rising gap between fited from the win-win nature of international the fortunes of urban and rural families. trade, members of this team have benefited from The aim of the project on which this book is one another's knowledge and understanding of based was to bring together information on the different facets of this enormously complex and implications of China's WTO accession agreement-- dynamic economy. The work is testimony to the an enormous and complex watershed agreement-- success of the partnership between the World Bank and to assess the overall impacts for households and the Development Research Centre of State throughout China. This required the collection of Council--a partnership that has contributed a information and the analysis of the policy reforms great deal to both of our organizations. We look involved in the accession agreement, followed by an forward to even stronger cooperation between our assessment of the effects of these policies on the two organizations in the future, as China enters the major elements of the economy. Finally, drawing on next phase of development. this analysis, the impact of the reforms on individual households--and especially those of the poor--was François Bourguignon evaluated. The results of the analysis provide a basis Senior Vice President and Chief Economist for recommending policies to capitalize on the bene- The World Bank fits resulting from accession and to mitigate the adverse impacts. The work on which this book is based has been Wang Mengkui disseminated in China through numerous presen- Director tations and seminars and through inputs into a Development Research Centre of State Council xi Acknowledgments We owe a debt of gratitude to the United Kingdom organizing the many workshops and interim publi- Department for International Development (DFID) cations associated with this project. for its generous support to this project. Particular Many others have assisted by commenting on thanks are due to the DFID staff who provided sup- drafts of this book, through participation in the port and encouragement at critical times in the life workshops, or by providing advice. Among others, of the project--in particular, Paul Whittingham for we would like to thank Long Yongtu, François his encouragement in the original design; Catherine Bourguignon, Christopher Hum, T. N. Srinivasan, Martin for her continuing advice and encourage- Kong-Yam Tan, L. Alan Winters, Justin Yifu Lin, ment from Beijing; and Fiona McConnon, Clare Wang Xinkui, Nicholas Lardy, Ke Binsheng, Wang Roberts, and Kebur Azbaha for their assistance with Jun, Bernard Hoekman, Liu Xiaofan, Xu Lin, Wang the implementation of the project. Yan, Lu Feng, Yu Jianhua, Sarah Cook, Yang In addition to our authors--from the Develop- Guohua, Yu An, Sheng Laiyun, Wang Pingping, ment Research Center (DRC) of the State Council Wang Yongjuan, Huang Langhui, Li Yan, Chen of China, the World Trade Organization, the World Xiaodong, Zhang Xiangchen, Wang Kan, Nicholas Bank, and other institutions around the world--we Bridge, Gillian Cull, Yukon Huang, Charles Piggott, would like to thank numerous staff within the Jeffrey Liang, Li Shi, Edwin A.L.M. Vermulst, and DRC, especially Sun Zhiyan and Liu Peilin, who Jianguang Shen for this help. assisted this project at every step, including the final We are very grateful to Mary Fisk for her excel- dissemination workshop in Wuhan. lent editorial assistance and to Santiago Pombo- At the World Bank, we would like to thank Bejarano for his enthusiastic support in converting Jianqing Chen, Leona Luo, and Rebecca Martin for the manuscript into a finished volume. We would the superb support and assistance they have pro- also like to thank the anonymous reviewers, who vided throughout the life of the project. The assis- provided excellent comments on the manuscript; tance of the Beijing Office was invaluable also in may their tribe increase! Deepak Bhattasali, Shantong Li, Will Martin xii About the Editors and authors Editors Athar Hussain, Department of Economics, London Deepak Bhattasali, Lead Economist, Poverty School of Economics, London Reduction and Economic Management Sector Elena Ianchovichina, Economist, Poverty Reduc- Development, World Bank, Beijing tion and Economic Management, World Bank, Shantong Li, Director, Department of Develop- Washington, D.C. ment Strategy and Regional Economy, Develop- Wenping Luo, Professor, Shanghai Maritime Uni- ment Research Center of the State Council, P.R. versity, Shanghai China, Beijing Keith E. Maskus, Professor, Department of Eco- Will Martin, Lead Economist, Development nomics, University of Colorado, Boulder Research Group, World Bank, Washington, D.C. Aaditya Mattoo, Lead Economist, Development Research Group, World Bank, Washington, D.C. Authors PatrickA.Messerlin,Professor of Economics, Insti- Kym Anderson, Centre for Economic Studies, Uni- tutes d'Etudes Politiques, Groupe d'Economie versity of Adelaide, Adelaide, Australia Mondiale de Sciences Po, Paris Min Chang, Graduate Student, Department of Debbie Mrongowius, Independent Researcher, Agricultural and Resource Economics, Univer- Shanghai sity of California, Davis Mari Pangestu, Director, Centre for Strategic and Shaohua Chen, Senior Information Officer, International Studies, Jakarta Development Research Group, World Bank, Martin Ravallion, Research Manager, Develop- Washington, D.C. ment Research Group, World Bank, Washington, Christopher Findlay, Professor, Asia-Pacific School D.C. of Economics and Government, The Australian Scott Rozelle, Professor, Department of Agricul- National University, Canberra tural and Resource Economics, University of Joseph F. Francois, Professor, Tinbergen Institute, California, Davis Rotterdam, and Fellow, Centre for Economic Terry Sicular, Associate Professor of Economics, Policy Research, London University of Western Ontario, London, Ontario Jeffrey L. Gertler, Legal Affairs Division and for- Dean Spinanger, Professor, Institute for World mer Secretary, Working Party on the Accession Economics, Kiel University, Kiel of China, World Trade Organization, Geneva Zhi Wang, Economic Research Service, United Thomas W. Hertel, Distinguished Professor, States Department of Agriculture, Washington, Department of Agricultural Economics, and D.C. Executive Director, Center for Global Trade Chen Xiwen,Vice President (Vice Minister), Devel- Analysis, Purdue University, West Lafayette, opment Research Center of the State Council, Indiana Beijing Jikun Huang, Director and Professor, Center for Fan Zhai, Department of Policy and Fiscal Affairs, Chinese Agricultural Policy Institute for Ministry of Finance, Beijing Geographical Sciences and Natural Resource Yaohui Zhao, Professor of Economics, Beijing Uni- Research, Chinese Academy of Sciences, Beijing versity, Beijing xiii Acronyms and Abbreviations ADSL asymmetric digital subscriber line NME non-market economy AQSIQ Administration for Quality NRP nominal rate of protection Supervision, Inspection and OECD Organisation for Economic Quarantine Co-operation and Development CAAC Civil Aviation Administration of PBC People's Bank of China China PCT Patent Cooperation Treaty CGA Customs General Administration PTA provincial telecommunications CGE Computable general equilibrium authority CHNS China Health and Nutrition Survey PTNR permanent normal trade relations CIRC China Insurance Regulatory SARFT State Administration of Radio, Film Commission and Television CJV contractual joint venture SCIO State Council Information Office CPI consumer price index SDPC State Development Planning CSRC Chinese Security Regulatory Commission Commission SETC State Economic and Trade DDA Doha Development Agenda Commission EC European Community SEZ special economic zone EJV equity joint venture SILG State Information Leading Group FDI foreign direct investment TCSC Tariff Commission under the State GATS General Agreement on Trade in Council Services TDSCDMA time division synchronous code GATT General Agreement on Tariffs and division multiple access Trade 3PL third-party logistics GDP gross domestic product TPS transitional product-specific GSM global system for mobile safeguard communications TRIPS Trade-Related Aspects of GTAP Global Trade Analysis Project Intellectual Property Rights HTS Harmonized Tariff System TRQ tariff-rate quota IP Internet protocol TVE township and village enterprises IPR intellectual property rights URAA Uruguay Round Agreement on ISDN integrated services digital network Agriculture MEI Ministry of Electronics Industry VAT value-added tax MFA Multi-Fiber Arrangement VoIP Voice over Internet Protocol MFN most favored nation WCDMA wideband code division multiple MII Ministry of Information Industry access MOFTEC Ministry of Foreign Trade and WIPO World Intellectual Property Economic Cooperation Organization MPT Ministry of Posts and WTO World Trade Organization Telecommunications xiv 1 impacts and policy implications of wto accession for china Deepak Bhattasali, Shantong Li, and Will Martin After 15 arduous years as a candidate, China's There is every indication that China's policy- application for membership in the World Trade makers see China's WTO accession agreement as Organization (WTO) was accepted at the WTO's a means to fulfilling broader goals. One broader Doha ministerial meeting in November 2001. As strategic goal is to facilitate the peaceful emergence expected by observers such as Panitchpakdi and of China as a great trading nation--and to avoid Clifford (2002), China has become an active and the trade tensions associated with the emergence of prominent member of the WTO, and has sharply major new traders in the past. Another goal is to increased its share of world trade since joining. accelerate the process of domestic reform--with Domestically, however, China faces many chal- the WTO acting, in the colorful words of Jin lenges, particularly in dealing with the widening gap (2002), as a wrecking ball for what remains of the between urban and rural incomes. The objective of earlier closed economy (Jin 2002). The process of this study is to analyze the impacts of China's acces- accession has also unleashed a massive process of sion to the WTO and the policy reforms needed to reforming the laws covering trade issues (WTO help facilitate China's economic development. 2002), with the result that many reforms go beyond Of the many approaches that might be used to those narrowly required by the WTO accession analyze China's accession to the WTO, an important agreement. In addition, China's policymakers have one deals with the legal rights and responsibilities-- also clearly recognized the need to deal with the that is, the challenges involved in meeting China's yawning gap between urban and rural incomes--a legal commitments and in ensuring that China's gap that has become more evident since China's rights are maintained. Another focuses on the trade accession to the WTO (Wen 2004). and policy changes required to open China's econ- The importance of China's accession to the omy and integrate it into the global economy. Yet WTO has been widely recognized and studied. Both even though these are important, they are not a the China Quarterly (see Fewsmith 2001) and complete characterization of the implications of China Economic Review (see Chun, Fleisher, and accession. For us, perhaps the most compelling Parker 2001) have published special issues on the aspect is one identified by Woo (2001): WTO acces- topic, and it has been the subject of books, includ- sion as a key component in the restructuring of the ing those by Panitchpakdi and Clifford (2002) and Chinese economy. Lardy (2002). Other investigators have undertaken 1 2 China and the WTO quantitative studies of the effects of trade liberal- balance between creating incentives for innovation ization on China (McKibbin and Tang 2000; and restricting competition by giving monopoly Ianchovichina and Martin 2001; Gilbert and Wahl power to the inventor. The TRIPS agreement is dis- 2002). Some studies, including those by Martin and cussed at the end of this section. Ianchovichina (2001) and Shafaeddin (2004) have The general principle of nondiscrimination looked into the implications of China's WTO requires that WTO members give equal treatment accession for other countries and for the WTO to competing suppliers (the most-favored-nation system itself. The studies presented in this volume principle), and that they not discriminate between attempt to go beyond previous studies in examin- domestically produced and imported goods or ing, within an integrated framework, China's policy services in their internal markets. The application reforms and their implications for the Chinese of this general principle to China involved some economy and for individual households. additional commitments, including the elimination This volume is divided into three parts. The first of dual pricing systems, the phasing out of restric- part contains studies that cover the most important tions on trading, and the introduction of more uni- policy reforms associated with accession to the form administrative arrangements and of judicial WTO. The second part consists of studies that review. These commitments are important not just examine the consequences of reform for the econ- for the central authorities, but also for the lower omy and for particular sectors. The studies in the tiers of government that deal with internal trade third part build on the earlier studies and examine and regulation. the implications of reform for households, and In China's accession, the market opening principle particularly for those households vulnerable to had three major dimensions: (1) commitments by poverty. Only with the insights available from these China to abolish nontariff barriers, reduce tariffs, three approaches is it possible to provide sound and open service sectors; (2) commitments by the policy recommendations. In the remainder of this importing countries to abolish the quotas on textiles chapter, we review and summarize some of the key and clothing originally imposed under the Multi- findings emerging from the entire set of studies. Fiber Arrangement (MFA); and (3) agreement by the United States and other countries to impose MFN tariffs on China. Analysis in chapter 13 by China's Policies and the Reforms Elena Ianchovichina and Will Martin suggests that Associated with Accession the tariff cuts offered by China were very substantial, In chapter 2 of this volume, Jeffrey L. Gertler de- and will result in a reduction in the weighted average scribes the long process that culminated in China's tariff from 12 percent in 2001 to 6.8 percent at the recent accession to the WTO. As in other acces- end of the implementation period.1 However, these sions, the process involved bilateral negotiations on reductions are quite small relative to the reduction issues of market access and a multilateral process in of 29 percentage points between 1992 and 2001. The which the accession agreement was formulated and analysis reported by Aaditya Mattoo in chapter 8 the best market access given to any one member suggests that China's commitments to liberalize was extended to all other members under the most- trade in services are extremely large relative to those favored-nation (MFN) rule. of almost all other countries.Indeed,Mattoo goes so The five basic principles of the General Agree- far as to say that China's accession commitments on ment on Tariffs and Trade (GATT) and the WTO services constitute the most radical reform of trade outlined by Gertler are useful in organizing our in services ever undertaken under the WTO. He assessment of the implications of China's accession. notes, however, that China's commitments were The five principles are: (1) nondiscrimination, more frequently subject to qualifications or reserva- (2) market opening, (3) transparency and pre- tions than those of other countries. Details of the dictability, (4) undistorted trade, and (5) prefer- liberalization in specific sectors are considered in ential treatment for developing countries. The the next section, as a prelude to assessing the likely agreement on Trade-Related Aspects of Intellectual impact of accession on the Chinese economy. Property Rights (TRIPS) incorporates some of Although most of the market opening agreed in these principles as well as others, such as striking a China's accession, like in other WTO accession Impacts and Policy Implications of WTO Accession for China 3 agreements, is to be carried out by the acceding China has made more stringent commitments, country, China's trading partners made an impor- including one to refrain from using export subsi- tant "concession" in agreeing to abolish the quotas dies on agricultural goods. on textiles and clothing originally imposed under The issue of antidumping and safeguard meas- the MFA. This agreement appears likely to be ures is troubling, particularly because it concerns extremely important for China given its strong China's market access opportunities. As Patrick A. comparative advantage in textiles and clothing, Messerlin notes in chapter 3, the WTO rules on although it is qualified by the ability of the im- antidumping are biased toward finding dumping, porters to impose special textile and clothing safe- even where no economically meaningful dump- guards for one year at a time during a transition ing exists. The situation is worse for China than period up to 2008. Another important "concession" for other WTO members, because 70 percent of by almost all existing members of the WTO was to China's exports are products most vulnerable to refrain from invoking nonapplication provisions antidumping measures. Furthermore, China will of the type widely invoked against Japan when it potentially remain vulnerable, for up to 15 years, to joined the GATT. As a result, China has received highly discriminatory non­market economy provi- most-favored-nation treatment in virtually all mar- sions. These provisions dramatically increase the kets. Clearly, this status is important because it frees probability that dumping will be found. And when China from onerous, one-sided review procedures,2 antidumping duties are applied, they are generally such as the former annual review of MFN in the much higher than when market economy provi- United States, and increases the confidence of sions are used--in the United States, the average investors in China's export industries that foreign 40 percent duty applied using nonmarket economy markets will be available to them on a continuing provisions was more than 10 times as high when basis. normal approaches were used. The transparency and predictability of trade pol- A particularly worrying feature of the accession icy are enhanced through both the general WTO agreement is the product-specific transitional safe- policy rules, such as the need to publish trade rules guard provisions that may be applied by any WTO and regulations, and some of the specific commit- Member, and may then trigger actions against ments made by China, such as those on uniform diversion of Chinese exports to other markets application of the trade regime and independent (Panitchpakdi and Clifford 2002). These provisions judicial review, as well as the provision of a mecha- are, in a sense, worse than the provisions on nism through which concerned parties can bring non­market economy treatment in that they intro- problems of local protectionism to the attention of duce an entirely new form of protection, targeted the central government.3 Another important con- specifically against China, more readily triggered tributing factor is China's binding of its entire tariff than regular safeguards and available to China's schedule for goods, almost always at tariff levels trading partners for up to 12 years from the date of below prior applied rates. The transitional review accession (Anderson and Lau 2001). The trade held each year for eight years after China's accession diversion measures allowed under this agreement provides additional information about China's are especially troublesome, because they provide trade regime and its reforms during that period. even less procedural protection than is available China's commitments to phase out restrictions under regular safeguards. on trading rights for all products except a short list Messerlin believes that China has two major of commodities that may remain subject to state choices in responding to these measures: to retaliate trading are important for increasing transparency. by, for example, increasing use of antidumping and Commitments to allow entry into distribution and contesting safeguard actions in dispute settlement; wholesale services provide further transparency as or to mount a concerted campaign for reform of a side effect of liberalization of these important the rules in the Doha negotiations, most particu- service activities. larly in the area of antidumping. Clearly, China will The WTO principle of undistorted trade involves have to contest unjustified actions through dispute disciplines on issues such as subsidies and counter- settlement. However, a retaliatory approach of vailing measures, antidumping, and safeguards. launching antidumping actions could be extremely 4 China and the WTO costly to China's economy both by reversing liberal- multilateral trade reform as a means of creating ization and by increasing the degree of uncertainty market access opportunities. about trade policy. Even though this approach is A central aspect of the WTO agreement for likely to be politically attractive, as is confirmed China is the TRIPS agreement. In response to its by the recent upsurge in antidumping actions in recognized need to stimulate innovation in China China--and the dramatic upsurge in these actions and access foreign technology and in response in developing countries such as Argentina, India, to pressure from its trading partners, China has Mexico, and South Africa--the economic costs to strengthened its intellectual property rights regime. China in particular suggest it needs be avoided as Since 1990, China has updated its laws on copy- much as possible. rights, trademarks, patents, and trade secrets, and If China elects instead to lead a push for reform adopted protection for new plant varieties and of the antidumping and safeguard rules that would integrated circuits. reduce the abuses of these protectionist measures, it Intellectual property rights involve a fine balance could greatly improve the performance of its own between creating the incentive to innovate and economy in the short run and the global trading restricting competition in the market for the goods system in the longer term. Here Messerlin provides and services that result from innovation. The bal- two specific suggestions. On non­market economy ance between these two objectives differs between treatment, he argues that China could press for new developed and developing countries, so that im- rules on the automatic granting of market economy plementing an appropriate intellectual property status in a particular commodity as long as the regime in a developing country is not simply a mat- country meets basic conditions such as low rates of ter of emulating the state-of-the-art regime of a protection, an absence of core gray-area measures, developed country. Such a regime may inhibit and an absence of state monopoly in the distribu- growth by limiting innovation and diffusion, and tion of that good. On antidumping measures more may result in excessive transfers to foreign producers generally, China could put forward, or strongly of intellectual property. Appropriate regulations are support, proposals to narrow the use of antidump- needed to ensure that markets remain competitive ing measures and to reduce their severity. China without excessive reductions in the incentive to might also seek similar relief on the product- innovate.The TRIPS agreement generally appears to specific safeguards. allow the needed flexibility, but taking advantage of Preferential treatment for developing countries such flexibility is a nontrivial task. was a particularly vexing issue throughout the In chapter 4, Keith E. Maskus assesses China's WTO accession negotiations. Although China is intellectual property regime--that is, its provisions clearly a low-income country, because of its size for patents, trademarks, trade secrets, and copy- and growth performance WTO members were rights. He bases his assessment in part on a com- reluctant to give it full developing-country treat- parison of China's, intellectual property rights ment. In many areas of the agreement, it is likely to (IPR) with benchmark standards for middle- have full access to the developing-country provi- income developing countries (World Bank 2001, sions, although in some others--such as agricul- and on interviews with market participants). He ture where it had to accept a limit of 8.5 percent on concludes that China's TRIPS regime is broadly de minimis domestic support versus the usual appropriate to China's situation. He believes, in 10 percent limit for developing countries--it faces particular, that China's policy of public procure- tighter restrictions than other developing countries. ment of pharmaceuticals at negotiated prices is At the same time, China obtained specific transi- appropriate for providing public health services. He tional arrangements in areas, such as the phasing also concluded that, with the reforms underway at out of quotas and licenses and phased entry of accession, the regime would be fully consistent with foreign enterprises, that are not generally avail- TRIPS requirements. able to developing-country Members. Special and Maskus does, however, raise some important differential treatment in the form of preferential concerns about China's policies related to TRIPS. access to industrial-country markets is not impor- For example, he believes proposals to extend tant for China, which increases the importance of patent protection to computer software--a level of Impacts and Policy Implications of WTO Accession for China 5 protection currently provided only in the United agriculture and services in China, and nontrivial in States, Japan, and Australia--may be excessive in a manufacturing. The evaluation must then take into young industry such as China's. Maskus also raises account the implications of the policy measures the serious problems of IPR enforcement, particu- being introduced. Only then can it take into account larly of trademarks, patents, and trade secrets. He the impact of these measures on economic variables believes these enforcement problems will inhibit such as output and trade levels by sector, and on the the transfer of technology in China and the devel- income levels of people, and particularly poor peo- opment of innovative domestic businesses. A ple. We first examine the impacts on agriculture, broader problem relates to China's current low then on manufacturing, then on services. allocation of resources to research and develop- ment and the therefore limited benefits of protect- Agriculture ing Chinese innovations. Important issues for the future include enhancing pricing regulations on Many authors have raised concerns about the pharmaceuticals as patent protection becomes impact of WTO accession on China's agricultural stronger and developing a broader competition sector and the many poor people engaged in this policy regime to deal with abuses of IPRs such sector. Much of this concern has been based on as monopoly pricing and restrictive licensing comparisons of China's statutory tariffs on agricul- arrangements. ture in the 1990s with the rates agreed in the In chapter 5, Chen Xiwen explores the situation accession process (e.g., Schmidhuber 2001). Other of the agricultural sector in China from the per- authors, such as Johnson (2000) and Lin (2000), spective of a senior policymaker. For basic foods, have recognized that these statutory tariff rates he emphasizes the shift from a supply-constrained bore little relationship to the protection (or taxa- environment to a demand-constrained environ- tion) actually experienced by China's agricultural ment. The demand for food has grown very slowly, sector. However, the evidence about the actual rates while the growth of supplies has accelerated. Mean- of protection applying to agriculture has remained while, farmers' incomes grew very slowly in the sec- extremely limited and frequently contradictory. ond half of the 1990s. In this environment, Chen Agricultural trade in China has been influenced suggests that Chinese agricultural policymakers by a bewildering array of policies on imports and give priority to, among other things, improving exports, including state trading, designated trading, the quality of food produced and moving toward quotas, licenses, tariffs, and tariff-rate quotas. Many higher-value labor-intensive commodities more studies have attempted to deal with this problem by suited to China's comparative advantage and to summarizing the protective impact of agricultural raising farmers' incomes. He sees very clearly that trade policies through the price distortions created it is important to allow many farmers to leave by these measures. The more restrictive the trade the agricultural sector, and believes the coordi- measure, in general, the larger will be the gap nated expansion of cities and small towns is between the domestic price and the international central to achieving this relocation of labor. Given price. that the shift from supply-constrained to demand- A number of studies have estimated the magni- constrained has been, to a degree, fortuitous and tude of the agricultural price distortions using the unexpected, a key question is whether the tradi- available series on domestic and international tional emphasis on food self-sufficiency will prices. Unfortunately, the results obtained have var- reemerge as a policy driver when imports of key ied wildly. Huang, Chen, and Rozelle (1999), for agricultural commodities increase. example, estimated the protection applying to rice, wheat, and maize in the mid-1990s at 4 percent, 20 percent, and 25 percent, respectively. By contrast, Economic Impacts of Accession Tuan and Cheng (1999) estimated these protection A comprehensive evaluation of the effects of major rates to be 29 percent, 62 percent, and 15 percent, changes in trade policy must be based on an assess- respectively. Carter (2000, p. 80) relied on producer ment of the stance of policy prior to accession-- price data and found generally negative price an assessment that was particularly difficult in distortions. More generally, Carter (2000) and 6 China and the WTO Martin (2001) believed that WTO accession would farm level. Rather, it is collected only on intermedi- require relatively little agricultural liberalization in ate inputs into production. China, while Schmidhuber (2001) and many others Once the implications of agricultural policies believed that the agricultural liberalization required have been assessed, it is possible to begin to assess by accession would be dramatic. Clearly, when the implications of China's accession commitments scholars reach such different conclusions from the for agricultural markets. For those products pro- same facts, a new approach is required, especially tected by ad valorem tariffs, this evaluation is given that the policy implications are so large. straightforward. The reduction in the tariff corre- For this project, Jikun Huang, Scott Rozelle, and sponds quite directly to the reduction in the domes- Min Chang adopted a new approach by basing tic price of the good, and this change, together with their analysis of policy impacts on detailed inter- information on the slope of the import demand views with participants in China's agricultural mar- curve, can be used to estimate the cost of protec- kets rather than on readily available price series (see tion. For those products protected by both a tariff chapter 6). This approach provides a much clearer and an export subsidy, it may be necessary to indication of the implications of agricultural trade consider changes in both variables. The situation policies for product prices, and of the real-world becomes considerably more complex, however, impacts of policies, than would otherwise be possi- when products are being protected, or are to be ble. Their research reveals, for example, that a protected, using tariff-rate quotas. In this situation, major source of the discrepancies in earlier the impact depends greatly on whether the in- research is differences in the quality of domestic quota or out-of-quota tariff determines the price of products and those traded internationally. They the good. When the quota will be filled in some also identify important features of the trade years but not in others, the average rate of protec- regime, such as export subsidies for maize and cot- tion may be a combination of the two tariff rates. ton, that have important impacts on product mar- Some key assessments that have been made of kets. Finally, they raise an important question the implications of the level of protection and the about the implications of the manner in which the changes associated with WTO accession are given value-added tax (VAT) is collected, at a 13 percent in table 1.1. The statutory tariff rates for 1998 used rate, for imported agricultural commodities. It by Schmidhuber and others are given in the first appears that, for administrative reasons, this tax is column of the table; Huang, Rozelle, and Min's esti- not collected on domestic agricultural output at the mates of the actual protection provided in 2001 are TABLE 1.1 Some Measures of Import Protection in China's Agricultural Sector (percent) Postaccession 1998 Statutory Tariffs 2001 Protection Protection Rice 127 3.3 3.3 Wheat 133 12.0 12.0 Maize 130 32.0 32.0 Vegetables and fruits 15 4.0 4.0 Oilseeds 132 20.0 3.0 Sugar 30 40.0 20.0 Cotton 3 17.0 20.0 Livestock and meat 35 15.0 15.0 Dairy 46 30.0 11.0 Note: The third column of the table shows anticipated average rates of import protection after accession taking into account the reforms required by accession and likely market outcomes. Sources: Average statutory rates taken from Schmidhuber (2001) and www.chinavista.com. Estimates of protection in 2001 provided by the authors. Impacts and Policy Implications of WTO Accession for China 7 given in the second column. The third column the importance of the discretion remaining to shows the anticipated average rates of import China's policymakers even after implementation of protection after accession taking into account the such comprehensive commitments on agriculture. reforms required by accession and the likely market Removal of the negative protection applying to outcomes. labor-intensive products would also be consistent Huang, Rozelle, and Min estimated the average with WTO rules, and is likely to be particularly rate of protection for rice to be slightly negative, helpful for employment in rural areas, as well as implying that China's system of state trading for rice economic efficiency. operated to tax rice exports slightly in 2001. After Participation by China in the WTO agricultural accession this rate is expected to remain the same, negotiations being conducted under the Doha because accession to WTO does not require reduc- Development Agenda could potentially reinforce tions in such protection for a state-traded commod- these benefits by opening large, and currently ity. For wheat, protection averaged an estimated highly protected, markets for China's labor- 12 percent. This rate of protection need not be intensive agricultural exports. Unfortunately, the greatly reduced, on average, because it seems likely high rates of agricultural protection that arose that wheat imports will exceed the tariff-rate quota when GATT rules on agriculture were extremely reasonably frequently (see Martin 2001), allowing weak mean that China faces barriers to its agricul- imposition of a tariff of up to 65 percent. For maize, tural exports that are four times higher than those it the rate of import protection utilized was higher, at faces on its other merchandise exports (Martin 32 percent, because of an export subsidy. Although, 2001). on average, the level of import protection need not In chapter 7, Kym Anderson, Jikun Huang, and change greatly because there is a significant proba- Elena Ianchovichina examine the implications of bility that the tariff-rate quota will bind by the end China's agricultural commitments for agricultural of the decade (Martin 2001), the export subsidy protection and for the agricultural sector. They find must be abolished, implying a potentially substan- that the basic WTO accession commitments to tial reduction in the price support given to maize. reduce agricultural import protection and elimi- Labor-intensive vegetables and fruits, like livestock nate agricultural subsidies will make those farm products, experienced negative rates of protection households dependent on agriculture worse off in 2001, and the WTO accession commitments are relative to urban households. If China elects as well unlikely to require changes in protection. to remove the negative protection of important Oilseeds present an entirely different picture in commodities such as rice, vegetables, and meats, which the principal form of protection has been a then returns to unskilled rural labor, and to farm- tariff, and the tariff is being reduced substantially. land would rise slightly, and the impact of accession The protection provided to sugar must be reduced and the elimination of negative protection on rural to meet China's commitments to a bound tariff of wages would be 0.5 percent instead of the 20 percent. As for cotton, import protection will ­0.7 percent observed in the accession case. They not change greatly, but export subsidies, such as the also highlight the importance of China pursuing 10-percent export subsidy observed in 2001, are improvements in agricultural market access, ruled out in the future. Protection of livestock and citing results from Yu and Frandsen (2002) that meat could remain negative as a consequence of suggest that agricultural liberalization by member export restrictions to markets such as Hong Kong. countries of the Organisation for Economic Finally, protection of dairy products can be Co-operation and Development (OECD) would expected to decline to meet China's tariff-binding benefit China and improve its agricultural trade commitments. balance. The reductions in protection between second and third columns of table 1.1 are just one possible Industrial Products outcome in a situation in which rates of agricul- tural protection can vary substantially, particularly In chapter 13, Ianchovichina and Martin examine if import levels exceed the tariff-rate quotas. How- the reductions in protection under way in manufac- ever, the agricultural protection measures illustrate turing and services as well as in agriculture. Within 8 China and the WTO industry, they find that accession builds on the sub- textiles and clothing. Because these barriers require stantial reductions in tariffs undertaken during the an exporter to purchase an export quota--or to 1990s, when weighted average tariffs on manufac- forego the opportunity to sell quotas already tures fell from 46.5 percent in 1992 to 25 percent in held--they impose a cost on exports that is similar 1995. By the time of accession in 2001, weighted to an export tax. Based on information on quota average tariffs on manufactures had fallen to around prices, this tax was estimated to be about 15 percent 13 percent. With full implementation of China's for clothing and 10 percent for textiles.5 accession commitments, they will fall to 6.9 percent. The estimates of the extent of merchandise trade The six-percentage-point reduction in average tar- liberalization used in this study omit some impor- iffs to be implemented after accession is important, tant and potential elements of trade policy. One but very small relative to the 33-percentage-point is the possibility of antidumping and safeguards reduction undertaken since 1992. measures being applied against China. Another is The largest reductions in industrial tariffs are the increasing use of measures of this type in required for beverages and tobacco--an almost China. The potential use of the product-specific 28-percentage-point reduction from 2001 levels. In safeguards against China would be particularly automobiles, the reduction was just over 15 per- important in this respect, because no such measure centage points. Although large, these reductions targeted specifically at China existed prior to were very much smaller than those undertaken China's accession. The risk that China will increase since 1995. The reduction in protection to the auto- its use of antidumping and safeguard measures mobile sector is particularly important because of beyond the current levels is also a concern for the high profile of this industry and its pervasive development policy. Such an action would be a linkages throughout the economy. Other industries triumph of a rules-focused approach to WTO in which substantial reductions in tariffs were implementation--it is legal; therefore, we should required included textiles, clothing, electronics, and do it!!!--over the sharp focus on promoting devel- light manufactures. Many of these industries are opment that has characterized China's trade policy relatively labor-intensive ones in which China has a agenda since the beginning of the reform era. comparative advantage and in which liberalization will help maintain efficiency and competitiveness. Services The liberalization considered in this project is, apart from the export quotas on textiles and Trade in services was a key area in China's WTO clothing, focused entirely on tariffs. This emphasis accession negotiations. To obtain a clear under- understates the degree of liberalization resulting standing of the implications of China's commit- from China's accession, because it ignores the abo- ments in this area, and the potential usefulness of lition of nontariff barriers such as designated trad- the WTO process for China's economic develop- ing, quotas, and licenses. The seriousness of this ment, we consider first Mattoo's overview of exclusion is much less than it would have been in China's commitments in services (chapter 8), and earlier years because of the dramatic reduction in then detailed case studies of logistics (chapter 9), the coverage of nontariff barriers in China--the telecommunications (chapter 10), and the financial frequency of import licenses, in particular, fell from sector (chapter 11). almost half4 in the late 1980s (Lardy 2002) to under Chapter 8 by Aaditya Mattoo compares the lib- 5 percent in 2001. Given the great uncertainty eralization undertaken by China in services with about the protective impacts of nontariff barriers that undertaken by other groups of countries, and in China and their limited remaining coverage, it concludes that China's services reform is the most Ianchovichina and Martin concluded that it was radical ever negotiated in the WTO. In cross-border perhaps better to use tariffs only when considering trade (mode 1), consumption abroad (mode 2), liberalization, and therefore to present something and establishment trade (mode 3), China has made of a lower-bound estimate of the benefits of the lib- more commitments in more service sectors than eralization associated with accession. the industrial countries, other developing coun- It is also important to take into account the tries, or the group of countries that recently reduction in the barriers facing China's exports of acceded to the WTO. Of course, any assessment of Impacts and Policy Implications of WTO Accession for China 9 the importance of these commitments depends on Chinese authorities from moving ahead faster than their implementation, which has drawn reserva- the minimum to which they have committed in tions from Whalley (2003). However, the first two some cases. Given the risks of exacerbating the reviews of China's services commitments at the already substantial inequalities between coastal and WTO appear to indicate that considerable progress interior provinces, there would appear to be good has been made (see WTO 2003a, 2003b). development reasons to phase out these geographic Mattoo notes, however, that China's commit- restrictions more quickly than is required by the ments to service market liberalization were not WTO commitments. indiscriminate. The number of sectors with guar- Imposing regulatory requirements is an impor- anteed unrestricted access is lower for the first two tantroleof government.Someof themostimportant modes than that in most other countries, and objectives of regulations are to make competition essentially zero for mode 3. An important feature of work,improve the availability of information to con- China's commitments is that they focus on market sumers, and ensure universal service. Making com- access and provide national treatment in not petition work is particularly important in network discriminating between domestic and foreign industries such as telecommunications, where indi- suppliers. vidual firms frequently do not find it in their interest Frequently applied restrictions on the establish- to allow interconnection by new firms. Improving ment of services enterprises include: the availability of information is especially impor- tant in financial services.Finally,developing efficient · restrictions on the form of establishment provisions on universal service is important for · restrictions on geographic scope ensuring that all parts of China have access to tele- · regulatory requirements. communications services. In chapter 9, Wenping Luo and Christopher Restrictions on the form of establishment, such Findlay provide a detailed assessment of the re- as requirements for joint ventures, have a long forms required by WTO accession in the range of history in China and have frequently been justified service activities that make up the logistics chain. as a means of achieving technology transfer or as They note that although China has made substan- means of obtaining a share of monopoly rents. tial progress in many of the component activities, Mattoo, however, points out that requirements to logistics costs are disproportionately high in China, form joint ventures might inhibit the transfer of and service quality is lower than is desirable, in part technology. He also points out that a more thor- as an enduring legacy of the planned economy. Luo ough approach to the problem of monopoly rents and Findlay conclude that logistics accounts for would be to ensure that competition between 30­40 percent of the wholesale cost of manufac- firms, whether domestic or foreign, eliminates tured goods, as contrasted with 5­20 percent in the these rents. United States, implying that the scope for gain from Restrictions on geographic scope also have a liberalization is particularly large. High logistics long history--from an era when experimentation costs are a particularly important problem for peo- with market-oriented approaches had to be isolated ple in the poorer areas of China, whose ability to because of the inconsistencies between, for exam- trade, and consequently their real incomes, are sig- ple, planned and market prices. Confining foreign nificantly reduced by these excessive costs. ventures, such as insurance, to five cities for five China's WTO commitments on logistics apply years might encourage agglomeration of these to a range of specific General Agreement on Trade activities in these cities that will not be reversed in Services (GATS) service sectors, including pack- when the geographic restrictions are later lifted. aging and courier services; maritime and rail trans- Moreover, such an approach may reduce the portation; freight forwarding; and storage and opportunities for other parts of China, such as inte- warehousing services. The commitments made in rior cities with a potential comparative advantage these sectors promise to increase competition in in these activities, to get started in these activities. some key areas, including road transport, rail trans- Yet China's WTO commitments represent only a port, warehousing, and freight forwarding. The lower bound, and so they have not prevented breadth of these commitments also provides a 10 China and the WTO much stronger basis for development of integrated will be phased out over several years, China has not third-party logistics firms able to reduce the costs committed to allowing more than 49 percent for- and increase the quality of logistics services in eign ownership for important services such as China. If the envisaged reductions in logistics costs mobile telephony. Because allowing higher levels are brought to fruition, Luo and Findlay conclude of foreign ownership would be consistent with that the costs of a wide range of goods and services China's GATS obligations, China may do so in the might be reduced by about 10 percent. future if such a move appears to promise worth- Yet the commitments outlined in the GATS agree- while gains in a particular activity. ment are not sufficient to achieve the full potential of Clearly, the challenge of establishing a telecom- logistics in China. Regulatory reforms are needed to munications sector that makes the maximum remove discrimination against particular enterprise contribution to China's overall growth and devel- types, to separate local administrations from enter- opment will require further expansion of the regu- prises, and to eliminate local protectionism. In addi- latory framework. Pangestu and Mrongowius tion, substantial investments in infrastructure are believe the central issues here will include ensuring needed to improve the timeliness and reduce the cost the independence of the regulator, ensuring that of providing logistics services. interconnection works adequately, and making In chapter 10, Mari Pangestu and Debbie pricing regulations more flexible. Mrongowius explore China's commitments in In chapter 11, Deepak Bhattasali examines the telecommunications--commitments that take on implications of China's commitments in the finan- particular importance given that China is expected cial sector, including banking, security trading, fund to be the largest market for telecommunications in management, and insurance. According to the world by 2010. These commitments are pro- Bhattasali, reforms up to 1997 focused on institu- found in that they allow foreign entry to a wide tional diversification and strengthened administra- range of activities that are currently closed to foreign tive oversight. Since 1997, reforms have addressed investment. Furthermore, this entry takes place in a the portfolio problems of the banks and governance sector that was monopolized by China Telecom until of the financial sector in preparation for WTO 1994 and is currently dominated by a small number accession. In 2001, however, the four large state of state-owned firms (see DeWoskin 2001). banks still accounted for 67 percent of bank deposits In basic telecommunications, China has com- and 56 percent of total financial assets. The share mitted to the disciplines of the WTO reference of loans going to small and medium-size non­ paper on the regulatory framework for telecommu- state enterprises--the most dynamic part of the nications. The objective is to create a competitive economy--remains small, a finding evident in environment in which interconnection between enterprise survey work reported by Dollar (2002). systems is allowed under reasonable and nondis- Another key concern is the dearth of the informa- criminatory conditions, and it allows for universal tion needed to assess the performance of the finan- service provisions. It also requires the existence of a cial sector. regulator independent of the telecom provider and Although a large number of foreign banks was sets criteria for licensing entry and for allocating active in China on the eve of accession, they scarce commodities such as the mobile telephone operated almost exclusively on an offshore or spectrum. enclave basis and accounted for less than 3 percent In the context of a basic telecommunications of bank assets. However, two years after accession system governed in line with the regulatory paper, they became able to provide local currency services China's other GATS commitments cover value- to Chinese enterprises, and they will become added services such as voice mail and online infor- eligible to offer these services to individuals within mation services; mobile voice and data services; five years. Other areas of financial services, such as and domestic and international services such as stockbroking, fund management, and insurance are private leased circuit services. Most of these ser- also being opened up on quite short timetables. vices are initially subjected to a combination of Like other scholars such as Langlois (2001), ownership restrictions and geographic restrictions Bhattasali believes that the reforms agreed upon in within China. Although the geographic restrictions this vitally important sector were a clear attempt to Impacts and Policy Implications of WTO Accession for China 11 increase competition, performance, and the range sharp decline in output achieved through wide- of products available in financial services. The key spread plant closures. challenges will lie in managing the transition to a Francois and Spinanger conclude that restruc- market-based system without serious problems or a turing in the industry to achieve scale economies in financial crises. He believes that the state banks will final assembly would reduce costs by about 20 per- come under serious pressures from their nonper- cent. This reduction would more than reverse the forming loans, weak management systems, low negative impact on output of the reduction in pro- operating margins, and the strong competition tection from 1997 levels, allowing the industry to they will face. His assessment is that financial re- expand relative to the no-accession case--and to engineering and fairly radical actions to reduce expand dramatically as China's growth and shifting operating costs will be required, but that, given comparative advantage shift resources into sectors suitable reforms, the rehabilitation of the state such as motor vehicles. There also would be impor- banks is unlikely to present major problems. tant changes within the motor vehicle sector. According to the analysis by Francois and Spinanger, the increase in efficiency of the final Impacts on the Economy assembly industry relative to the production of Because China's reforms are so broad-ranging, and intermediate parts is likely to increase the demand their economy-wide interactions are so extensive, for imported parts substantially, with their share they must be evaluated on an economy-wide basis. rising from 39 percent to 52 percent. If the restruc- In chapter 13, Ianchovichina and Martin assess the turing of the industry is achieved successfully, impacts of liberalization on agriculture, manufac- exports of finished motor vehicles are projected to tures, and services. Previous research, however, increase very rapidly, resulting in an increase in revealed that undertaking any such assessment in a total exports of vehicles and parts of over US$4 bil- satisfactory manner required special attention to lion6 a year. Clearly, the reforms associated with the motor vehicle sector. In an analysis in which accession will require major changes through- they did not account for the restructuring of this out the industry and considerable restructuring, industry, Ianchovichina and Martin (2001) found although this process has perhaps been eased by the that output in this industry would decline abso- very rapid growth in demand during recent years. lutely over the period to 2007, despite the strong In chapter 13, Ianchovichina and Martin ana- increases in the Chinese demand for automobiles lyze the impacts of liberalization associated with and the shift in China's comparative advantage to accession in agriculture, manufactures, and serv- more capital- and skill-intensive products such as ices,7 and the opportunities arising from the elimi- motor vehicles. nation of the quotas against China's (and other In their contribution to this study, Joseph F. countries') exports of textiles and clothing. Their Francois and Dean Spinanger examine China's analysis takes into account China's important motor vehicle industry (chapter 12). Like previous export processing arrangements, and it builds on studies such those by Harwit (2001) and the the labor market study by Sicular and Zhao (chap- Chinese Academy of Engineering and National ter 14), a study by Shi Xinzheng (2002) on labor Academy of Engineering/National Research Coun- markets, and the Francois and Spinanger analysis of cil (2003), they conclude that it has been shaped by automobile industry restructuring (chapter 12). policies that have encouraged market segmentation The resulting changes in the specification of their and suboptimal plant size--a structure that is seen model greatly increase the realism of their analysis frequently in countries with highly protected auto- and have very important implications for their mobile industries. As a result, the industry is very results. inefficient, with most plants operating well below Ianchovichina and Martin divide the effects of global standards for efficient production. Unless WTO accession into a component of liberalization these problems can be overcome, the industry is undertaken between 1995 and 2001 in preparation expected to respond to the fall in protection, and for accession and the remaining component to be the increased competition for labor from expand- undertaken after 2001 in order to meet China's ing sectors such as textiles and clothing, with a accession commitments. The choice of 1995 as a 12 China and the WTO starting period is somewhat arbitrary in view of the trade barriers, there are also substantial increases in fact that China reformed its trade regime through- imports of services. The increase in imports of bev- out the 1990s. However, 1995 was an important erages and tobacco is the largest because of the sharp turning point, when China had to forgo its hope reduction in the tariffs for these commodities. of resuming its seat in the GATT and apply as a The biggest change in employment after acces- newcomer to the WTO under a process much more sion is an increase in employment in the apparel focused on the commercial implications of the sector of more than 50 percent. Employment in the accession package. Ianchovichina and Martin rec- textile sector and in plant fibers also increase to ognize that China's product mix and the world's meet the demand from the apparel sector. Trade demand for China's exports are changing rapidly. reform leads to small reductions in employment in They therefore superimpose the impact of liberal- most agricultural sectors and in manufacturing ization on a situation in which China's industrial sectors such as petrochemicals, metals, and auto- structure and output and trade patterns are veering mobiles. Overall, however, the movements of labor sharply toward more capital- and skill-intensive between sectors are generally small relative to the goods in response to high rates of investment and changes in trade patterns. rapid growth in educational levels. The overall welfare gains to China are substan- The authors find that the liberalization associ- tial, particularly from the liberalization undertaken ated with WTO accession results in substantial between 1995 and 2001 to prove China's bona fides growth in trade relative to output, with the total and to prepare for accession. The benefit from the volume of exports rising by 17 percent as a conse- liberalization undertaken during this period is quence of the liberalization after 2001. The most estimated to be a continuing gain of about $30 bil- rapid growth in exports is in apparel; exports lion per year. The smaller reduction in protection expand by over 100 percent in response to the abo- between 2001 and the end of the implementation lition of the export quotas on clothing. Exports of period will generate incremental gains to China of most agricultural products rise because of a decline $10 billion a year. The measured gains in export in agricultural input costs and constraints on the and income growth are very much lower-bound out-migration of labor from the agricultural sector. estimates in that they ignore the benefits from abo- However, exports of plant-based fibers (predomi- lition of nontariff barriers and they involve serious nantly cotton) are projected to fall in response to aggregation biases. Moreover, models of this type the increase in demand for use in export produc- appear to greatly understate the implications of tion. Feed grain exports also fall because of the abo- major trade liberalizations, particularly those asso- lition of the export subsidy on exports of maize. ciated with the rapid emergence of new products Exports of automobiles rise substantially because (Kehoe 2002), which Martin and Manole (2003) of the increase in the efficiency of the sector as it have found to be a major feature of China's reform exploits economies of scale and the greater trade experience. exposure of the industry in the more liberalized The gains from reform do not accrue evenly postaccession economy. None of the increases in within China, with wages of skilled and unskilled exports takes into account the possible benefits urban workers rising modestly and wages for to China of being able to expand its market unskilled farm workers declining by 0.7 percent in access through participation in WTO negotiations. real terms. Ianchovichina and Martin also examine The expansion in textile and clothing exports is, the implications of some potential complementary in fact, a delayed benefit from the Uruguay Round, policies that might be able to deal with this prob- previously denied to China as a nonmember of lem. Reform of the hukou system (a system of the WTO. residence permits regulating movement between Imports rise in a range of sectors in which there urban and rural residence) and other labor market are substantial reductions in trade barriers, includ- barriers is highly desirable because of the ineffi- ing beverages and tobacco, processed food, textiles, ciencies and inequities associated with this segmen- clothing, oilseeds, dairy products, and sugar. tation of the labor market. However, such reform Because Ianchovichina and Martin have repre- increases in importance after WTO accession sented trade liberalization in services as reducing because of the partial liberalization of the Impacts and Policy Implications of WTO Accession for China 13 agricultural sector, which increases the pressures between urban and rural workers and because of for workers to leave agriculture. Adding abolition the explicit policy barriers resisting the movement of the hukou system to the policy package is found of labor between urban and rural sectors. to result in an increase in rural wages of almost In chapter 14, Terry Sicular and Yaohui Zhao 17 percent. Hukou removal completely overwhelms examine the options that households in China face the reduction in rural wages associated with trade in selling their labor--the most important resource liberalization, as an estimated 28 million agricul- of the poor. They find substantial differences tural workers leave the agricultural sector. Com- between the earnings of rural and urban house- plete abolition of the hukou system would put holds. Although some of these differences stem downward pressure on unskilled wages in urban from differences in the quality of labor and other areas, although this reduction of an estimated resources, much is attributable to differences in 3.8 percent would be much smaller than the gain to the earnings received for resources of the same rural workers. quality--a finding that is consistent with the exis- Another complementary reform that tence of substantial barriers between urban and Ianchovichina and Martin consider is an expansion rural labor markets. These barriers appear to be in access to education--a change that would help particularly strong for poorer households, con- to raise unskilled wages in both urban and rural tributing both to depressed income levels and to areas. Improving agricultural technology is another difficulty in adapting to changes in economic policy option with potentially large benefits for opportunities. poor rural households that are able to adopt the Another important feature of Chinese labor new production techniques. markets captured by Sicular and Zhao is the reluc- The results of this quantitative analysis are tance of China's farm households to transfer labor highly stylized, because they assume that enter- from agriculture to other sectors. Although such prises and households are able to adjust success- reluctance is frequently observed in countries that fully to the changes in incentives created by WTO are attempting to transfer labor out of agriculture, accession. As the OECD (2002) has pointed out, it is exacerbated in China by restrictions on the sale successfully making these changes is likely to of farmland usage rights. Permanent movement is require considerable strengthening of the economic likely to require families to relinquish their land system in areas such as enterprise governance and rights without compensation. A main objective of reform of the banking system. the study by Sicular and Zhao is to provide esti- mates of the responsiveness of rural workers to changes in relative returns. They find rural house- Impacts on Households and holds quite responsive to movements in agricul- on Poverty tural returns relative to market wage rates, with The simplest possible approach to capturing the an estimated response elasticity of 2.67, but the effects of WTO accession at the household level response to changes in the ratio of agricultural and requires an assessment of its impacts on the prices nonfarm business returns is much lower. Although consumers pay; on the prices that owners of labor, labor is not perfectly mobile between urban and capital, and other factors receive for their resources; rural labor markets, these results make clear that it and on the government's ability to provide transfers is also far from immobile. or public goods. In addition, it is useful to be able Two studies in this volume integrate the findings to assess the ability of households to adjust to the of the different studies and analyze their implications changes resulting from accession, perhaps by for poverty. They make these assessments by using changing occupations or activities. general equilibrium models of China's economy to Because ongoing work on trade and poverty has evaluate the implications of reform for the prices and found that impacts through factor markets are con- factor returns faced by China's households and then sistently more important than impacts through examining the effects on particular households. consumer prices, it is important to examine factor The study by Shaohua Chen and Martin market effects carefully. This is especially true in Ravallion presented in chapter 15 considers the China because of the large income differences impact of WTO accession on income distribution 14 China and the WTO and poverty drawing on the simulation model takes a longer-run perspective than the Chen and results provided by Ianchovichina and Martin. This Ravallion study, and it assumes that households are study uses a sample of 84,000 households--17,000 more able to move labor between agriculture and urban and 67,000 rural--from surveys by China's other activities, with an elasticity of transformation National Bureau of Statistics. The price impacts of 2.67 for the movement of labor between agricul- implied by Ianchovichina and Martin's Global tural and nonagricultural activities. Like Chen and Trade Analysis Project (GTAP) model analysis are Ravallion, the authors find that urban households applied to the households, taking into account the benefit substantially more than rural households impacts on the prices households must pay for from WTO accession. However, in this longer-run their consumption goods and purchases of inputs, analysis even agriculture-specialized households and the prices they receive for their sales of goods gain on average. Only in a shorter-run experiment, and of labor and other factors. The loss of govern- where households are much less able to move ment revenue from falling tariffs is restored very between sectors, do some agricultural households simply a tax that raises the price of all consumption suffer small losses. goods to maintain government revenues. Hertel, Zhai, and Wang also consider a range of Chen and Ravallion's analysis is focused on the complementary policy reforms that might help to short run, in which households are limited in their deal with adverse impacts on rural households. responses to changes in prices, and on the period Abolition of the hukou system is found to be after accession in 2001, when the impact on the strongly beneficial for rural workers, as are expan- average household and on poverty rates was very sions in education. Both of these reforms are found small. They consider the impacts on individual to be substantially favorable for rural households. households and find sharp differences, particularly In chapter 17, Athar Hussain analyzes the between urban and rural households. Most urban Chinese system of social protection and its value as households, and especially the relatively poor a safety net for shocks of the type involved in WTO urban households, gain from WTO accession, with accession. In his view, China's social protection sys- the poorest urban group gaining by about 1.5 per- tems should not attempt to compensate losers from cent of initial incomes. This is not true for rural the reforms; it is simply too difficult to determine households. The very poorest rural households the magnitude of compensation that is appropriate. experience a sharp reduction in their living Rather, Hussain believes the focus should be on standards--about 6 percent for the poorest per- preventing poverty. He notes that the wide distri- centile. This reduction reflects a combination of bution of land ownership by rural households has falling rural wages and increases in the prices of played an important role in maintaining living consumption goods--items consumed in substan- standards and providing insurance against adverse tial quantities by members of this group. Overall, shocks, but views this insurance as less valuable almost 90 percent of urban households gain from against shocks to grain prices resulting from liber- WTO accession, while over three-quarters of rural alization than against shocks such as the loss of an households lose, although the losses are generally urban job. quite small. The estimated losses to rural house- Hussain observes that China's systems for holds are larger in the Northeast than in other income maintenance and poverty relief have three areas, with average losses to rural households of broad characteristics: a stark dichotomy between over 2 percent in Heilongjiang and Jilin. urban and rural systems; a focus on the reduction The study by Thomas W. Hertel, Fan Zhai, and of absolute poverty; and a high degree of decentral- Zhi Wang in chapter 16 uses a model of the ization in financing. He finds the system in urban Chinese economy that takes into account impor- areas relatively comprehensive, but views the sys- tant features such as the duty exemptions for tem in rural areas as seriously deficient. The locally intermediate goods used in the production of funded system in place in many areas is unable to exports. Because of limitations on the availability deal with a large-scale shock of the type likely to of household data for analytical purposes, it occur with accession to WTO--"the social safety focuses on the three relatively diverse provinces of net is full of holes." Hussain goes on to argue for Liaoning, Sichuan, and Guangdong. This study extending basic safety net protection into rural Impacts and Policy Implications of WTO Accession for China 15 areas. One first step might be to extend a mecha- software and about problems with inadequate nism like the urban unemployment insurance enforcement of IPRs that reduces the incentive to schemes to wage employees in township and village provide innovations. enterprises. However, Hussain believes the central Agriculture is being liberalized by less than was element is the development of a national scheme, suggested by some earlier studies that began from such as the Minimum Living Standard Scheme, the assumption that agricultural protection would that targets poor rural households. be reduced from initial statutory tariff levels. Signifi- cant liberalizations have been achieved in areas such as maize, cotton, and sugar, and it appears that the Conclusions adjustment pressures in these industries will be sig- China's WTO accession agreement calls for sub- nificant. There will, however, be opportunities to stantial reductions in protection, for a strengthen- expand exports of some labor-intensive exports ing of intellectual property rights protection, and as part of a broader policy reform and by seeking for adoption of a framework of trade rules at home increases in agricultural market access in the Doha and abroad to facilitate trade growth. While requir- negotiations. ing many policy changes, it leaves open a wide The industrial sector will face substantial adjust- range of policy choices and, in these, China should ment pressures in key sectors such as automobiles, continue to focus strongly on its development and beverages and tobacco, where external protec- needs--the central perspective that has guided the tion is being substantially reduced. Restructuring of steady transformation of its trade regime, and its scale-intensive sectors such as the automobile sec- economy generally as it moves from a planned to a tor will be essential and can generate substantial market economy. productivity gains. Overall, however, most of the One key concern in the accession agreement is adjustment in this industry has already occurred, the provisions on antidumping and safeguards. The and what remains is an expansion of both imports nonmarket provisions that countries are permitted and exports. to invoke against China for up to 15 years are likely China's GATS commitments represent perhaps to result in antidumping duties substantially higher the most thorough-going liberalization of the serv- than those in other countries--in a situation in ices trade ever undertaken in the GATT. The range which China faces seven times as many antidump- of commitments is extremely broad, although some ing actions per dollar of exports as the United commitments involve restrictions on ownership, States. The special product safeguards applicable business scope, or region. Critical sectors such as against China for the next 12 years are a new form telecommunications, logistics, and financial serv- of protection, applicable only against China. The ices are to be confronted with renewed competition, provisions on trade diversion lack even basic proce- and they are likely to see a burst of innovation and dural restraints, and pose a potentially serious productivity growth as restructuring proceeds. threat to China's export development. The tempta- The impacts of China's WTO commitments-- tion to retaliate, particularly with antidumping and China's successful economic development-- actions of its own, will be strong but would damage depend heavily on the ability of China's labor China more than other countries. A better option markets to reallocate labor from agriculture to for China would be to seek reform of WTO rules in other activities. Analysis reported in this volume these areas. concludes that these markets are adversely affected China's TRIPS arrangements generally appear to by a range of regulations such as the hukou system. have been implemented in a manner consistent Other features of the labor market, such as the "tie" with both the legal requirements of the agreement to the land where households have use rights to and China's development needs. They generally land but cannot sell it because property rights are provide a reasonable balance between incentives to not sufficiently well defined, restrict the mobility of innovate and access to those innovations. Some labor out of agriculture. These features of the econ- concerns have emerged about proposals that might omy inhibit the adjustment needed after accession strengthen protection too much to allow access to and increase the vulnerability of poor people to innovations such as those for patent protection on downturns in agricultural prices. 16 China and the WTO We report in this volume two studies that evalu- process--with its emphasis on continuing reform ate the impacts of trade reform on poverty. Both designed to improve economic performance--with studies find that the rural sector is more vulnerable the rules-oriented approach of the WTO. The ben- to this reform than is the urban population, and efits of accession to China will be substantial, but both conclude that urban households benefit more reaping them will require continuing reforms in the than rural households from the package of reforms adventurous spirit of China's economic reforms-- associated with accession. The study that looks at reforms that focus on economic development and the shorter term finds rural households are more go beyond trade policy into areas such as labor subject to negative shocks. Although these impacts market reform and rural development. are in general relatively small, they affect most rural households. The effects are largest in proportional Notes terms among the poorest households, and tend to 1. See chapter 13 for a discussion of the methodology used be geographically concentrated, particularly in the for these computations. northeastern provinces such as Heilongjiang and 2. Although China's transitional review at the WTO allows Jilin. The second study provides a relatively opti- other WTO members to seek information and raise concerns mistic longer-run picture, where virtually all about China's trade policies, the process is multilateral, and China can raise concerns about implementation of other Mem- households gain from the reform. bers' commitments in the protocol. These models and the overall modeling studies 3. Although internal protectionism appears to have abated suggest that the answer to mitigating these prob- in recent years, survey results suggest that problems remain, par- ticularly local government interventions in labor markets and lems is to include reforms such as reductions in the administrative discrimination (DRC 2003). These interventions barriers against movement of labor out of agricul- appear to hinder foreign investment (Amiti and Javorcik 2003). ture. Removal of the hukou system would increase 4. Two-thirds of exports were also subject to licensing. 5. See www.chinaquota.com. rural wages substantially, with a relatively modest 6. All dollar amounts are current U.S. dollars. negative effect on higher-income households. A 7. The estimated liberalization of services is very crudely range of other policy options directed toward approximated by halving the barriers to trade in these activities improving the welfare of rural people, such as rural estimated by Francois. research and development, improvements in rural infrastructure, expansion of educational opportu- References nities in rural areas, and improvements in rural The word processed describes informally reproduced works that health services, are all fully WTO consistent and may not be commonly available through libraries. could have powerful beneficial effects in both the Amiti, M., and Javorcik, B. 2003. "Trade Costs and Location of Foreign Firms in China." Paper presented at Workshop on short and the long run. As an important WTO National Market Integration. Development Research Centre member, China also has an opportunity to press for of State Council and World Bank. Beijing, September 6. greater market access opportunities for labor- Anderson, S., and C. Lau. 2001. "Hedging Hopes with Fears in intensive agricultural exports that could help gen- China's Accession to the World Trade Organization: The Transitional Special Product Safeguard for Chinese Exports." erate employment in rural areas. Powell, Goldstein, Frazer, and Murphy LLP, Geneva. The network of social protection measures in Processed. China is quite underdeveloped and a constraint on Carter, C. 2000. "China's Trade Integration and Impacts on Fac- tor Markets." In Organisation for Economic Co-operation the country's ability to grow while dealing with and Development, China's Agriculture in the International widely held concerns about the need to compensate Trading System, OECD Proceedings. April. the potential "losers" from the policy reforms. Par- Chinese Academy of Engineering and National Academy of Engineering/National Research Council. 2003. Personal Cars ticular attention should be given to strengthening and China. Washington, D.C.: National Academies Press. the social welfare systems available to rural Chun, C., B. Fleisher, and E. Parker. 2001. "The Impact of residents. China's Entry into the WTO: Overview." China Economic China's accession to the WTO was a defining Review 11(4): 319­22. DeWoskin, K. 2001. "The WTO and the Telecommunications moment for both China and the WTO. It necessar- Sector in China." China Quarterly 167: 630­54. ily moves the process of trade reform in China away Dollar, D. 2002. "Improving the investment Climate in China." from an incremental approach to one incorporat- Presentation to the Conference on China's Investment Climate, Beijing, December 3. ing quite detailed rules for trade policy. It is impor- DRC (Development Research Centre). 2003. "Measures, Objects tant to combine the spirit of China's reform and Degrees of Local Protection in the Chinese Market--An Impacts and Policy Implications of WTO Accession for China 17 Analysis Based on a Sample Survey." Paper presented at McKibbin, W., and K. Tang. 2000. "Trade and Financial Reform Workshop on National Market Integration. Development in China: Impacts on the World Economy." World Economy Research Centre of State Council and World Bank. Beijing, 23(8). September 6. OECD (Organisation for Economic Co-operation and Develop- Fewsmith, J. 2001. "The Political and Social Implications ment). 2002. China in the World Economy: The Domestic of China's Accession to the WTO." China Quarterly 167: Policy Challenges. Paris. 573­91. Panitchpakdi, S., and M. Clifford. 2002. China and the WTO: Gilbert, J., and M. Wahl. 2002. "Applied General Equilibrium Changing China, Changing World Trade. Singapore: John Assessments of Trade Liberalization in China." World Wiley and Sons. Economy 25(5): 697­731. Schmidhuber, J. 2001. "Changes in China's Agricultural Trade Harwit, E. 2001. "The Impact of WTO Membership on the Policy Regime: Impacts on Agricultural Production, Con- Automobile Industry in China." China Quarterly 167: sumption, Prices, and Trade." In Organisation for Economic 655­70. Co-operation and Development, China's Agriculture in the Huang, J., M. Chen, and S. Rozelle. 1999. "Reform, Trade Liber- International Trading System, OECD Proceedings. April. alization and Their Impacts on China's Agriculture." Paper Shafaeddin, S. 2004. "Is China's Accession to WTO Threatening presented at International Agricultural Trade Research Con- Exports of Developing Countries?" China Economic Review, sortium Symposium on "China's Agricultural Trade Policy: forthcoming. Issues, Analysis and Global Consequences." San Francisco, Shi Xinzheng. 2002. "Empirical Research on Urban-Rural June 25­26. Income Differentials: A Case of China." Peking University. Ianchovichina, E., and W. Martin. 2001. "Trade Liberalization in Processed. China's Accession to WTO." Journal of Economic Integration Tuan, F., and G. Cheng. 1999. "A Review of China's Agricultural 16(4): 421­45. Trade Policy." Paper presented at International Agricultural Jin, Liqun. 2002. "China: One Year into the WTO Process." Trade Research Consortium Symposium on China's Agricul- Paper presented at World Bank, Washington, D.C., October tural Trade Policy: Issues, Analysis and Global Conse- 22. Available at www.worldbank.org/wbi/B-SPAN/docs/ quences. San Francisco, June 25­26. IMF-WB_address_final.pdf. Wen, J. 2004. "Report on the Work of the Government." Report Johnson, D. G. 2000. "The WTO and Agriculture in China." delivered at the second session of the National People's Con- China Economic Review 11: 402­4. gress. Beijing, March 5. Kehoe, T. 2002. "How Well Did Computable General Equilib- Whalley, J. 2003. "Liberalization in China's Key Service Sectors rium Models Predict the Effects of NAFTA?" Paper pre- Following WTO Accession: Some Scenarios and Issues of sented at Conference on General Equilibrium Modeling. Measurement" NBER Working Paper 10143, National Cowles Foundation, Yale University. April. Bureau of Economic Research, Cambridge, Mass. Lardy, N. 2002. Integrating China into the Global Economy. Woo, W. T. 2001. "Recent Claims of China's Economic Excep- Washington, D.C.: Brookings Institution Press. tionalism: Reflections Inspired by WTO Accession." China Langlois, J. 2001. "The WTO and China's Financial System." Economic Review 12: 107­36. China Quarterly 167: 610­29. World Bank. 2001. Global Economic Prospects 2002: Making Trade Lin, J. Y. 2000. "WTO Accession and China's Agriculture." China Work for the World's Poor. Washington D.C: World Bank. Economic Review 11: 405­8. WTO (World Trade Organization). 2002. "Transitional Review Martin, W. 2001. "Implications of Reform and WTO Accession under Article 18 of the Protocol of Accession of the Peoples' for China's Agricultural Policies." Economics of Transition Republic of China." G/C/W/438. Geneva. 9(3): 717­42. ------. 2003a. "Report of the Meeting Held on 5 December Martin, W., and E. Ianchovichina. 2001."Implications of China's 2003." Council for Trade in Services, S/C/M/69. Geneva. Accession to the WTO for China and the WTO." World Econ- ------. 2003b. "Report of the Meeting held on 1 December omy 24(9). 2003. Council for Trade in Financial Services, S/FIN/M/43. Martin, W., and V. Manole. 2003. "China's Emergence as the Geneva. Workshop of the World." Paper presented at Stanford Center Yu, W., and S. E. Frandsen. 2002. "China's WTO Commitments for International Development Conference on Chinese in Agriculture: Does the Impact Depend on OECD Agricul- Economic Policy Reform. Stanford University, September tural Policies?" Paper presented at Fifth Conference on 18­20. Global Economic Analysis. Taipei, June 5­7. Part I Policy Reforms Associated with Accession 2 What China's WTO Accession is All About Jeffrey L. Gertler With the gavelling of the accession package at China would leave the GATT. Although the govern- the conclusion of the Working Party meeting on ment in Beijing never recognized this withdrawal September 17, 2001, the negotiation on China's decision, nearly 40 years later, in 1986, China noti- accession to the World Trade Organization (WTO) fied the GATT of its wish to resume its status as a finally was concluded. The WTO Ministerial Con- contracting party and its willingness to renegotiate ference subsequently approved the terms of China's the terms of its membership. accession in Doha, Qatar, on November 10, 2001, A working party to examine China's status and the Chinese Government notified its accept- established in March 1987 met for the first time in ance on November 11. In line with customary prac- October of that year. The GATT Working Party on tice and as set out in China's Protocol of Accession, China's Status met on over 20 occasions but with- China formally became a Member of the WTO out conclusion. With the formation of the WTO in 30 days later on December 11, 2001. 1995, the GATT Working Party was converted into Each accession to the WTO is a unique event, a WTO Working Party on the Accession of China. but China's accession has been particularly note- The Working Party, chaired by Ambassador Pierre- worthy. China had been one of the 23 original con- Louis Girard of Switzerland, met 18 times. tracting parties to the General Agreement on Tariffs Article XII of the Marrakesh Agreement Estab- and Trade (GATT) in 1948. However, the applica- lishing the WTO (the "WTO Agreement"), which tion process for readmission to the multilateral governs accessions, is striking in its brevity. The trading system took 15 years from its submission in operative provision reads: "Any State or separate July 1986, making it easily the longest and most customs territory . . . may accede to this Agree- arduous accession negotiation in the history of the ment, on terms to be agreed between it and the GATT/WTO. A review of the application process WTO." By early April 2004, more than 20 new contributes to an understanding of the overall members had joined the WTO since its establish- process and highlights some key features of China's ment, bringing membership to 146, including Accession agreement. three separate customs territories (Chinese Taipei,1 After China's revolution in 1949, the govern- Hong Kong [China], and Macau). Cambodia and ment in Taiwan (China) announced in 1950 that Nepal had been accepted for membership but 21 22 China and the WTO were still in the process of completing ratification BILATERAL NEGOTIATIONS procedures. China's importance as a trading nation generated The final stages of the China accession process considerable enthusiasm for its participation in the can be classified rather naturally under three accession negotiations. Some 44 WTO members, headings: including the 15 member States of the European Union as one entity, expressed interest in conclud- · conclusion of bilateral market-access negotia- ing bilateral market-access negotiations with China. tions As these bilateral deals were struck and notified · conclusion of multilateral negotiations in the to the WTO, China's consolidated Schedule of Con- Working Party and the corresponding documents cessions and Commitments on Goods and its con- that together stipulate the terms of China's solidated Schedule of Specific Commitments on accession, including the draft Protocol and its Services were prepared with assistance from the Annexes, and the Working Party Report WTO Secretariat. Thereafter, they were reviewed · approval and acceptance of these terms of in the Working Party and multilateralized, that accession by WTO members and by China, is, extended on a most-favoured-nation (MFN) respectively. basis to all WTO members, as China's Goods and Services Schedules, annexed to the Protocol of Before delving into a description of these three Accession. In this way, these bilateral commitments steps, it is worth reviewing briefly the many ups and became part of the multilateral treaty terms of downs China experienced in the course of the China's membership in the WTO. accession process. I refer in particular to significant China was only able to make rapid progress in progress made just prior to the Tiananmen "inci- concluding its bilateral negotiations with most dent,"followed by almost two-and-a-half years with other WTO member governments once it reached virtually no accession activity, China's subsequent bilateralagreementwiththeUnitedStatesinNovem- participation in the Uruguay Round negotiations, ber 1999 and then with the European Communities and its failure to conclude negotiations on its status in May 2000. Thereafter the negotiations proceeded as a GATT-contracting party in time to become an apace. original member of the WTO. Next were the con- Only in the final days of the Working Party was version of the GATT Working Party into a WTO China able to conclude negotiations with Mexico, accession working party in December 1995, the the last of the 44 WTO members seeking bilateral considerable optimism about an accelerated pro- market-access commitments. The sticking point cess in early 1997, the near-conclusion of a bilateral had been the hundreds of antidumping orders that agreement with the United States in April 1999, the Mexico continued to maintain against products of U.S. bombing of the Chinese Embassy in Belgrade Chinese origin. Mexico eventually agreed to termi- the following month, and the conclusion of a bilat- nate these allegedly WTO-inconsistent measures eral agreement with the United States in November six years after China's accession. In addition, 1999. These developments were followed by a spate El Salvador, which recognizes Taiwan (China) and of additional bilateral agreements in the first half of not China (as is true of other WTO members), 2000, including that with the European Union in did not request bilateral negotiations with China May. Further hiccups occurred subsequently, in- and invoked the non-application provision of cluding the downing of a U.S. spy plane over the Article XIII of the WTO Agreement against China. Straits of Taiwan, the Bush administration's aggres- One element contributing to pressure on China sive enthusiasm for a new form of the Strategic and Members to conclude bilateral accords was the Defense Initiative called the Missile Defense System, U.S. administration's agreement with China. As a and finally, the terrorist attack in the United States quid pro quo for China's market-access concessions, and the U.S.-led retaliatory actions since that time. the United States would provide China with per- Clearly, China and WTO members have been on a manent MFN status, thereby eliminating the annu- roller-coaster ride of major proportions over much ally renewed conditional MFN provided under the of the accession period. Jackson-Vanik Amendment to the U.S. trade act of What China's WTO Accession Is All About 23 1974. After much debate, the U.S. Congress finally · trading rights passed unconditional MFN status (what it calls · standards and technical regulations. permanent normal trade relations or PNTR) for China in September 2000. Additional agreements were reached on the following: MULTILATERAL STEPS · a special transitional provision lasting 15 years With the bilateral market-access negotiations near- on price comparability in determining subsidies ing completion, members and China showed and dumping renewed interest in wrapping up the many out- · the establishment of both a transitional standing multilateral elements of the accession product-specific safeguard mechanism and a package. In order to finalize the negotiated package separate transitional textile safeguard and in recognition of the fact that much of the · immediate implementation of the Agreement information China had submitted to the Working on Trade-Related Aspects of Intellectual Prop- Party was incomplete or out of date, the Working erty Rights (TRIPS) Party requested that China submit updated · a transitional review mechanism to oversee information--including notifications of laws, reg- compliance with the terms of the Protocol ulations, and other policy measures--on all key · a host of technical sectoral issues in trade in aspects of China's trade regime. services. Clearly, such information was indispensable to assessing the congruency of China's trade regime While China has reserved the right to exclusive with WTO rules. It was also indispensable to final- state trading for products such as cereals, tobacco, izing negotiations on key provisions of the Protocol fuels, and minerals, and to maintain some restric- and the Report. Identifying the trouble spots and tions on transportation and distribution of goods agreeing on the timing--including possible tran- inside China, many of the restrictions that foreign sition periods--for China to bring any WTO- companies currently face in China will be phased inconsistent policy measures into compliance with out over a three-year transition period. During a WTO obligations, presented major challenges. 12-year period after accession, WTO members will The final meeting of the Working Party in have access to a transitional safeguard mechanism September 2001, with informal sessions the week in cases where imports of products of Chinese ori- of September 10 and the formal meeting on gin cause or threaten to cause market disruption to September 17, was devoted to completing the tech- member's domestic producers. nical cleanup and verification of the Goods and Upon accession China became a party to the Services Schedules, followed by an overall review of Agreement on Textiles and Clothing; accordingly, the documents to ensure consistency among the as for all WTO members, quotas on textiles will end various elements of the accession package. on December 31, 2004. However, a special, negoti- China and Working Party members finally ated, safeguard mechanism will remain in place reached agreement on all outstanding issues: until the end of 2008, permitting WTO members to take action to curb imports in case of market · preambular and general provisions disruption caused by Chinese exports of textile · the administration of the trade regime, includ- products. ing uniform administration, special economic Several areas of negotiation were problematic areas, transparency, and judicial review until very late in the process. First, there was lack of · nondiscrimination, special trade arrangements, agreement on the availability to China of WTO state trading, nontariff measures, tariff-rate provisions for developing countries in areas such as quota administration, import and export licens- domestic support to agriculture and industrial sub- ing, price controls, taxes and charges levied sidies. The United States, in particular, objected to on imports and exports, export subsidies and providing China with the full benefit of developing domestic support in agriculture, and sanitary country provisions in these areas. The matter even- and phytosanitary measures tually was resolved; China agreed not to resort to 24 China and the WTO certain of these WTO provisions and accepted a cap The Protocol and Working Party Report essen- of 8.5 percent on domestic support in agriculture, tially contain a one-way set of commitments (from below the 10 percent available generally to develop- China's side only), although these documents also ing country members under the WTO Agreement. contain some "soft" commitments by Members, for Because this 8.5 percent cap is still well above exist- example, regarding the non-abuse of domestic pro- ing budgetary outlays by the Chinese government, cedures in anti-dumping actions and restraint in it is doubtful whether the cap will have a major the use of the special safeguard. Additionally, there impact on China's ability to help its farming sec- is an unusual Annex to the Protocol, containing tor adapt to the new and evolving conditions of commitments by certain Members to phase out competition. inconsistent measures maintained against China A second area that was among the last resolved over a transition period lasting up to five years. was the regime relating to trading rights. The basic regime eventually agreed to calls for a fully liberal- APPROVAL AND ACCEPTANCE ized right for foreign companies to gain trading rights in China after a three-year transition period; Once consensus was achieved in the Working Party state-traded products remain an exception. on the final accession package, this was forwarded A third area of considerable difficulty was agree- to the General Council for decision. Given the ment on China's handling of its regime dealing with timing of the Fourth Ministerial Conference in technical regulations and standards; the key issue mid-November 2001, it was decided that these doc- was how to ensure nondiscrimination (national uments should be forwarded to Doha for approval treatment) in the application of this regime. After by Ministers, rather than be decided upon at the repeated urging spanning several years, China level of the General Council in Geneva. finally committed to unifying its administrative In accordance with established procedures, the structure responsible for the inspection and con- Ministerial Conference approved the Decision on formity assessment procedures (under the Admin- Accession and the Protocol on the terms of China's istration for Quality Supervision, Inspection and accession on November 10, 2001. China accepted Quarantine [AQSIQ]) for both domestic and the WTO's Protocol of Accession on November 11, imported goods, thereby allaying many of the con- 2001, and became a Member of the WTO 30 days cerns raised by the dual and separate systems of later. Notifying the Director-General that the Stand- inspection that previously existed. ing Committee of the People's Congress had rati- As part of the concluding phase, the Working fied the terms of accession, China became the 143rd Party also reviewed and obtained amendments and member of the WTO on December 11, 2001.2 clarifications to many of the transitional Annexes Article XII:2 of the WTO Agreement provides of the Protocol. This was done on the basis of that "the Ministerial Conference shall approve the updated and revised drafts of these Annexes agreement on the terms of accession by a two- provided by China. They included the following thirds majority of the Members of the WTO." annexes: However,pursuant to Article IX:1 of the same agree- ment and a 1995 decision of the General Council, · Products subject to state trading all accession decisions are to be approved by con- · Products subject to designated trading sensus (with possible recourse to voting only where · Nontariff measures subject to phased consensus is not achievable). All 16 WTO accession elimination decisions to date, including China's, have been · Products and services subject to price controls taken by consensus. · Notification and phase-out of subsidies · Export taxes and charges CHINA'S ACCESSION IN THE CONTEXT · Restrictions maintained against China OF THE WTO'S BASIC PRINCIPLES · Issues to be addressed in the transitional review · the Schedule of Concessions and Commitments It could be interesting to consider briefly how on Goods, as well as the Schedule of Specific China's accession fits within the context of the five Commitments on Services. fundamental principles of the GATT and the WTO. What China's WTO Accession Is All About 25 Simply put, these principles cover the following: progressively and substantially open up its services sectors to foreign competition also is undeniable. · Nondiscrimination Moreover,China has already demonstrated its inten- · Market opening tion to play a significant role in the Doha Develop- · Transparency and predictability ment Agenda. · Undistorted trade · Preferential treatment for developing countries. Transparency and Predictability These are key elements of the multilateral trading Non-Discrimination system. The basic transparency principle, contained Two types of nondiscrimination are of interest: in GATT Article X, calls on member governments to the MFN principle and the national treatment prin- promptly publish all trade-related laws, regulations, ciple. Under the MFN principle, a member may not judicial decisions, and administrative rulings of discriminate between its trading partners: goods general application; to administer all such measures and services and service providers are to be accorded in a uniform, impartial, and reasonable manner; MFN, that is, equal, treatment. At the same time, a and to provide for independent judicial review pro- member must provide national treatment: it may cedures for the prompt review and correction of not discriminate on its internal market between its administrative actions. The predictability principle own and foreign products, services, and nationals. is ensured through a legal hierarchy giving prefer- Where do things stand in terms of China's acces- ence to tariffs over less transparent and less secure sion vis-à-vis the principle of nondiscrimination? non-tariff measures such as quotas and licences, China, like other members, has committed itself and through encouraging members to "bind" their to abide by all WTO agreements, including those market opening commitments in goods and serv- provisions requiring application of MFN and ices. In the goods area, this binding amounts to national treatment. In its Protocol of Accession, setting ceilings on customs tariff rates. China has agreed to undertake additional commit- China has committed to abide by the WTO's ments to ensure the smooth phasing in of these non- transparency obligations across the board-- discrimination principles. Of particular note are including with respect to uniform application of its commitments to eliminate dual pricing practices trade regime and independent judicial review-- and to phase out within three years most of the and has made additional commitments in each of restrictions on importing, exporting, and trading these areas. While difficulties may exist with respect currently faced by foreign enterprises. All foreign to variations in treatment in different parts of enterprises, including those not invested or regis- China's customs territory, as well as with the per- tered in China, are to be accorded treatment no less ceived lack of independence of the judiciary, there favorable than that accorded to enterprises in China. can be little doubt that the Chinese government is committed to carrying through the necessary reforms to implement these obligations in a uni- Market Opening form and impartial manner. Also, as noted, China's The principle of market opening is promoted in accession commitments will be the subject of a the WTO through successive rounds of multilateral special transitional review mechanism for the first trade negotiations aimed at the progressive lower- 10 years of membership. ing of trade barriers. New members are pressed to With China as a member, her producers and liberalizetheirtraderegimes during accession nego- exporters will more confidently be able to make tiations. Trade ministers also initiated the latest long-term business decisions on the expansion of round of multilateral negotiations at the Doha their activities.The more open the Chinese economy Ministerial Meeting in November 2001. becomes, the more China will benefit from the legal With respect to market opening, China has sig- security of the rules-based trading system. Not just nificantly reduced its tariff and nontariff barriers as foreign investors, exporters, and importers, but also part of its bid to join the WTO. The breadth and all Chinese citizens,will benefit from the more open, depth of the cuts are evident. China's willingness to nondiscriminatory reforms China is undertaking. 26 China and the WTO In terms of producing a more predictable and countries and countries in transition to market secure trading environment, China has bound all economies to adjust their systems to many of its import tariffs in the goods area. China also has the new obligations resulting from the Uruguay committed to the phased reduction and removal of Round. A Ministerial Decision gives additional tariff barriers, mostly by 2004, but no later than flexibility to the least-developed countries in imple- 2010. China's average bound tariff level will menting the various Uruguay Round agreements decrease to 15 percent for agricultural products, and calls on developed country Members to accel- ranging from 0 to 65 percent, with the higher rates erate their implementation of market access applied to cereals. For industrial goods, the average commitments on goods exported by the least- bound tariff level will go down to 8.9 percent, with developed countries. a range from 0 to 47 percent, with the highest rates Although China has not been granted across- applied to photographic film and automobiles and the-board preferential treatment as a developing related products.In services,China has made a more country, it has negotiated specific transitional comprehensive set of initial commitments than arrangements in certain areas of its trade regime. those offered by most developed countries during Examples include the phasing out of quotas and the Uruguay Round. Of particular note are China's import licenses, and the phased liberalization of commitments in services sectors covering telecom- the right for foreign entities to trade in China. In munications, banking, and insurance. contrast, despite the availability of more preferen- tial treatment under the WTO agreements, China has accepted a special cap on its ability to provide Undistorted Trade domestic production subsidies in agriculture, has The WTO system also promotes undistorted agreed not to use export subsidies, and has com- trade through the establishment of disciplines on mitted to immediate implementation of the TRIPS subsidies and dumping, allowing members to Agreement. respond to unfair trade through the imposition of countervailing or antidumping duties. The treaty INSTITUTIONAL IMPLICATIONS allows individual members to impose temporary safeguard measures, under strict rules, when faced No one can contest that China's participation in the with a sudden surge in imports causing serious WTO will affect the operations of this organization injury to a domestic industry. in substantial ways and over the long term. China is As in other areas, China has agreed to abide by joining as the seventh largest exporter and eighth all WTO disciplines relating to subsidies and coun- largest importer of merchandise trade and as the tervailing measures, protection provided under twelfth leading exporter and tenth leading importer rules for antidumping, and safeguard measures. of commercial services. It has the largest popula- As noted above, it has also committed not to use tion and largest potential market of any WTO export subsidies on either industrial or agricultural member. goods and has accepted special provisions sought Undoubtedly, China's membership will have by other members in relation to determinations of implications for the regular work of the WTO's dumping or subsidies, as well as a special product- many committees administering the many agree- specific safeguard mechanism and a separate textile ments of this institution. China will surely be active safeguard. China has indicated its intention to join in the newly launched and future rounds of multi- the plurilateral Agreement on Government Pro- lateral trade negotiations, in agriculture, and in curement, which is aimed at ensuring fair competi- services, but also in other areas of mutual concern. tion rules in purchases by government procurement It will participate in fashioning the improved insti- agencies. tutional operations of the WTO. Clearly, China's membership is likely to result in expanded recourse to the dispute settlement procedures of the WTO, Preferential Treatment for Developing Countries both by China and by other members in relation to This principle permeates the entire WTO Agree- China's implementation of its WTO commitments. ment, providing transition periods to developing The first such case was that recently brought by What China's WTO Accession Is All About 27 China, similar to that brought by many other WTO it has had to use until now, to shape its trade members, against the U.S. steel safeguard measures. with its major trading partners, with a single, We should, of course, expect to see some new faces multilateral trade relationship with the rest of in the Secretariat. the world. Of course, what lies beyond China's accession is the major, continuing, and, in many ways, impon- CONCLUDING REMARKS derable task of implementation by China of its Since the mid-1980s, the process of reform in WTO accession commitments. A question upper- China has matured considerably, and China's trade most in the minds of many Chinese and foreigners performance has reflected this maturation. China alike is whether and how China will be able to has become a very important player in interna- ensure uniform and impartial implementation of tional trade, on both the import and the export trade commitments. sides. Moving from an earlier phase of import plan- At this stage, it is difficult to predict the speed ning to one of import licensing and then, more sig- with which WTO members and China will resort to nificantly, to one of import tariffs, has brought the WTO's dispute settlement procedures. There is prices and the market mechanism into play as key little doubt, however, that China and its trading determinants of China's future trade relations with partners will eventually take full advantage of WTO the rest of the world. mechanisms to resolve trade disputes. While con- Accession should allow China to lock-in the siderable emphasis has been placed on the 10-year accumulated benefits of the trade reform process transitional review mechanism provided in China's that the Chinese government has undertaken to Protocol, this mechanism does not contain any date, and it should provide a platform from which enforcement provisions. China can sustain its reform process into the future. It is hard to overstate the difficulties many sec- By placing China's reforms within the broader tors of Chinese society will face in the months and context of trade liberalization by all WTO mem- years following accession. The impact on loss- bers, Chinese producers and exporters can increase making state industries, less developed agricultural the returns from trade reform in China through communities, and myriad government-financed reciprocal market access abroad and help the projects throughout the country will be dramatic. Chinese government resist pressure domestically to Moreover, the adjustment to new, more competi- reverse the process of reform. tive market conditions, will, for many millions of For China, WTO accession will provide the individuals and families, mean unemployment and 1.3 billion Chinese people with secure, predictable, significant displacement. It will take many years for and nondiscriminatory access to the markets of large segments of China to establish a new equilib- 143 trading partners. It will also give this same rium, during which time many citizens may well enormous population secure and nondiscrimina- face considerable hardships. tory access to the goods and services of these other However, we should bear in mind that this WTO members. adjustment process has already begun. The Chinese However, from China's perspective, membership people are hardly strangers to this process. Already also for the first time commits this new WTO in the early 1990s China had introduced a player at the international level to implement legal bankruptcy law and other legislation making state and domestic policy reform, ensuring much greater industries in principle responsible for their own transparency and security on a uniform basis. profits and losses. For at least the past decade, China has made impressive strides at reform over China has radically reduced state subsidies and the past 20 years. However, committing itself to encouraged development of private enterprise in abide by international treaty rules and the rule of many sectors. Since the mid-1990s in particular, law in the conduct of trade and in domestic policy and as a member of the International Monetary reform is likely to take this process forward at an Fund, China has rationalized and liberalized even more impressive pace. handling of its foreign exchange market. In addi- Accession will also mean that China can replace tion, China has progressively and dramatically the many risky and uncertain bilateral relationships reduced its import tariffs and other nontariff 28 China and the WTO restrictions on foreign participation in the Chinese Notes market. 1. The nation of Taiwan (China) became a member of the The difficult adjustment is far from complete. At WTO as Chinese Taipei. this stage, therefore, we can do little more than wish 2. Chinese Taipei, whose terms of accession were approved by Ministers on 11 November 2001, became a WTO member on China and its people "bon courage" as they venture 1 January 2002. down the extremely challenging path that stretches before them. 3 China in the WTO: Antidumping and Safeguards Patrick A. Messerlin A few weeks prior to China's acceptance as a full These include continued use of non-market econ- member of the World Trade Organization (WTO), omy (NME) status in antidumping investigations chief negotiator Long Yongtu had ranked stricter and the use of a special transitional product-specific antidumping rules second among China's priorities safeguard (TPS) provision. These two provisions in the WTO. Antidumping rules define the condi- are scheduled to last 15 and 12 years, respectively.1 tions under which a WTO member can, if it wishes, Second, China, by far the main target of existing counter-balance dumping, that is, the fact that for- antidumping measures,to date is one of the smallest eign firms exporting to the markets of this WTO users of such measures. However, the past decade member price their products more cheaply in these has shown how quickly large developing countries markets than in their own domestic markets. At that willing to use antidumping rules can become inten- time, the United States was still fighting to exclude sive users of this instrument, and the evolution of antidumping from the topics to be discussed at China's antidumping enforcement in 2002 and early the WTO Doha Ministerial, and the European 2003 raises legitimate concerns in this respect. Community (EC) was adopting an ambiguous The chapter is organized as follows. The first sec- position. tion describes the current situation. Antidumping is In the early stages of the negotiations under the used massively by only 10 countries (four industrial Doha Development Agenda (DDA), China finds and six developing), and a strong asymmetry, best itself in an unique situation on the antidumping and illustrated by China, exists between countries safeguard issues. (Safeguard rules define the condi- enforcing antidumping measures and those targeted tions under which a WTO member can, if it wishes by antidumping measures. The second section so, give a transitional relief to a domestic industry examines how China could minimize exposure to facing an unforeseen surge in competing foreign foreign antidumping cases--an option that would imports.) First, the WTO accession protocol of be a recipe for both trade success and China's lead- China includes very special provisions that China's ing role in reforming WTO antidumping rules. trading partners may use against Chinese exports. The third section analyzes China's antidumping I would like to thank Mike Finger, Edwin Vermulst, Will Martin, and two anonymous referees for very useful comments on earlier drafts of this paper. 29 30 China and the WTO regulations and first cases, including their crucial two main results. First, worldwide antidumping relations with the existing web of the U.S. and EC enforcement is thus highly concentrated in less antidumping cases. The fourth section examines the than a dozen countries. The top 10 antidumping opportunities that the DDA offers to China for nego- users enforce 90 percent of the antidumping meas- tiating stricter disciplines both on WTO contingent ures notified in the WTO; they represent 70 per- protection, and on the use of the NME and TPS pro- cent of the world GDP and 50 percent of world visions by China's trading partners. The conclusion trade. summarizes the crucial choices to be made by China Second, the situation prevailing during the with respect to antidumping and safeguard policies. Uruguay Round--antidumping users were almost exclusively industrial countries--is no longer the The Current Situation case. Six "new" antidumping-intensive users (all of them developing countries: Argentina, Brazil, During the November 2001 Doha WTO Ministerial India, Mexico, South Africa, and Turkey) have Conference, antidumping was perceived as an issue almost caught up with the four major "old" users. pitting developing countries anxious to discipline These new users implement more than one-third of the use of this instrument against the United States, the total number of antidumping measures in force which was (and still is) very reluctant to change its in 2002, compared with less than one-fourth in own antidumping regulations. However, a close 1995. Meanwhile, the share of measures of the four examination of the current situation suggests a old users has declined from more than two-thirds much more complex picture. This examination is to half of the total number of antidumping meas- based on the antidumping measures in force at the ures in force during the period. A last worrisome end of each year of the period 1995­2002 that are sign is that the rest of the developing countries, notified to the WTO Secretariat by members.2 while still small users individually, have doubled their global share of measures in force during the Antidumping Users vs. Targeted Countries: observation period.3 A Key Asymmetry Table 3.2 presents the stock of antidumping Table 3.1, which presents the stock of antidumping measures in force by targeted country for the measures in force by antidumping users, illustrates period 1995­2002. It shows a marked asymmetry TABLE 3.1 Top 10 Antidumping Users, Measures in Force, 1995­2001 Number of Measures in Forcea Average Average Number per Applied User Country 1995 1996 1997 1998 1999 2000 2001 2002 Importb Tariffc U.S. 265 271 271 281 282 202 227 239 0.29 4.3 EC 140 138 138 139 159 175 175 183 0.19 4.6 Australia 78 46 40 49 39 44 59 38 0.77 5.8 Canada 79 78 78 65 72 71 85 83 0.38 4.8 Mexico 93 92 81 86 80 77 61 55 0.72 12.6 South Africa 12 29 42 56 87 96 94 80 1.81 15.0 India 13 15 24 44 58 94 115 181 1.28 39.6 Argentina 15 30 33 39 45 42 45 60 1.17 13.7 Turkey 37 37 34 34 35 14 16 30 0.61 12.7 Brazil 21 28 23 28 38 43 49 54 0.51 12.5 All Others 50 59 84 102 122 117 97 -- 0.04 -- of which China -- -- 0 0 4 8 11 17 0.03 15.8 All countries 803 823 848 923 1017 975 1023 -- 0.21 -- a. Antidumping duties and undertakings in force as of 31 December of the year. b. Per thousand of US$ of the 1997 imports of the user country. c. Average applied tariffs (WTO 2001). Sources: WTO Reports on Antidumping (G/ADP/N series), WTO trade data 2001; Author's computations. China in the WTO: Antidumping and Safeguards 31 TABLE 3.2 Major Antidumping Targets, Measures in Force, 1995­2001 Number of Measures Targeted Number per Countries 1995 1996 1997 1998 1999 2000 2001 2002a Exportb U.S. 60 66 66 68 66 62 57 67 0.09 EC 77 88 89 102 132 149 99 (98) 0.13 Australia 5 6 5 6 5 5 5 5 0.08 Canada 19 19 19 19 20 18 8 8 0.08 Mexico 11 15 17 17 19 21 17 17 0.15 South Africa 7 10 11 11 12 15 16 24 0.39 India 15 15 15 21 29 35 42 44 0.72 Argentina 9 8 7 7 7 7 9 8 0.30 Turkey 9 9 6 8 10 13 12 18 0.37 Brazil 48 51 52 45 42 43 34 51 0.85 All others 543 536 561 619 675 611 724 -- 0.29 of which China 143 148 180 193 202 207 199 212 0.99 All countries 803 823 848 923 1017 979 1023 -- 0.22 a. Incomplete estimates for the EC. b. Per thousand of US$ of the 1997 exports of the targeted country. Sources: WTO Reports on Antidumping (G/ADP/N series); WTO trade data; Author's computations. between antidumping users and targets. The top 10 an indication of the incentives of these export users are the targets of less than one-third of all the interests to contribute to the opening of the mar- measures in force, with a strong amplification of kets of their country. These numbers can be use- this gap in 2001 and 2002. In sum, antidumping is fully compared to the symmetrical numbers of an instrument enforced by a few large countries antidumping measures in force by thousands of against the smaller economies of the rest of the U.S. dollars imported by the country in question world--hence, the absence of incentives coming (shown in table 3.1), which can be interpreted as an from the rest of the world that would induce the indication of the strength of the incentives existing antidumping-intensive users to restrain of these import-competing interests to induce their use of antidumping actions. their own government to use antidumping. The Combining tables 3.1 and 3.2 shows that for all observed imbalance between export interests and top 10 antidumping users, with the exception of import-competing antidumping beneficiaries in Brazil, the country's interests hurt by foreign anti- the top 10 antidumping users suggests that it also is dumping measures are smaller than those benefit- unlikely that domestic coalitions are strong enough ing from domestic antidumping protection. This is to support antidumping reforms in the WTO in based on the well-known economic proposition users that are key WTO players. that analyzes protection more as a conflict between This situation raises a question that should be domestic forces (namely export interests versus carefully examined in the future. Table 3.1 suggests import-competing interests) than as a conflict that one could reasonably argue that antidumping between countries. To capture this aspect, one measures enforced by the six major developing can calculate the number of foreign antidumping country antidumping users impose welfare costs measures in force imposed on exports from a top on their own domestic economies higher than the user adjusted by the size of the country's exports costs imposed on industrial economies by the (in thousands of U.S. dollars). Such a trade- developed country­imposed antidumping meas- adjusted number of measures (shown in table 3.2) ures for two reasons. First is the marked difference mirrors the intensity of the foreign pressures between the number of measures imposed by imposed on the export interests of a country, giving developing countries and industrial countries, 32 China and the WTO once adjusted by trade size. The average number China's Special Situation of measures in force per thousands of U.S. dollars Table 3.2 shows that China has been the main target of goods imported in 1997 by an antidumping of all the antidumping measures enforced in the user is a better indicator of the potential harm world--18 percent of the antidumping cases in done by antidumping on the domestic economy 1995, almost 20 percent in 2001­2002. Table 3.3 than the mere absolute number of measures. This refines this information by presenting the unad- indicator is much higher for developing countries justed number of antidumping measures imposed than for industrial countries: it ranges from 0.5 on Chinese exports by the top 10 antidumping (Brazil) to 1.8 (South Africa), whereas it ranges users, and this number once adjusted by the trade from 0.2 to 0.4 for industrial countries (with the value between each trade partner and China (in exception of Australia). These differences would be other words, the average number of cases imposed much larger if the number of antidumping meas- by foreign antidumping users per thousand U.S. ures to be adjusted were to take into account the dollars of exports from China to these users). Both number of tariff lines concerned, since developing sets of numbers show that China is much more countries tend to cover many more tariff items targeted by the developing countries than by the with antidumping cases than industrial countries. industrial countries--especially when one looks at The second reason for arguably higher welfare the figures adjusted for trade size. In fact, China is costs from antidumping measures in major devel- almost exclusively targeted by the top users, all of oping country users is that available (though not them relatively large economies. systematic) information suggests that antidump- The above information on antidumping users ing duties enforced by developing countries are, versus targets raises a key question about China's on average, more severe than those imposed by role in future world antidumping activities. Will industrial countries--and economic analysis China follow the same evolution as the other large shows that welfare costs increase more rapidly developing countries, that is, will it increase rapidly than tariffs. TABLE 3.3 Shares of Antidumping Measures in Force against Imports from China, 1995­2002 Average Shares of Antidumping Measures Against Chinaa Average Number per Share Export from User Countries 1995 1996 1997 1998 1999 2000 2001 2002 (excl. 2002) Chinaa U.S. 12.8 13.7 15.1 14.6 14.5 16.5 18.5 18.0 15.1 1.72 EC 20.7 21.7 23.2 23.0 20.8 19.4 19.4 15.8 21.2 1.04 Australia 9.0 4.3 10.0 8.2 7.7 4.5 5.1 7.9 7.0 0.71 Canada 7.6 7.7 10.3 10.8 8.3 8.5 10.6 10.8 9.1 1.83 Mexico 33.3 28.3 40.7 38.4 36.3 35.1 44.3 41.8 36.6 60.27 South Africa 8.3 27.6 28.6 23.2 19.5 18.8 19.1 22.5 20.7 11.98 India 38.5 46.7 33.3 27.3 32.8 22.3 25.2 27.1 32.3 10.66 Argentina 33.3 20.0 30.3 35.9 31.1 21.4 15.6 20.0 26.8 9.98 Turkey 13.5 13.5 14.7 17.6 17.1 14.3 25.0 40.0 16.5 4.37 Brazil 14.3 14.3 21.7 28.6 28.9 25.6 22.4 20.4 22.3 4.67 All Others 34.0 28.8 26.2 22.5 18.9 12.8 15.5 -- 22.7 -- Total (number)b 143 148 180 193 202 179 199 -- 174 -- in % of total 17.8 18.0 21.2 20.9 19.9 18.3 19.5 -- 19.4 -- a. in percent of the total number of antidumping measures in force imposed by the user country. b. Per thousands of US dollars of the 1997 imports from China of the user country. c. provisionally based on Lindsey and Ikenson 2001. Sources: WTO Reports on Antidumping (G/ADP/N series). WTO trade data; Author's computations. China in the WTO: Antidumping and Safeguards 33 the number of antidumping cases? Or will China ing this hypothesis is shown below. In such a case, adopt a different approach, that is, will China the worldwide spread of antidumping regulations minimize its own antidumping use and invest and the increase of antidumping measures are its negotiating strength in the WTO to obtain seen as a positive development by the firms in ques- stricter antidumping rules, as claimed by its chief tion, rather than as incentives to discipline anti- trade negotiator? Clearly, China's choice between dumping use. Examining such deeper aspects of these alternatives will have a decisive impact on the antidumping protection requires looking at the evolution of world antidumping enforcement and distribution of antidumping measures in force by on WTO trade disciplines. sector or product rather than by country. Minimizing China's Exposure A Few "Antidumping-Intensive" Sectors to Foreign Antidumping Table 3.4 provides a broad picture based on the China's second alternative will be a difficult exer- number and average shares of antidumping meas- cise, particularly in the coming years. The usual ures in force by the most aggregated section of the slowness of WTO negotiations means that China Harmonized Tariff System (instead of specific prod- could not get reforms of WTO antidumping rules ucts, as below). It shows that antidumping measures before two (and more likely four) years. In the are concentrated in a handful of Harmonized Tariff interim, it will be hard for Chinese authorities System (HTS) sections. Antidumping measures in to resist pressures from import-competing firms metals, chemicals, machinery and electrical equip- based in China that demand intensive use of China's ment, textiles and clothing, and plastics and rubbers antidumping procedures. This difficulty raises the amount to 75 percent of the total number of meas- following question: could China adopt policies ures, whereas trade in these sectors amounts to less minimizing, as quickly as possible, its exposure to than half of total world trade. These few HTS sec- foreign antidumping measures--thereby alleviat- tions happen to be key sources of exports for ing the political costs of playing a reforming role in dynamic developing countries in the first stages of WTO antidumping rules? their industrial development. Second, these HTS Trade problems are fundamentally domestic sections are characterized by a high proportion of conflicts between export-oriented and import- relatively standard products and by oligopolistic competing industries. Country-based data, as those market structures. The metals and chemicals sec- provided above, do not permit examination of this tions clearly fit these features.The other HTS sectors deeper aspect of protection. They can even be mis- require a closer look at subsectors.Few antidumping leading. For instance, they can suggest that the actions in the machinery and clothing subsectors are strong asymmetry between antidumping users and characterized by many firms and highly differenti- targets observed in tables 3.1 and 3.2 would ated products, and most of the antidumping actions progressively disappear because each WTO mem- in the electrical equipment and textiles subsectors ber would be induced to introduce antidumping are characterized by relatively standard products regulations and to enforce them--with such a and by oligopolistic firms.These features explain the worsening of the situation ultimately inducing wide use of antidumping measures in the textile WTO members to adopt collectively stricter disci- subsector since the late 1980s, before any effective plines. (Arguments have already been presented liberalization of the Multi-Fibre Arrangement quo- suggesting that the top antidumping users will tas,as decided by the Uruguay Round (GATT,1994). be unlikely to follow this path.) Taking into account Such a pattern strongly suggests that complain- firms' behavior suggests that an even darker ing firms use antidumping as an additional--cheap scenario is quite plausible. Firms that are the and powerful--instrument for segmenting the petitioners--hence the driving forces in antidump- markets that ongoing or scheduled trade liberaliza- ing enforcement--may well want to lodge similar tions aim at making more competitive. It also sug- antidumping complaints in several key countries to gests that the spread of antidumping cases to segment the world markets of their products certain sectors, such as clothing, should be expected through antidumping measures; evidence support- to the extent that these markets will be increasingly 34 China and the WTO c Lines ariffsT 0.0 1.2 5.9 0.0 68.6 78.2 30.0 12.6 18.0 43.4 12.1 26.1 15.7 29.2 30.1 61.9 21.5 96.3 25.0 67.2 25.0 38.7 <10% ariffT with not b are s verageA Applied ariffT 9.6 9.4 9.8 4.1 14.4 21.0 15.9 23.8 17.7 13.4 28.4 21.3 23.2 14.3 20.2 18.9 40.6 13.0 16.4 12.6 10.9 15.8 states China' 9.2 8.0 7.4 7.3 2.8 0.8 3.1 0.8 1.9 0.2 3.6 0.4 1.7 1.0 0.5 0.0 4.1 5.0 8.0 37.8 10.9 Import Pattern 00.01 member EC 6.7 4.7 3.2 3.7 1.6 0.7 2.1 2.1 4.8 3.4 7.1 1.1 1.7 0.1 7.3 0.0 0.3 0.1 2.0 29.2 19.8 Export Pattern 100.0 various by a 88.9 70.3 89.7 44.4 -- 66.7 -- -- -- product ADMs Growth 117.0 104.1 116.1 114.2 108.4 101.6 115.5 207.6 129.4 104.0 122.2 161.9 112.0 Indexes . same HS 9.2 6.4 6.1 4.0 4.0 3.1 3.1 2.1 1.6 1.5 1.4 8.0 5.0 5.0 5.0 2.0 1.0 0.0 0.0 the 33.5 21.6 of ADMs Pattern by Section 100.0 sector by 81 79 74 30 40 93 27 22 61 32 14 11 401 228 lines, imports 2000 1103 on es 1995­2000, 70 66 43 44 44 28 23 31 02 12 11 tariff 372 217 100 1999 1081 Measur total imposed 95 54 53 42 42 44 31 24 71 9 of 13 Sector 316 210 976 1998 cent ADMs by 94 61 55 40 37 12 9 per 29 19 10 14 in 3.3, orld Anti-Dumping 305 204 921 1997 c. to W of 85 60 54 40 37 71 3.1 31 19 16 11 15 the 290 203 899 2001, 1996 in in Number tables b. ce 81 48 53 42 34 81 6665 434677 999211 000999 333300 001111 000000 000000 35 15 21 14 14 276 201 874 with For 1995 in 1997­1995. contrast case. articles In EC and products Measures y cement, products oils 2001. one equip. products in equipment industr plaster paperboard, fats, instruments wood leather articles manufactured of y 1998­2000 Ikenson animal allied SystemSections electrical,y paper stone, products skins, and metal clothing rubbers transport foodstuffs headwear, of cinema aggregated Antidumping articles vegetable products ammunitions jeweller between animals, hides, objects Lindsey 3.4 Metals, Chemical, Machiner extiles,T Plasters, ehicles,V oodpulp, Articles W reparedP egetableV ood, Footwear Optical, Miscellaneous W Live Animal, Mineral Arms, Raw Gems, Art ce: Growth 6 7 4 2 9 1 3 5 8 Harmonized 15 16 11 17 13 10 12 18 20 19 14 21 otalT a. Sour systematically ABLET China in the WTO: Antidumping and Safeguards 35 subjected to product differentiation and imperfect sion but rather should expect to continue to face a competition (based on such factors as trademarks, large number of antidumping cases. Of course, goodwill, and distribution channels). In sum, the WTO membership gives China access to the WTO observed sectoral pattern of antidumping reflects dispute settlement mechanism, which could pro- the increasing privatization of trade policy by firms vide some relief to Chinese exporters harassed by that enjoy enough initial oligopolistic power to foreign antidumping. However, this relief probably fully use the pro-collusive bias embedded in will be only marginal and short-lived. For instance, antidumping regulations--a key lesson that should the 2001 WTO dispute settlement ruling banning be kept in mind when implementing these regula- the averaging method is already in the process of tions, in China as elsewhere. being circumvented by alternative procedures (hastily developed by creative petitioners) that may hurt the defendants differently, but not necessarily China's Sensitivity to Antidumping-Intensive less than the condemned averaging method. Sectors A more promising route for minimizing China's Does China tend to import products pertaining to exposure to foreign antidumping measures may be antidumping-intensive HTS sections (that is, to Chinese policies--and pre-eminently China's own industries subjected to many antidumping meas- trade policy. If China still imposes high tariffs in ures in the world)? If that is the case, one should antidumping-intensive sectors at the end of the expect Chinese authorities to be under strong pres- accession period, it may not adopt many or severe sures for taking and enforcing antidumping meas- antidumping measures on imports from the rest of ures on imports of such goods--that is, for the world in these products (though section 1 participating to the worldwide segmentation game shows that the top developing country antidump- in this sector that constitutes the ultimate goal of ing users have not hesitated to add high antidump- antidumping enforcement. Table 3.4 shows that the ing measures on the top of unfinished trade five most antidumping-intensive HTS sections rep- liberalization). However, such a costly tariff policy resent almost 70 percent of total Chinese imports, for China will not ensure less foreign antidumping opening the possibility that Chinese firms or foreign cases against Chinese exports. On the contrary, firms producing in China could table antidumping high Chinese tariffs will facilitate the introduction complaints to segment world markets. of foreign antidumping measures against Chinese China's sensitivity to antidumping-sensitive exports in antidumping-intensive sectors. This is sectors could also be assessed from an export per- because high Chinese tariffs will allow the existence spective. Table 3.4 underlines how sensitive Chinese of high prices in China, making the existence of exports are to the worldwide antidumping activity, dumping by Chinese exporters easier to prove, all particularly in machinery-electrical equipment and the more because the Chinese exports in question in textiles and clothing. The above observations on consist mostly of basic products, for which Chinese the key role of economic characteristics (type of exporters are likely to align their prices to those product, market structure) as the dominant factors prevailing in foreign markets since there is minimal driving world antidumping activity suggest China or no premium for differentiation. as a target of foreign antidumping measures much Accordingly, a first way to minimize exposure to more because of the intrinsic economic features of foreign antidumping measures would be the adop- her exported products (differentiation level and tion by China of a relatively uniform and moderate oligopolistic markets) than because it was not a tariff policy to reduce distortions in the domestic WTO member. production pattern, foreign antidumping investi- gators interpret such distortions as sources of dumping as much as possible. Table 3.4 provides the Chinese Policies as the Best Ways to Minimize average applied tariffs by HTS section that China Exposure to Foreign Antidumping imposed in 2001. Though averaging tariffs for such A key corollary of this conclusion is that China widely defined sectors gives very imperfect infor- should not expect to face less antidumping meas- mation on the tariff structure (peak tariffs within ures in the coming years because of her WTO acces- each sector are eroded by low tariffs), it suggests that 36 China and the WTO tariff peaks in China are concentrated in HTS sec- instrument that petitioners are successfully tions that are not antidumping-intensive activities, using in a strategic way to segment the two with the exception of textiles and clothing. largest world markets. This observation leads to Implementing a uniform tariff policy may an important corollary. Assessing the welfare help China to shift export products away from costs of antidumping measures should take antidumping-intensive sectors for another reason into account not only the severity of the meas- already observed in certain countries, such as Hong ures per se (the high level of antidumping Kong (China), Japan, the Republic of Korea, duties, or the restrictiveness of quantitative Singapore, and Taipei. These five industrial coun- restrictions), but also the fact that the level tries have reduced their exposure to foreign of competition prevailing in the importing antidumping measures by upgrading their markets is dramatically reduced by such exported products--to the point that one could measures. "Indirect" welfare costs generated by argue that, in fact, foreign antidumping measures antidumping-caused collusion compound the may have accelerated the economic development of direct welfare costs generated by the antidump- these industrial countries by inducing them to shift ing duties. faster to anticipated comparative advantages in · Second, U.S. antidumping duties are, on average, highly differentiated products. higher than EC duties: 104 percent and 38 per- cent, respectively. However, comparing in detail the measures in the EC and U.S. echoing cases is EC and U.S. Antidumping "Echoing" against China difficult because of the differences in regulations Exports and the lack of sufficiently detailed information. Table 3.4 relies on sectors too broad for capturing For instance, some EC cases are terminated by antidumping protection as embedded in the product the withdrawal of the complaint by the petition- strategies of a few large firms. This aspect requires ers. The effective impact of such withdrawals is information at the product level. Table 3.5 provides hard to assess. It may be limited to a "chilling" such detailed information for the EC and U.S. effect on Chinese exporters who may be forced antidumping actions against China. More precisely, to export less, or at higher prices, or both, Table 3.5 provides the list of the EC and U.S."echo- to limit the risks of facing new antidumping ing" cases, that is, those cases that have targeted the complaints. However, withdrawals may merely same products exported by China between 1980 and reflect a lack of cooperation of the domestic 1999. However, the European Community and the industry or a failure to get the minimum num- United States are the top antidumping users in terms ber of EC member states to support an EC Com- of the absolute number of measures in force, and mission proposal to impose measures. They also they represent the two largest markets for Chinese may reflect the fact that petitioners have been exports (15 and 21 percent, respectively). able to impose quantity or price restraints on Table 3.5 provides two general observations: Chinese exports on a "private" basis--with the corresponding full-fledged impact to be · First, the number of echoing cases is large: 58 expected from these hidden restraints. cases, that is, 75 and 68 percent of the total number of antidumping cases initiated against Chinese exports by the United States and the China's Antidumping Enforcement: European Community, respectively. All of these At the Crossroads cases echo each other generally within a year or less, though of course they subsequently may The following first assessment of China's anti- follow different developments in these markets. dumping regulations and enforcement procedures All but three (cycles, hammers, and pocket should be taken as provisional because it is con- lighters) resulted in antidumping measures of strained by the limited number of ongoing cases to some kind. Such a large proportion of echoing date; a substantial number of cases is often needed cases and the similarity of their outcomes con- for a robust assessment of the effective antidump- stitute signs that antidumping is a protectionist ing enforcement by a country. China in the WTO: Antidumping and Safeguards 37 TABLE 3.5 EC and U.S. Echoing Antidumping Cases against Chinese Producers, 1980­99 Decisions Dumping Antidumping Initiating Margin Positive Negative Duties Years Country Products (EC) EC U.S. EC U.S. EC U.S. 1992 EC antimony trioxide 43.2 noi 1991 U.S. antimony trioxide N 1988 EC barium chloride 50.1 D 25.8 1982 EC barium chloride 75.0 Ds 1983 U.S. barium chloride N 1983 U.S. barium chloride A 14.5 1999 EC brushes, hair 1992 U.S. brushes, hair T 1986 EC brushes, paint 100.0 Und 1992 EC brushes, paint noi 1985 U.S. brushes, paint A 127.1 1999 U.S. brushes, paint 1994 EC coumarin 50.0 Ds 1994 U.S. coumarin A 160.8 1991 EC cycles 30.6 D 30.6 1995 U.S. cycles N 1984 EC cycles, chains 45.0 Und 45.0 1999 EC cycles, forks 1999 EC cycles, frames 1996 EC cycles, parts 30.6 D 30.6 1999 EC cycles, wheels 1992 EC ferrosilicon 49.7 D 49.7 1992 U.S. ferrosilicon A 137.7 1995 EC furfuryl alcohol 1994 U.S. furfuryl alcohol A 45.3 1999 EC glycine 1994 U.S. glycine A 155.9 1985 EC handtools: hammers 1990 U.S. handtools: hammers A 45.4 1994 U.S. lighters, disposable N 1990 EC lighters, pocket 16.9 D 16.9 1991 EC magnesium oxide 27.7 Und 1994 U.S. magnesium, alloy A 79.4 1994 U.S. magnesium, pure A 108.3 1997 EC magnesium, unwrought 31.7 D Und 31.7 1994 U.S. manganese, metal A 143.3 1992 EC manganese, unwrought 1996 U.S. persulfates 1994 EC persulfates, peroxodisulfates 110.1 D 83.3 1982 U.S. polyester, cotton cloth A 36.2 1990 EC polyester, yarn 23.5 D 23.5 1986 EC potassium permanganate 94.5 D Und 28.0 1983 U.S. potassium permanganate A 39.6 1984 EC silicon carbide 31.5 Und 1993 U.S. silicon carbide N 1989 EC silicon, metal 38.7 Ds 18.7 1991 EC silicon, metal 178.0 Ds 1990 U.S. silicon, metal A 139.5 38 China and the WTO TABLE 3.5 (Continued) Decisions Dumping Antidumping Initiating Margin Positive Negative Duties Years Country Products (EC) EC U.S. EC U.S. EC U.S. 1994 EC steel, pipe or tubes fittings 58.6 D 58.6 1999 EC steel, pipe or tubes fittings 49.4 D 49.4 1991 U.S. steel, pipes A 182.9 1985 U.S. steel, pipes N 1988 EC tungstate, ammon. para- 75.7 noi 1988 EC tungsten, carbide 73.1 D Und 1988 EC tungsten, metal powder noi 1991 U.S. tungsten, ore A 151.0 1989 EC tungsten, ores 50.3 D Und 1988 EC tungstic, oxide & acid 85.8 D Und Total 58 Average margins and duties 59.2 29 15 5 7 37.8 104.5 Sources: Bloningen 2001 for the U.S.; Official Journal of the EC. Notes: EC decisions: D ad valorem duty; Ds specific duty; Und undertaking; with withdrawal; noi no injury. ep expired deadline. U.S. decisions: A affirmative; N negative. China's Antidumping Regulations complaints (a condition that domestic monopo- lies, oligopolies, or cartels fit much more easily China adopted its first antidumping regulations on than do competitive industries) March 25, 1997, and the antidumping guidelines · possibility of ex officio initiation of cases by the necessary for implementing the law later the same Chinese authorities year. Following accession to the WTO, this "old" · screening of the complaints by the antidumping regulation was replaced by the "new" regulation in office, exposing this office to strong and hidden January 2002 with another set of guidelines (see, pressures by vested interests on the WTO website, G/ADP/N/1/CHN1 and · possibility of withdrawal by the petitioners, G/ADP/ N/1/CHN2; see also Wang 2003 for a legal facilitating private collusion between them and analysis). the defendants Chinese regulations follow the usual structure of · cumulation of imports, facilitating the demon- antidumping legislation, including proof of the stration of injury and widening the geographical existence and estimate of the magnitude of dump- scope of protective measures ing and of material injury, and proof of the causal · recourse to constructed normal values, enabling relation between dumping and injury. However, manipulation of costs and reasonable profit in they have a few striking features. First, many details case of the absence of comparable prices in the are left to the detailed guidelines or to case-by-case exporting country practices. This is often the case among countries · broad definition of the confidentiality of the that have recently adopted antidumping regula- information limiting the rights of the defen- tions. However, it has the great inconvenience of dants (and arguably having equal impacts on generating a high level of legal uncertainty in the domestic and export firms) process as a whole. · possibility of imposing undertakings as anti- The second feature is that all the well-known dumping measures and the mandatory provi- protectionist biases of the WTO antidumping pro- sion that antidumping duties shall be borne by visions are included in China's regulations, as the importers (the no-absorption provision) shown by the following non-exhaustive list: · possibility of imposing retroactive antidumping · use of the concept of "a major proportion" of duties in the case of a history of dumping, that the industry as the threshold level for accepting is, recurrent antidumping complaints China in the WTO: Antidumping and Safeguards 39 · possibility of taking "appropriate" measures in WTO for the cases under the new regulation. case of circumvention of the antidumping Sixty-nine cases have been initiated, with a very measures by the foreign firms. uneven time pattern: after a slow start in 1997­98, the number of cases has increased rapidly (some The third feature of the Chinese antidumping cases under the Old Regulation are left with regulations prior to the 2003 government reform unknown status), reaching 24 cases in 2002 and 11 was a highly complex coordination of the various cases for the first five months of 2003. This increase administrative agencies involved in antidumping can only be a source of serious concern. It remains investigations. The Fair Trade Administration to be seen whether it simply mirrors cases that for Import and Export (FTA) of the Ministry were "in the pipeline" for a long time, or whether of Foreign Trade and Economic Cooperation China has begun to follow the drift observed in the (MOFTEC) received complaints, decided those it six top developing country antidumping users and would accept or reject, and was involved in the will soon become another antidumping intensive whole investigation and process. The State Eco- user, putting in danger her so-far successful nomic and Trade Commission (SETC) joined liberalization. MOFTEC for determining the existence of injury at Although it is too early to know whether the preliminary stage and for the final investiga- China's first antidumping measures will be repre- tions. The Customs General Administration (CGA) sentative of future antidumping enforcement, joined MOFTEC for some parts of the investiga- these measures deserve a few observations. First is tions. For the imposition of antidumping duties, that antidumping measures have been taken in MOFTEC made proposals to the Tariff Commis- almost all the cases. This is a very high percentage, sion under the State Council (TCSC), which took compared to generally observed levels of 60­70 per- the decisions. The merger of SETC and MOFTEC cent in industrial countries. The second observa- in early 2003 brought these two functions into one tion is the relatively high level of the measures ministry. adopted by the Chinese authorities, although it Lastly, the Old Regulation included the follow- seems that the most recent measures are less severe ing Article 40: "In the event that any country or than the ones taken under the Old Regulation. region applies discriminatory antidumping or The main countries targeted are industrial and countervailing measures against the exports from advanced developing countries--not quite the pat- the People's Republic of China, the People's tern observed for the other developing country Republic of China may, as the case may be, take antidumping users. counter-measures against the country or region in The cases initiated since 2001 deserve an addi- question." It is not known whether this provision tional remark. Their product pattern is closer to has led to cases, but clearly it opened the possibil- what is observed for the other antidumping users, ity for China to use her antidumping rules as a as best illustrated by the steel cases (echoing the EC retaliatory instrument. Interestingly, Article 56 of and U.S. safeguards) and by the ethanolamine cases the New Regulation is only slightly more diplo- (observed in several other antidumping users). This matic: "Where a country (region) discriminatorily increasingly similar product pattern suggests that imposes antidumping measures on the exports China's antidumping enforcement may be starting from the People's Republic of China, China may, on to be part of the ongoing process of segmenting the basis of the actual situations, take correspon- world markets through worldwide antidumping ding measures against that country (region)." activity. It also raises the issue of a progressive cap- ture of China's trade policy by firms, similar to what is observed in the 10 major antidumping Antidumping Enforcement by China users. In this context, it would be important to Table 3.6 provides the list of the antidumping cases know whether petitioners in China are Chinese initiated by China between 1997 and May 2003, firms (private or state-owned) or whether they are based on information provided by MOFTEC for firms having strong links (such as joint ventures, the cases under the old regulation and on China's technical relations, vertical integration) with for- latest notification (G/ADP/N/105/CHN) to the eign firms well experienced in "antidumping art." 40 China and the WTO TABLE 3.6 China's Antidumping Measures in Force and Investigations, as of 30 June 2003 Provisional Definitive Antidumping Antidumping Duties (percent) Duties (percent) Initiation Year Country Products Minimum Maximum Minimum Maximum Cases initiated under the old regulation 1997 Canada newsprint 17.1 78.9 9.0 78.0 1997 Korea, Rep. of newsprint 17.1 78.9 9.0 78.0 1997 U.S. newsprint 17.1 78.9 9.0 78.0 1999 Russia steel, cold-rolled silicon 11.0 73.0 0.6 62.0 1999 Korea, Rep. of polyester film 21.0 72.0 13.0 46.0 1999 Japan steel, cold-rolled stainless 4.0 75.0 17.0 58.0 1999 Korea, Rep. of steel, cold-rolled stainless 4.0 75.0 17.0 58.0 1999 EC Germany acrylates 24.0 74.0 0.0 0.0 1999 Japan acrylates 24.0 74.0 31.0 69.0 1999 U.S. acrylates 24.0 74.0 31.0 69.0 2000 EC Britain methylene chloride 7.0 39.0 6.0 39.0 2000 EC France methylene chloride 28.0 75.0 28.0 75.0 2000 EC Germany methylene chloride 67.0 67.0 66.0 66.0 2000 EC Netherlands methylene chloride 10.0 58.0 9.0 57.0 2000 Korea, Rep. of methylene chloride 7.0 28.0 4.0 28.0 2000 U.S. methylene chloride 49.0 58.0 49.0 58.0 2001 Japan polystyrene a a a a 2001 Korea, Rep. of polystyrene a a a a 2001 Thailande polystyrene a a a a 2001 Indonesia lysine a a a a 2001 Korea, Rep. of lysine a a a a 2001 U.S. lysine a a a a 2001 Korea, Rep. of polyester, chips a a a a Cases initiated under the new regulation 2001 Korea, Rep. of polyester, staple fiber 4.0 48.0 2.0 48.0 2001 Korea, Rep. of pet chips 6.0 52.0 5.0 52.0 2001 Indonesia acrylates 11.0 24.0 11.0 24.0 2001 Korea, Rep. of acrylates 11.0 20.0 2.0 20.0 2001 Malaysia acrylates 13.0 38.0 4.0 38.0 2001 Singapore acrylates 46.0 49.0 30.0 49.0 2001 EC Belgium caprolactam 6.0 16.0 6.0 16.0 2001 EC Germany caprolactam 28.0 38.0 28.0 38.0 2001 EC Netherlands caprolactam 9.0 18.0 6.0 18.0 2001 Japan caprolactam 5.0 21.0 5.0 18.0 2001 Russia caprolactam 6.0 29.0 7.0 16.0 2002 Korea, Rep. of polyester, film -- -- 0.0 0.0 2002 India anhydride, purified 33.0 33.0 b b 2002 Japan anhydride, purified 66.0 66.0 b b 2002 Korea, Rep. of anhydride, purified 14.0 33.0 b b 2002 Japan styrene butadiene 0.0 33.0 b b 2002 Korea, Rep. of styrene butadiene 10.0 27.0 b b 2002 Russia styrene butadiene 16.0 46.0 b b 2002 Kazakstan steel, cold-rolled products 21.0 48.0 b b 2002 Korea, Rep. of steel, cold-rolled products 9.0 40.0 b b China in the WTO: Antidumping and Safeguards 41 TABLE 3.6 (Continued) Provisional Definitive Antidumping Antidumping Duties (percent) Duties (percent) Initiation Year Country Products Minimum Maximum Minimum Maximum 2002 Russia steel, cold-rolled products 9.0 29.0 b b 2002 Taiwan (China) steel, cold-rolled products 8.0 55.0 b b 2002 Ukraine steel, cold-rolled products 12.0 22.0 b b 2002 Japan polyvinyl chloride 32.0 115.0 b b 2002 Korea, Rep. of polyvinyl chloride 10.0 76.0 b b 2002 Russia polyvinyl chloride 34.0 82.0 b b 2002 Taiwan (China) polyvinyl chloride 10.0 27.0 b b 2002 U.S. polyvinyl chloride 25.0 83.0 b b 2002 Japan toluene 19.0 49.0 b b 2002 Korea, Rep. of toluene 6.0 22.0 b b 2002 U.S. toluene 23.0 28.0 b b 2002 Japan phenol 7.0 144.0 b b 2002 Korea, Rep. of phenol 10.0 10.0 b b 2002 Taiwan (China) phenol 7.0 20.0 b b 2002 U.S. phenol 29.0 29.0 b b 2003 EC Germany ethanolamine b b b b 2003 Iran ethanolamine b b b b 2003 Japan ethanolamine b b b b 2003 Malaysia ethanolamine b b b b 2003 Mexico ethanolamine b b b b 2003 Taiwan (China) ethanolamine b b b b 2003 U.S. ethanolamine b b b b 2003 EC chloroform b b b b 2003 India chloroform b b b b 2003 Korea, Rep. of chloroform b b b b 2003 U.S. chloroform b b b b 69 All cases average antidumping 17.2 43.4 9.8 23.7 duty a. Status unknown; b. Ongoing investigations. Sources: MOFTEC, WTO Semi-annual report G/ADP/N/105/CHN, 22 August 2003. China's Options in the WTO sound "interpretation" of the specific provisions on Negotiations on Contingent antidumping. Such an interpretation could create Protection strong incentives in China for both restraining its own antidumping use and fighting for stricter China's antidumping enforcement needs to take WTO rules on antidumping. It also could give into account an aspect that the six top developing strong incentives to China's trading partners to ease country antidumping users have not had to China's transitional period of accession. It is fair to address: China's WTO accession protocol incorpo- say that dealing with the provision on safeguards rates specific provisions on antidumping and safe- seems more difficult. This provision is so much in guards (for a legal analysis, see Vermulst 2000). The contradiction with the general WTO rules and section argues that this special feature that is ini- spirit that its use will raise a large systemic risk for tially a handicap can be turned into a positive the whole WTO. However, various options are instrument. More precisely, it could induce China examined at the end of the section. to negotiate, in the DDA, a more economically 42 China and the WTO Linking China's Effective Liberalization to a Better negotiate an economically sound interpretation of Treatment of Exports by Antidumping Users the use of the NME status by WTO members. The benefits from such an interpretation rely on the fact China's protocol of accession allows China's trading that China and her trading partners have a com- partners to use the non-market economy (NME) mon interest in establishing the strongest possible status in their investigations against allegedly link between Chinese effective liberalization and dumped Chinese exports. This NME status, which the elimination of the NME use by foreign is scheduled to last 15 years until 2017, allows for- antidumping authorities against Chinese exporters. eign antidumping investigators to use proxies for It is important to underline that what is at stake is estimating the home market prices or costs of not the elimination of the NME provision itself Chinese exporters. Such proxies make the proof (once again impossible to obtain) but its effective of the existence of antidumping much easier use in the future.5 than under the antidumping rules for market The argument aims at mobilizing the export economies, and they inflate the magnitude of the interests in both China and the rest of the world estimated antidumping margins compared to during the implementation period of China's those, already high, imposed on market economies. accession to the WTO. On the one hand, if Chinese Table 3.7 gives a sense of the intrinsic biases of exporters know that they will face a less unfair the NME procedure by summarizing the informa- treatment (i.e., that they will not be subject to the tion available from 208 EC and U.S. antidumping NME status) in foreign antidumping cases, they cases initiated between 1995 and 1998 (Lindsey will be induced to monitor China's liberalization 1999; Messerlin 2000a). It shows that the further more closely--to check whether China is effectively from pure price comparisons the methodology opening her markets in accordance with her acces- used for estimating dumping margins stays, the sion protocol--and to provide stronger support to higher these margins have been: from 3 percent it, including stricter use of China's antidumping (U.S.) and 22 percent (EC) under pure price com- regulations. On the other hand, if foreign exporters parisons, to 14 percent (U.S.) and 24 percent (EC) are convinced that they will get more effective and under a mix of price comparisons and constructed stable access to Chinese markets, they will be values, and to 25 percent (U.S. and EC) under the induced to fight for restraining the antidumping various constructed value methods. Moreover, use enforcement by their own authorities, in particular of NME status is clearly linked to the highest for limiting the use of the NME status. dumping margins found (40 percent in the U.S. and 46 percent in the EC).4 Doing so could be achieved by implementing the following simple rule. Foreign antidumping It is almost impossible to eliminate a provision investigators will automatically grant the market- included in a country's protocol of accession. economy status to Chinese exporters of any given However, during the DDA, it may be possible to TABLE 3.7 Do Antidumping Investigations Really Look at Dumping? U.S. Cases (1995­98) EC cases (1995­97) Number Average Number Average Bases of Estimated Normal Values of Cases Dumping of Cases Dumping of Exporters Examined Margins Examined Margins Price comparisons only 5 3.2 8 22.7 Price comparisons and constructed values 33 14.2 33 24.4 Constructed values 20 25.1 12 25.1 Non-market economies 47 40.0 12 45.6 "Best available facts" 36 95.6 2 74.5 All cases examined 141 44.7 67 29.6 Sources: Lindsey 1999 for the U.S. cases; Messerlin 2000a for the EC cases. China in the WTO: Antidumping and Safeguards 43 product that would meet the three following partners should grant to China is the unconditional conditions: benefit of market economy status in the antidump- ing investigations faced by the Chinese exports · The Chinese MFN tariff on the product involved meeting these conditions.6 is moderate (e.g., 10 percent or less). This thresh- Three final observations are necessary. First, the old tariff will be one of the core components of a negotiations on the improved implementation of more economically sound "interpretation" of the NME status for China should be as swift as pos- China's NME status to be agreed upon during sible. Second, the conditions can be easily defined the DDA. It could be stable over time, or it could on a tariff line (HTS) basis. For instance, China increase as time passes--showing an increasing could notify the WTO, on a regular basis, of the confidence in the ongoing Chinese liberalization tariff (HTS) lines for which these conditions are process among China's trading partners. met; this procedure could readily be included in the · No "core grey measures" shall be imposed on the general monitoring procedures of China accession product in question by the Chinese authorities. process. Cross-notifications by China's trading The list of the core gray-area measures to be partners could be added to the process, under the introduced in the interpretation agreement also condition that they will not slow it. Finally, weaker should be negotiated during the DDA. It should variants of the above suggestions could be exam- be short (e.g., specific tariffs, quantitative ined, if necessary. For instance, the NME status restrictions, and minimum prices) and only the could only be eliminated for the goods notified measures listed should be considered as parts of after a specified period, for example, one year the conditions. instead of immediately. Suggesting weaker variants · No state-owned monopolies shall distribute in is beyond the scope of this paper and is left to trade China the competing foreign and Chinese vari- negotiators. However, it is worth noting that any eties of the product in question. Chinese state- weakening of the suggested approach may have owned sole producers are acceptable because, huge costs in terms of decreasing incentives for as shown by economic analysis, a protection export interests in both China and the rest of the granted exclusively by a moderate tariff elimi- world to support the transition process of the nates the risk of monopoly power of the domes- Chinese accession to the WTO. tic sole producer. Stricter Rules on Antidumping Table 3.4 shows that Chinese ad valorem tariffs applied in 2001 lower than 10 percent amount to Ambassador Long Yongtu's desire to introduce 38 percent of all the tariff lines. The fact that the stricter rules on antidumping in the WTO is a Chinese tariff schedule has only roughly 7,000 tariff "natural" extension of the negotiation on the NME lines suggests that it does not offer many opportu- use proposed above because it focuses on the nities to create narrow niches of protection for antidumping rules faced by allegedly dumped carefully defined tariff items. Applying the above exports from market economies. China's efforts to conditions thus would substantially reinforce the introduce stricter rules on WTO antidumping could rights of Chinese exporters in the antidumping follow two very different approaches. cases lodged in these two HTS sections. However, A cautious approach would be to merely table a China should take some initiatives for improving series of proposals or to support those already the situation in the other antidumping-intensive tabled in Geneva by a few WTO members for HTS sections, in particular in textiles and clothing. improving WTO-based antidumping regulations at The rationale for the three conditions is that the margin. For instance, the following suggestions, they make unlikely the fact that Chinese exporters derived from proposals tabled by the Swedish would dump for other reasons than those econom- Kommerzkollegium 1999, could receive China's ically sound, such as differences in demand pattern, support: and need to meet foreign demand and to make Chinese products known in foreign markets. · Dumping should be the principal cause of mate- Hence, the minimum that WTO China's trading rial injury. 44 China and the WTO · Double protection (for instance, antidumping much easier for WTO members to impose safeguard measures imposed on the top of quantitative measures against Chinese exports during the next restrictions) should not be allowed. 12 years, until 2014 (for a detailed description from · Measures should last a maximum of five years a legal perspective, see Andersen and Lau 2001). (implying stronger limits to review). All the terms defining the use of a safeguard action · Repeated initiations in a short period of time in the traditional GATT-WTO context (under should not be allowed. Article XIX) have been systematically weakened: no · Cumulation of imports from different countries requirement of "unforeseen circumstances," no should be banned or severely restricted, unless MFN requirement, the need for "material" rather they come from the same firms or from the sub- than "serious" injury, fewer factors related to the sidiaries of the same firms. condition of the domestic industry, a weaker causal · Aggregating products under the"one single prod- link between increased imports and injury, the uct"procedure should be severely restricted. absence of a non-attribution causation analysis. · All zeroing practices (only export transactions The most important--and potentially the most that have been found to be dumped are used to devastating for the WTO--provision of Section 16 calculate dumping margins) should be banned; of the WTO China Protocol states that WTO mem- all export transactions should be included in the bers have the unprecedented possibility to use a investigation. "trade-diversion" clause. This means that, as soon · The antidumping authorities should produce as one WTO Member implements a TPS measure disclosure documents. against Chinese exports, all the others could · The use of the de minimis rule should be enforce a similar measure at almost no cost in expanded in an economically sound way. terms of legal procedures (e.g., no investigation, no prior notification, no input from Chinese parties). Alternatively, China could adopt a bolder In short, the trade-diversion clause makes almost approach. For instance, reforming antidumping unchallengeable the hypothesis that Chinese could take, as often as possible, the form of negoti- exports will be diverted from the first closed market ated "quantitative thresholds" (Messerlin 2000b). to the rest of the world. WTO members could agree that no antidumping All these features put the TPS largely in contra- measure should be imposed in cases where the diction with the WTO's usual deep concerns about level of injury losses is less than an agreed thresh- a fair balance of rights and obligations. It is a provi- old of the complainants' revenues of the year(s) sion so unbalanced that one can fear that its use used as the reference (pre-dumping) period. An will trigger feelings in China close to those associ- approach based on quantitative thresholds is con- ated to the "Unequal Treaties" of the nineteenth ceptually equivalent to a tariffication process. This century. As a result, it represents a serious systemic approach tends to give a sense of the magnitude risk to the WTO regime. of the concessions granted by both sides, bringing One may argue that the TPS is such a politically it more in line with usual WTO negotiating tech- aggressive instrument that one may wonder niques. It is also flexible enough to permit incre- whether it will be ever used. This argument is far mental reforms and to deliver the progressive from convincing. Its politically explosive content liberalization that WTO members seek through makes the TPS unlikely to be initiated by a WTO progressive increases of the thresholds, thereby member other than a very large industrial country avoiding the current deadlock of binary choices (the United States, the European Community, or between fully enforcing antidumping regulations Japan). However, all the other WTO members that or rejecting them totally. would have waited for such a move would benefit from the "trade diversion" clause as soon as a large industrial country takes the lead. The TPS Provision and the General Issue What are the possible actions left to China if the of Contingent Protection in the WTO TPS were to be invoked? A first possibility would be Section 16 of the China Protocol of Accession cre- to negotiate the same "preemptive" approach as for ates the TPS mechanism. The TPS makes it legally the NME provision by negotiating during the Doha China in the WTO: Antidumping and Safeguards 45 Round an interpretation agreement on the effective use the dispute settlement mechanism as soon as a use of this procedure. WTO members would agree WTO member notifies TPS use to the WTO Secre- not to use the TPS when Chinese products would tariat. Lawyers tend to overstate the benefits of such meet the three conditions listed above. Rather, they an approach, by ignoring the full development of would use the normal WTO safeguard provision the trade conflicts. It is almost certain that WTO under Article XIX. dispute settlement cases dealing with the TPS Putting antidumping and safeguard regulations would leave the two parties in a difficult political on a par would make a lot of sense from the per- situation. A more aggressive approach by China spective of the global WTO architecture. Most would be based on Article 56 of the New Regula- WTO members use antidumping as a substitute for tion. However, such an approach should not ignore safeguards, with many antidumping measures in the basic principle of deterrence: trade deterrence, fact being safeguard actions dealing with industries as nuclear deterrence, is good as long as it remains a in difficulties. In fact, the TPS provision expands threat--as long as it stops short of being an effec- China's strong stake in substantial improvements of tive action. antidumping rules to the whole WTO contingent protection regime--that is, antidumping and Conclusion safeguard regulations. As a result, during the DDA, China may not only be very active in the antidump- The chapter provides two results of prime impor- ing negotiations but could also try to expand these tance for China. First, the countries that would gain negotiations to safeguards--so far not explicitly most from better disciplines on antidumping rules included in the Doha negotiating program--to are the few developing countries which, since the make the whole WTO contingent protection regime 1995 Uruguay Round, are intensive antidumping more consistent. users. Because antidumping measures imposed by A promising way to improve the current WTO these developing countries tend to be more fre- Safeguard Agreements would require tying together quent and severe than those imposed by industrial the concept of temporary protection embedded in countries, they hurt developing countries' domestic safeguards and the basic concept of renegotiation economies much more than industrial country- under GATT Article XXVIII (Messerlin 2000b). imposed antidumping measures harm theirs. If For instance, at the end of the second period of China wants to continue to enjoy successful liberal- enforcing a safeguard measure allowed by the cur- ization, it should not become an intensive anti- rent Safeguards Agreement (based in turn on dumping user. GATT Article XIX), the country shall be requested Second, few economic and political forces would to choose between two alternatives: either to rene- introduce some automatic counterweights and gotiate the tariff of the product subjected to the restraints to the situation currently existing in terms safeguard measure in question, or to eliminate the of antidumping enforcement. Major current users safeguard measure (using then antidumping or any have few incentives to reform their very discrimina- other trade remedy should be prohibited since all tory use of the antidumping instrument, mostly the instruments of contingent protection are sub- targeting other countries than the top 10 users if stitutable). The logic of the mandatory aspect of one accepts, of course, the economically sound such a choice is that current safeguard and welfare-based incentive that protection is harmful antidumping procedures transform what should be to the economy of the country imposing it. Mean- transitory protection into permanent protection. while, smaller countries have few incentives to use The possibility remains that such a global the antidumping instrument in a retaliatory move. approach to antidumping, safeguards, and TPS As a result, China is at crossroads--and with measures may face entrenched hostility from WTO China, so is the rest of the world, and in particular, members, in particular from the top 10 antidump- the United States and the European Community, ing users. In such a case, the remaining options for the world's largest economies. On the one hand, China would rely on some kind of threats of retali- China may be increasingly tempted to use anti- ation. The least aggressive approach would be for dumping rules more intensively for several reasons: China to announce its intention to systematically as a back door of old-time protection, at the risk of 46 China and the WTO unraveling her scheduled trade liberalization; as a Notes progressive integration in the worldwide collusive 1. In Spring 2002, the EC and the United States declared that dimension of antidumping, used as an instrument they will consider Russia as a market economy for purposes of for segmenting world markets for the benefit of antidumping investigations. However, the case for the introduc- tion of a TPS provision in Russia's accession protocol seems still large firms; and as a retaliatory instrument against open. It would be interesting to make a parallel between the con- foreign antidumping measures. This last reason is ditions imposed on Japan's accession to GATT, and those specific to China, and it deserves a comment. Retal- imposed on China in its accession to the WTO. iatory antidumping measures can have some appeal 2. Tables 3.1 to 3.3 treat measures taken against different individual member-states of the EC as one aggregated measure only for countries with markets large enough to if adopted at the same time and for the same product (data generate some deterrence power. Chinese markets for 2002 are still not totally complete). By contrast, table 3.4 are not yet large enough to meet this condition. follows the notifications of EC trading partners that vary in their treatment of the EC (as one entity or as a set of distinct However, strong Chinese growth will change this member-states). situation relatively rapidly, and China's increasingly 3. Are antidumping measures used by developing countries credible threat of antidumping retaliation implies as a "safety valve" for softening the impact of their ongoing trade liberalizations? Rather, comparing the intensity of antidumping that WTO disciplines on antidumping should use with three indicators capturing key aspects (simplicity, irre- become increasingly attractive to major trading versibility, and openness, as estimated by Laird and Messerlin partners (if the United States and the European (2002) of the trade policy of the six developing countries, which Community are not myopic). Since WTO Rounds are intensive antidumping users, suggests a correlation between antidumping intensity and a flagging trade liberalization. tend to last for extended periods, this evolution 4. For instance, under the NME status, it is possible to use should be taken into account by the United States industrial countries (such as the United States or Sweden) as and the European Community as early as the cur- "reference" countries for China. That introduces systemic errors about the product and/or the production process. For instance, rent Doha Round--thereby triggering a potentially it makes no sense to consider, without deep economic analysis, "virtuous" dynamic game. the calcium metal produced in small quantities by a U.S. On the other hand, being still a small antidump- monopolist for its own use as similar to the calcium metal produced by China and Russia in large quantities for sales on ing user and a key target of foreign antidumping international markets. The U.S. product is likely to have charac- measures, China has expressed the desire to impose teristics in terms of quality and/or availability making it very stricter WTO antidumping rules. It will be one of different from the Russian or Chinese calcium metal, and it is the main beneficiaries of such a move because it sold and bought in a market structure very different from the markets of its Russian and Chinese counterparts. In the same would allow it to preserve its successful trade liber- vein, trying to estimate production costs by combining input alization. Such a move may have another key posi- prices in industrial countries and input quantities used in a tive outcome: it will help China to negotiate an developing country makes little economic sense. 5.Antidumping authorities in the EC and the Untied States in economically sound interpretation of the special some cases already adopt more liberal interpretations vis-a-vis provisions on antidumping and safeguard included Chinese exporters than the Protocol allows them. In the European in its WTO accession protocol. This new interpreta- Community, for example, this would be the case where the authorities accept individual treatment or market-economy tion of these special provisions should be based on status with respect to individual Chinese exporters. a few key and economically sound conditions to 6. It could be argued that market forces in China for these be met by China: low tariffs, no "core gray" meas- relatively unprotected products could be distorted by Chinese ures, and no distribution monopolies. Such an regulations on inputs for such goods (for instance, subsidies). However, there are WTO instruments for addressing such interpretation relies on the following strong eco- an argument (which could be applied to most China trading nomic and political argument. China and its trad- partners). ing partners have a common interest in establishing the strongest possible link between Chinese effec- tive liberalization and the elimination of the use References of these special provisions by foreign authorities Andersen, Scott, and Christian Lau. 2001. "Hedging Hopes with against Chinese exporters. In other words, the sug- Fears in China' Accession to the World Trade Organization: gested interpretation of the specific antidumping The Transitional Special Product Safeguard for Chinese Exports." Mimeo. Geneva: Powell, Goldstein, Frazer and and safeguard provisions included in China's acces- Murphy. sion protocol is an effort to mobilize the export Bloningen, Bruce. 2001. Database on U.S. Antidumping Cases. interests in both China and the rest of the world (www.uoregon.edu). GATT Secretariat. 1994. The Results of the Uruguay Round during the difficult implementation period of the Multilateral Trade Negotiations. Geneva: World Trade Chinese accession to the WTO. Organization. China in the WTO: Antidumping and Safeguards 47 Kommerzkollegium. 1999. Further Actions within the EU in Messerlin, Patrick A. 2000b. "Antidumping and Safeguards." In the Field of Antidumping. Stockholm: Kommerzkollegium. Jeff Schott, ed. The WTO After Seattle. Washington, D.C.: Photocopy (17 June). Institute for International Economics. Laird,Sam,and Patrick A.Messerlin.2002."Trade Policy Regimes Vermulst, Edwin. 2000. "Contingent Protection Provisions in and Development Strategies: A Comparative Study." Mimeo. China's Draft Protocol on Accession to the WTO." Seminar Washington, D.C.: InterAmerican Development Bank. on Antidumping and Safeguards. Department of Interna- Lindsey, Brink. 1999. The US Antidumping Law: Rhetoric versus tional Trade and Economic Affairs. Beijing: Ministry of Reality. Washington, D.C.: Cato Institute. Foreign Trade and Economic Cooperation. Lindsey, Brink, and Dan Ikenson. 2001. Coming Home to Roost: Wang, Lei. 2003. "China's New Antidumping Regulations: Proliferating Antidumping Laws and the Growing Threat to Improvement to Comply with the WTO Rules." Mimeo US Exports. Washington, D.C.: Cato Institute. (1wang@ealawfirm.com). Beijing: East Associates. Messerlin, Patrick A. 2000a. "An Economic Perspective on World Trade Organization (WTO). 2001. Market Access: Unfin- Antidumping and Safeguard." Seminar on Antidumping ished Business, Post-Uruguay Round Inventory and Issues. and Safeguards. Department of International Trade and Special Studies 6. Geneva. Economic Affairs. Beijing: Ministry of Foreign trade and Economic Cooperation. 4 Intellectual Property Rights in the WTO Accession Pack age: Assessing China's Reforms Keith E. Maskus After a long period of rapid growth and significant property laws. Since 1990 the government has structural change, the Chinese economy increas- updated its laws covering copyrights, trademarks, ingly makes use of advanced production technolo- patents, and trade secrets (or "anti-unfair competi- gies, while demand shifts toward higher-quality tion") and has adopted protection for new plant goods and services. Chinese enterprises place grow- varieties and integrated circuits. The country has ing emphasis on developing brand-name recogni- joined nearly all major international IPRs conven- tion, reputation for quality, and product innova- tions and is a member of international agreements tion. In such an environment, the provision and on classification of patents and trademarks and the enforcement of intellectual property rights (IPRs) deposit of microorganisms. help promote further economic development. With Most recently China made further revisions to substantial structural reform ongoing in Chinese conform to the requirements of the Agreement on enterprises, it is important to establish incentives Trade-Related Aspects of Intellectual Property for the development and expansion of businesses in Rights (TRIPS) in the World Trade Organization high-growth sectors and to support innovation in (WTO). In its accession agreement with WTO consumer products. Properly structured, IPRs can members, China agreed to implement intellectual help achieve these goals. property laws that are fully consistent with TRIPS. The Chinese central government recognizes the When legislative reforms are fully implemented, need for a workable IPRs system. Support is also China will have a modern structure for IPRs on a growing among innovative Chinese enterprises, par with many developed economies. which are likely to suffer the largest losses from China also has established education and train- trademark and copyright infringement in the econ- ing programs in IPRs, upgraded its administrative omy. Chinese enterprises understand that their and legal systems for enforcing these rights, and access to new foreign technologies depends partially undertaken numerous anticounterfeiting pro- on IPRs. grams. Nevertheless, problems remain in the areas In response both to changes in internal prefer- of administration and enforcement. Victims of ences and to considerable external pressure, China infringement complain about weak monetary has undertaken a dramatic reform of its intellectual and civil penalties, delays in administrative and 49 50 China and the WTO court procedures, and "local protectionism" Chinese economic development could be affected that makes enforcement difficult in regional by stronger IPRs jurisdictions. · conclusions and recommendations. The evolving system presents both opportuni- ties and challenges for the Chinese economy. The WTO Commitments opportunities arise from the improved environ- ment for technical innovation, product develop- Since the mid-1980s China has implemented a ment, and inward technology and investment number of laws and administrative regulations cov- flows. The challenges include moving resources out ering intellectual property protection (see Lacroix of infringing activities into legitimate businesses, and Konan 2002; Maskus, Dougherty, and Mertha coping with higher costs of imitating products and 1998; Potter 2001). External pressure has been an technologies, and absorbing the costs of adminis- important impetus for legal change. Many of these tering a stronger system. changes were made as a result of three agreements, Over time, stronger IPRs will shift incentives each a Memorandum of Understanding, with the away from encouraging static competition through United States. copying and imitation toward promoting dynamic This regulatory process has culminated in the competition through innovation, technology introduction of numerous changes in China's IPRs absorption, and product design. The latter incen- regime, both in anticipation of joining the WTO tives are increasingly appropriate for China, which and in becoming fully compliant after the country's plans to become a leader in technology develop- accession in late 2001. For example, a substantial ment. However, stronger intellectual property pro- second revision of the patent law was achieved in tection will place competitive pressures on lagging 2000, becoming effective in July 2001. This revision enterprises and will raise concerns about the distri- establishes full TRIPS compliance in patent regula- bution of costs and benefits among individuals, tions and clarifies and strengthens certain adminis- enterprises, and regions. trative and judicial procedures, including the use of The ultimate objectives of an IPRs system are to preliminary injunctions. In June 2003 the govern- increase competition through innovation and tech- ment issued procedures for issuing compulsory nology acquisition and to encourage innovators licenses for patented inventions, consistent with to make their products available to consumers. TRIPS Article 31, while the patent law itself was Strengthening IPRs improves such incentives but is revised in August 2003 (WTO 2003). not sufficient on its own. Rather, the system needs Similarly, in 1997 the government promulgated to be developed within a broader set of policies, new rules on the protection of new plant varieties, including further enterprise reform, development establishing sui generis, TRIPS-consistent protec- of financial and innovation systems, expansion of tion along the lines of the 1978 Union for the educational opportunities, and means for sustain- Protection of New Varieties of Plants (UPOV) ing competition in Chinese markets. treaty.1 A new set of regulations for the protection This chapter contains the following sections: of layout designs of integrated circuits came into force October 1, 2001. · a description of the progress in China's WTO Just prior to China's accession to the WTO at the commitments in the IPRs area Doha Ministerial in November 2001, some differ- · an overview of the economic rationale for intel- ences with required TRIPS standards remained.2 lectual property rights Table 4.1 provides a list of areas in which standards · a discussion of the intricate relationships are mandated, the norms (minimum standards) in between IPRs and economic development, TRIPS, the status of China's laws and regulations including a review of available evidence on that just before Doha, and China's actions in cases subject and assessment of the potential of the where a divergence exists. Most of these discrepan- China's regime to encourage growth cies arose in the trademark and copyright areas. For · an analysis of recent trends in the use of IPRs this reason, on October 27, 2001, the People's Con- in China, considering both data and informa- gress enacted a substantial revision to its trademark tion learned from a series of interviews in 1998 and copyright laws to make both consistent with and 2001, and some simple indications of how TRIPS obligations.3 Intellectual Property Rights in the WTO Accession Package: Assessing China's Reforms 51 TABLE 4.1 Substantive Requirements of the TRIPS Agreement in the WTO Pre-WTO Status of TRIPS Norm Chinese Law Actions General Obligations National Treatment Applied for persons Discrimination in copyright Remove enforcement, trademark discrimination agents, and trade secrets protection Most Favored Nation MFN with reciprocity Member of Berne Convention; exemptions for copyright TRIPS-compliant Copyrights and Neighboring Rights Term of protection Life 50 years; 50 years TRIPS-compliant corporate Computer software CR Copyright TRIPS-compliant Discussing patents Data compilations Copyright Not protected Protect with copyright Phonogram producer Right to prevent fixation, Inconsistent with TRIPS Clarify compensation and performer rights reproduction, or system; strengthen broadcasting for rights 50 years Broadcast rights Right to prevent fixation, Inconsistent with TRIPS Provide right of reproduction, or communication to broadcasting for 20 years, public or copyright Rental rights Right to prohibit rental of Not protected Provide rental rights computer programs and movies Discrimination in National Treatment Foreigners could not use Remove discrimination enforcement local copyright bureaus procedures Trademarks Well-known marks Protected without requiring No criteria for defining Protect well-known registration "well-known"; none granted marks; establish to foreigners criteria Use restrictions Use not required for registration; Law is unclear on prior use Comply with TRIPS import restraints cannot be used to invalidate use Symbols protected Rights extend to distinguishing Certain signs are ineligible Comply with TRIPS names, letters, numerals, colors Geographical Indications Basic protection Prevent misleading claims Not protected Comply with TRIPS of origin Wines and spirits Prevent use of such words as Not protected Comply with TRIPS "style" or "like" Patents Eligibility Basic exemptions Probably TRIPS-compliant Clarify compatibility with TRIPS Pharmaceutical Covered; interim marketing TRIPS-compliant products rights Living organisms Micro-organisms and biological TRIPS-compliant; plant and Considering patents production processes covered; animal varieties excluded "higher-order" life optional Term of protection 20 years from filing TRIPS-compliant 52 China and the WTO TABLE 4.1 (Continued) Pre-WTO Status of TRIPS Norm Chinese Law Actions Patents Rights Exclude others from production, TRIPS-compliant use, or distribution Compulsory licenses Wide scope for use with TRIPS-compliant compensation and limiting conditions Burden of proof in Falls on defendant TRIPS-compliant process patent infringement Industrial Designs Term of protection 10 years from filing TRIPS-compliant Textile designs Covered Protected by copyright Considering design patents Plant Varieties Basic protection Plant breeders' rights to Plant breeders' rights with prevent commercial farmer's privilege and production or marketing research exemption of propagating material (UPOV 1978) Stronger protection Patents optional No patents Integrated Circuits Term of protection 10 years from filing TRIPS-compliant Rights Prevent distribution of IC's TRIPS-compliant or IC-using products Exceptions Non-voluntary license TRIPS-compliant Trade Secrets Protection from Defines boundaries of unfair TRIPS-compliant unfair disclosure practices Test data for Protection from disclosure Unfair use not prohibited Protection for six pharmaceuticals for unspecified period and years from date and agricultural unfair use of undisclosed of marketing chemicals data approval Control of Anti- Competitive Practices Compulsory licenses Wide latitude for use subject to Weak consultation provisions Comply with TRIPS conditions and consultations Exhaustion No standard Depends on form of IPR Enforcement Sanctions Civil and criminal sanctions In existence but weak Enhance and border measures enforcement action enforcement Provisional measures Preliminary injunctions and Not fully available Comply with TRIPS seizures Damages Adequate to compensate Generally low or no Comply with TRIPS victim of infringement compensation Administrative actions Enforcement may be through Available but costly and tends Enhance administrative actions to result in small fines enforcement Judicial review Must be available Not widely available Enhance review procedures The trademark law establishes the right of indi- association (e.g., for producing fruits and vegeta- vidual Chinese persons to apply to register trade- bles of a given quality standard) may use jointly. It marks. It clarifies the definition of collective marks, also establishes joint ownership and protects collec- which are marks that several firms in a collective tive marks and certificate marks for the first time. It Intellectual Property Rights in the WTO Accession Package: Assessing China's Reforms 53 further broadens the range of symbols that may be The Nature of Intellectual used as a distinctive mark, extends protection of Property Rights well-known trademarks to their unauthorized use To analyze how IPRs influence economic develop- on different products, and sets out criteria for ment, it is important to understand their economic ascertaining well-known marks. These regulations underpinnings. The need for IPRs arises from the on well-known marks were further updated and social objective of promoting the creation of new adopted by the National Industrial and Commer- types of information adding to the economy's cial Administration in June 2003 (WTO 2003). The knowledge base. These types of information trade mark law also sets out protection for geo- include new products and technologies; new liter- graphical indications in accordance with TRIPS. ary, musical, and artistic expressions; and indica- The copyright law establishes a communication tors of product quality. An intellectual property right over the Internet, sets out fair use limitations right defines the extent to which its owners may for electronic content consistent with two new legally prevent others from taking actions that treaties reached under the aegis of the World Intel- infringe or damage the property. It also may be lectual Property Organization (WIPO), clarifies defined as the legal ability to set terms on which it broadcast and rental rights, and recognizes that may be used, subject to public-interest limitations databases are copyrightable. Further changes were on the scope of that ability. For example, patents made in August 2002 with respect to rental rights, provide exclusive use of a technology, but they are performance rights, rights of communication and granted only for a fixed period of time and in broadcast, and database protection. return for disclosing sufficient technical informa- Both the trademark and copyright laws clarify tion to allow competitors to understand it and try the amounts of compensation available to plaintiffs to improve on it. and the methods for their calculation. In copy- Public intervention is needed because the out- rights, foreign firms are now permitted to plead come of some intellectual effort may be potentially their cases with local copyright bureaus in addition valuable but also easily copied and used by others, to the National Copyright Administration, remov- leaving little incentive to incur the original invest- ing a prior form of discrimination. A number of ment costs or to improve it. Without public sup- clarifications regarding enforcement, administra- port for innovation, the economy suffers from tion, and penalties were made in 2002 and 2003 insufficient incentives to develop new products, (WTO 2003). technologies, and cultural works, making citizens This recent and ongoing activity should make worse off in the long run. An additional problem is China's IPRs legal regime fully consistent with that the social value of information is often greater TRIPS in the near term. It is evident that the prom- than its private value because there are external ise of WTO membership significantly accelerated benefits from new inventions. Examples include China's legal reforms in IPRs, because consistency spillover cost reductions from new technologies to with at least minimum TRIPS standards is a neces- input users and network efficiencies from software sary condition for accession (Maskus 2000a). systems. Moreover, China's commercial interests in innova- At the same time, IPRs generate costs. Legal tion, development of consumer products, and excludability imposes a static cost on users, reflected access to high-level foreign technologies provided by the excess of price over marginal production significant domestic motivation for such reforms. cost. In the case of intellectual creations, this distor- Nevertheless, because China remains a country tion can be significant because the marginal cost of with relatively low per-capita income and techno- supplying additional blueprints, digital video discs logical development has not spread widely through (DVDs), and computer programs is small. Addi- the economy, it is likely that significant problems tional costs are incurred because IPRs encourage with enforcement will remain for some time. China duplication of investment in research and develop- may become one focus of dispute resolution cases ment (R&D) through patent races, and generate at the WTO as intellectual property interests from wasteful efforts to assert and defend ownership developed countries demand stronger enforcement rights and to extend those rights beyond the scope and protection. granted. Enforcement costs may be substantial. 54 China and the WTO Thus, there are complex tradeoffs facing the A related form of industrial property is plant design of IPRs. China needs to strike an appropriate breeders' rights (PBRs), which permit developers to balance among the needs of creators, developers, control the marketing and use of new plant vari- and users, in accordance with minimum standards eties. These rights operate much like patents, pro- required by TRIPS. viding for fixed terms, requirements of distinctive- One alternative to IPRs is direct government ness in new plant strains, and disclosure rules. They support for invention, including public monopolies are intended to encourage development of new in technology development, research institutes in seed strains for agricultural use. agriculture and industry, and government sub- Rights to market goods and services under sidies to university research. Many such programs exclusive names and symbols are protected by are important complements to IPRs in national trademarks and service marks. Registration may be innovation systems. However, the performance renewed indefinitely, subject to use requirements. of China's public organizations and state-owned An important related device is the geographical enterprises in commercializing products and serv- indication, which permits the use of a particular ices has been poor. This failure arose largely from place name where a good was produced to ensure inadequate incentives within public research insti- that the product embodies quality characteristics tutions to focus on new and marketable inventions. of that region. Trademarks provide incentives for firms to invest in brand-name reputation and Intellectual Property Rights and product quality and for their licensees to produce Economic Development and sell high-quality goods. If marks were not pro- tected, rival firms would ruin their value by selling Relationships between IPRs and economic develop- cheaper items under those marks. Thus, the social ment are extremely complex and available evidence benefits from trademarks include greater product is difficult to interpret (see Evenson and Westphal variety and lower consumer search costs due to the 1997). However there is a growing consensus that absence of confusion. stronger IPRs can improve development prospects Firms develop technological know-how that is if they are structured properly. important for production but may not be patentable or may have greater economic value if it remains Types of Intellectual Property Rights undisclosed. Such trade secrets are protected by It is useful to provide a brief description of the legal rules against unfair misappropriation. There is main forms of IPRs. Rights to exploit inventions no exclusive right to the process if it is discovered with commercial uses are granted through patents, by fair means, such as reverse engineering. Trade utility models or petty patents; these terms refer to secrets protection is important because it supports short-term exclusive rights for small and incremen- the introduction of sub-patentable technological tal innovations, and industrial designs. Patents pro- processes into commerce and also promotes compe- vide the right to prevent for a fixed time (20 years tition through reverse engineering. under TRIPS) the unauthorized making, selling, or Artistic and literary creations are protected by using of the product or process described. Utility copyrights, which are exclusive privileges to copy models provide exclusive rights for a shorter period and distribute particular creative expressions for a to inventions that embody only a small inventive fixed term. Related IPRs include neighboring rights step. The scope of patent coverage is limited to uses of performers and broadcasters and moral rights of of the novelty claimed by the inventor and recog- original creators to prevent future alterations of nized by patent examiners. The technology must their works. Copyrights are subject to limitations meet technical criteria for novelty (or non- for social purposes, with the most prominent obviousness), inventiveness, and industrial utility exception being the fair-use doctrine, under which and must survive procedures challenging validity. others may use copies for scientific and educational Patent applications are published for inspection by purposes. Countries vary widely in the scope of interested persons. The essential tradeoff in patents their fair-use exceptions to copyrights, particularly is to create a protected market position in return with regard to reverse engineering of computer for disclosure of technical knowledge. programs. Intellectual Property Rights in the WTO Accession Package: Assessing China's Reforms 55 Some new technologies do not fit easily within stimulates the entry of small and medium-sized this framework. Computer software contains ele- enterprises into specific markets. Second, it encour- ments of both literary expression and industrial ages more successful enterprises to grow and take utility, raising questions about whether it should be advantage of scale economies through interregional covered by copyrights or patents. The TRIPS stan- production and marketing. Some may even become dard is for copyright protection; the United States, significant exporters as they improve quality. Japan, and other countries also provide patent pro- Similar comments apply to copyrights. Sectors tection. Similar questions arise regarding semicon- that are dependent on copyrights, such as publish- ductor chip designs, which achieve a special form of ing, recorded entertainment, and software, will protection under TRIPS. Electronic transmissions experience limited entry by local firms in their of Internet materials, databases, and broadcasts absence. Creation of new films, music, and software also raise concerns about the adequacy of copyright is expensive and little worth the investment by local protection to encourage their development. entrepreneurs if their products will be copied. IPRs are enforced both to deter and punish Accordingly, society's long-run cultural and eco- infringement and to discipline rights holders that nomic development is impaired. abuse their market power. Enforcement against Innovation goes beyond developing new prod- basic infringement involves seizing unauthorized ucts to establishing marketing and distribution net- copies, destroying the associated facilities if neces- works. It is difficult to do this in an environment of sary, and imposing fines and criminal sanctions. weak IPRs because rights holders cannot readily However, as claims about infringement or abuse, prevent infringement. IPRs improve the certainty particularly in the patent realm, become more com- of contracts, and permitting better monitoring plicated, courts must decide on their legality, which and enforcement of rights at all levels of the supply requires considerable legal and scientific expertise. network. In turn, both innovative firms and their distributors are more willing to invest in marketing and brand-name reputation. How IPRs Stimulate Economic Development As firms build reputations through trademarks, IPRs protection can spur economic development incentives grow to deter false use of those marks. and growth in several ways. A weak regime can stifle Fake products sold under a misappropriated trade- both invention and innovation even at low levels of mark can ruin reputations, particularly for new economic development. Most inventions are spe- firms, and overcoming such damage can be costly. cific to local market circumstances and can benefit Thus, effective trademark enforcement should from patent or utility model protection. Innovation increase the average quality of products over time usually involves minor adaptations of existing tech- and permit consumers to be less wary of counter- nologies, management systems, and quality control feit goods. This is particularly important in cases of mechanisms, which can stimulate growth. Such beverages, foodstuffs, and medicines. investments tend to have high economic and social Finally, IPRs can help disseminate knowledge. returns by raising productivity toward international Patent claims are published, and competitors may levels. Evidence from Brazil and the Philippines use the disclosed technical knowledge to develop suggests that effective systems of utility models can further inventions. This cumulative process of in- promote innovation (see Maskus and McDaniel vention, which depends on the narrowness of 1999). Another study demonstrated econometri- patent claims, can be an important source of tech- cally that Japan's system of utility models con- nical change (see Maskus and Penubarti 1995; tributed positively and significantly to its postwar Smith 1999). Moreover, patents provide a legal rise in productivity (see Scotchmer et al. 1991). basis for trading and licensing technologies. Trade- Trademarks provide incentives for the entry of marks and trade secrets facilitate information new firms and the development of new products, exchange through ensuring that licensees do not even in poor nations. Firms find it easier to innovate abandon their contracts. cumulatively as they grow larger and their trade- Considerable evidence suggests that interna- marks are better recognized. This process has two tional flows of technology depend on the strength positive effects on industrial development. First, it of IPRs, among many other factors. For example, 56 China and the WTO international trade in manufactures is positively and rapid growth, making it easier to shift workers affected by the strength of patent regimes in large into legitimate activities. China has mixed prospects developing countries (see Coe, Helpman, and on this score, in part because of restrictions on Whoffmaister 1997). This trade often embodies internal labor mobility. It is conceivable that copy- technical knowledge that may be learned in recipient right and trademark enforcement will contribute countries and adapted to local technological capa- significantly to looming unemployment problems bilities (see Maskus 1998a; Lee and Mansfield 1996). associated with economic reform. Foreign direct investment, joint ventures, and Because IPRs raise the costs of copying and imi- technology-licensing contracts also transfer produc- tating products and technologies, acquiring tech- tion knowledge. It is clear that the strength of IPRs nological knowledge through simple imitation influences choices by multinational firms on where could become more expensive in China. Consider- to invest and whether to transfer advanced tech- able anecdotal evidence suggests that firms operat- nologies. Studies of U.S.-sourced foreign direct ing in China lose technologies to potential rivals investment (FDI) find evidence that firms limit their through such avenues as defection of technical per- investments in countries with weak patents (see sonnel, misappropriation by input suppliers, and Mansfield 1995; Contractor 1980). Survey evidence copying of blueprints. Without effective trade indicates that the level of technology transferred secret protection, these activities are common and depends on the ability to maintain control over help establish competition. However, some impor- the technology through defense of intellectual prop- tant practical effects are that foreign firms tend to erty (see Yang and Maskus 2001). Licensing also transfer older technologies, engage in less techni- tends to rise with stronger IPRs because of reduced cal training, hide key aspects of know-how from contracting costs and greater legal certainty.4 subcontractors and suppliers, and not establish first-rate R&D facilities. Accordingly, stronger IPRs entail a balancing How IPRs Limit Economic Development act. Stronger IPRs make uncompensated imitation Tightened IPRs also impose economic costs. Poor more difficult but improve the quality of technol- countries may experience net losses in the short- ogy flows. Countries wishing to become significant run because the dynamic gains tend to take a long technology developers should favor IPRs for that time to emerge. reason. The costs of administering and enforcing a Stronger IPRs create market power, from which modern IPRs system are high. For China they easily firms may be able to raise prices to monopolistic will amount to annual sums in excess of $10 mil- levels. This concern is particularly relevant in devel- lion (see Watal 1999; Lanjouw 1998). These costs oping countries for two reasons. First, applications include training of examiners, judges, lawyers, and for protection come overwhelmingly from foreign enforcement officers, along with the costs of run- firms, meaning that the associated profits are trans- ning various offices. Many of these costs may be ferred abroad. Second, market competition may be covered by administrative fees charged to apply for weak, supporting monopolization. As will be dis- and register patents and trademarks, while others cussed in the next section, China is in an intermedi- may be reduced by using international registration ate position in this regard. Patent applications by agencies such as the Patent Cooperation Treaty domestic firms are rising rapidly but lag behind (PCT), as does China. foreign applications. Broader economic reforms The most visible aspect of IPRs infringement in have improved competitive processes, but the econ- China is unauthorized copying of recorded enter- omy remains far from a situation of free entry and tainment and software and selling of products vibrant competition in technology and product bearing counterfeit trademarks. Undoubtedly, sig- markets. nificant amounts of labor are employed in copying Chinese authorities may be especially concerned and retailing illegitimate products in China, and an about the implications of patent rights for the important short-run cost of stronger IPRs will be prices of medicines. There is evidence that patents labor displacement. The adjustment costs tend to can support markedly higher prices for pro- be smaller in economies with flexible labor markets tected drugs than for copied and generic drugs.5 Intellectual Property Rights in the WTO Accession Package: Assessing China's Reforms 57 However, the extent of these price increases A Scorecard for China's IPRs Regime depends on the competitive aspects of markets. The With this background it is useful to assess infor- more competitive the local drugs market is before mally the legal standards China has adopted in patents are awarded, the larger is the share of drug terms of how these standards may affect prospects production that consists of copies of patentable for technical change and the provision of certain drugs; the more inelastic the demand for medicines public goods.6 For this purpose, China's regula- is, the higher will be the price increases caused by tions in several critical areas are compared against patents. These conditions suggest that countries benchmark standards consistent with TRIPS with extensive drug imitation and high demand requirements for middle-income countries as set could experience substantial price increases for out by the World Bank (2001). China has character- protected drugs. In this regard, China's policy istics of the dynamic middle-income countries, of public procurement at negotiated prices is including a growing base of human capital and appropriate for ensuring the continued provision sophisticated capabilities in science and technol- of public health services. ogy, suggesting that this comparison may be appro- Another area of concern is computer software. It priate. However, it also suffers from substantial is often claimed that software would be much more rural and urban poverty, and many of its enter- expensive with enforced copyrights because the prises use technologies that lag significantly behind current prices of legitimate copies in developing those in the modern sectors. Accordingly, it is diffi- nations are very high compared to prices of unau- cult to provide a definitive analysis of the nation's thorized copies. However, in countries with high laws in this context, and the following analysis piracy rates software producers or their distributors is offered largely for purposes of illustration and tend to sell at low volumes and high markups, discussion. reflecting small markets with inelastic demand. In An overview is provided in table 4.2. In the China such markets are largely limited to foreign- copyrights area, China provides TRIPS-standard owned enterprises and government agencies. As terms of protection for creative goods, such as a markets broaden under copyright enforcement, once-renewable 25-year term for software copy- foreign and domestic firms should supply more rights (50-year maximum period). The essential legitimate copies at lower prices, suggesting that question is whether the particular conditions of ultimate price increases could be modest (see legal protection are appropriate for China's devel- Potter 2001). opment prospects. For example, it is recommended Thus, there are legitimate concerns about that middle-income countries provide fairly wide market power supported by IPRs. However, com- exceptions for fair use of copyrighted materials in petitive markets and appropriate regulation can education and scientific research. China's Copy- mitigate these impacts without unduly reducing right Law embraces this concept, permitting free innovative incentives. IPRs need to be introduced use of copyrighted material in journals, periodicals, into markets in which other competitive processes, and broadcast media for purposes of disseminating such as firm entry, labor flexibility, distribution news. It also allows the uncompensated making of systems, and international trade, are operating copies for classroom use and scientific research and effectively. allows a free right to translate works from Han into Competition is important because IPRs may be minority languages. In software, users are permitted abused, as shown by litigation problems in the a limited right of decompilation for purposes of United States, the European Union, and elsewhere. developing new programs, which should help the Such abuses include bad-faith lawsuits, hidden industry remain fairly open to incremental innova- ownership of intellectual property, restrictive tion and competition. However, the government patent pooling agreements, refusals to license tech- is considering extending patents to computer nologies, tie-in sales in related markets, and insis- programs, a standard that would exceed TRIPS tence on exclusive rights to competing technolo- requirements and is found only in the United States, gies. Thus, China must develop mechanisms for Japan, and Australia (see Caves 2000). For a soft- ensuring competition maintenance in markets ware sector that remains young and subject to affected by IPRs. 58 China and the WTO TABLE 4.2 An Assessment of China's IPRs System for Development Purposes China's Post-WTO Area of IPRs Middle-Income Standards Standards Commentary Copyrights Fair use Liberal exceptions for Same Important for research and exceptions education and research technology access Computer Copyrights with decompilation Permit limited decompilation Patents may be overly software and fair use and fair use; considering protective patents Market Improve collection societies Weak institutions Improvement would be institutions and rights contracts beneficial WIPO treaties Adopt minimum standards Considering ratification Could improve internet content and access Data Copyrights with creativity Same Avoid EU-type protection compilations requirement Trademarks Well-known Narrow definition of sectoral Unclear Clarify scope and recognize marks coverage well-known marks Domain names Protect against fraudulent Same Important for promoting registration internet use Confidential Information Test data Short period of protection Six years of protection Stronger than U.S. from disclosure standard Disclosure Limited restraints on what Unclear limitation on Ambiguity may be restraints employees may reveal to "promise of gain" detrimental new employers Patents Exemptions Exemptions for discoveries, Same Appropriate to retain from eligibility algorithms, medical exemptions treatment methods, plants and animals Novelty Consider oral prior art Oral prior art in China Scope could be widened Inventiveness High step Notable progress Unclear Scope of claims Narrow claims and narrow Single claim; scope Depends on examiners doctrine of equivalents unspecified Experimental use Permit experimental use Not permitted May be overly strong Exhaustion International exhaustion National exhaustion Strong protection Government use For clear public interest Same Appropriate Compulsory Permitted under TRIPS Same Appropriate licenses conditions Utility models Short duration, low 10-year duration, same Appropriate and designs inventiveness Plant Varieties Farmers' privilege Recognize farmers' privilege Recognized Appropriate Breeders' Use for breeding and Permitted Appropriate exemption scientific research permitted considerable cross-fertilization through learning institutions for realizing economic returns to and reverse engineering, such a choice seems ques- creative activity, including collection societies for tionable for the medium term. Neither should licensing recorded music and gathering royalties. China extend patents to methods of instruction or Indeed, only under the Amended Copyright Law of doing business. of 2001 were copyright owners given the right to Other important issues exist in the copyright authorize collection associations to administer their area. China has relatively weak complementary rights. Further, because contract enforcement can be Intellectual Property Rights in the WTO Accession Package: Assessing China's Reforms 59 weak and uncertain, the ability of enterprises to allo- may be costly in terms of sorting out the scope of cate rights in creative works is limited. Such institu- unfair competition in this area. tions are important for providing a full framework China does not patent higher-order life forms or within which the development of artistic and literary biological research tools. In its 2001 patent law, work can grow beyond intermediate stages (see also China retains appropriate exemptions from cover- Dahlman and Aubert 2001). To date China protects age for discoveries of nature, mental methods of databases solely with copyright protection, as man- arriving at results (such as computer algorithms dated by TRIPS, and clarifies that such protection and mathematical formulas), diagnostic and surgi- cannot interfere with the independent rights of cal treatments, and plant and animal varieties. those who develop components of compilations. These limitations are widely advocated for develop- This minimum standard is appropriate for a country ing countries with emerging biotechnology sectors with a strong interest in access to information data- in order to avoid locking up critical technologies bases, and China should be wary of moving toward that support additional research and learning. the much stronger standards of the European China's patent law does not permit experimental Union. Finally, ratification of the WIPO Copyright use of patented materials, however, which may be Treaty and the Treaty on Performances and Phono- overly strong in the context of its development grams could be beneficial in sorting out copyright strategies. The country's standards covering gov- protection for Internet transmissions, so long as ernment use, compulsory licenses, and utility mod- appropriate fair-use limitations are provided. els and designs are typical of middle-income The Amended Trademark Law of 2001 clarifies economies. the definition of well-known trademarks by setting Finally, China's plant variety law also seems out five criteria for achieving protection without appropriate for its needs. The patent law excludes registration. It remains to be seen whether these such inventions from coverage, leaving them to criteria will limit the unauthorized use of such variety protection. China's regulation permits the marks by others, and the regulation is unclear farmers' privilege to use propagating materials about how widely the restrictions on use will apply for re-planting and also permits experimental use across sectors of business. This law also recognizes for science and for rival breeders to develop new that well-known marks should be applicable to varieties. domain names on the Internet, a provision that should promote wider content available to Chinese The Issue of Enforcement Internet users. Protection of confidential test data from disclo- The largest remaining obstacle to effective use of sure provides applicants for patents in pharmaceu- IPRs in China is weakness in enforcement proce- tical and chemical products a period of exclusivity dures. Recent legal changes increased the scope of in the use of results from clinical trial. It seems enforcement considerably. As noted in table 4.1, advisable for middle-income economies with criminal sanctions are now available for cases of domestic research capabilities in medicines and willful infringement, while the maximum permis- biotechnology to provide such exclusivity for some sible monetary sanctions were increased. Prelimi- period, but TRIPS is silent on the length of any nary injunctions and orders for seizure of sus- such requirement. Surprisingly, China has opted to pected infringing goods may be issued, which are protect it for six years from application date, in important components of timely relief for IPRs comparison to the U.S standard of five years. owners. Courts may order compensatory damages, China's law may be overly protective from the though this standard is weaker than what might be standpoint of the encouragement of domestic required to deter infringement a priori. Finally, competition. Next, China's law on unfair competi- enhanced access to judicial review is provided, tion has no explicit language on the legality of which is consistent with prior efforts by the restraints on the ability of employees to reveal tech- Chinese government to strengthen its enforcement nical secrets to rivals that may hire them, other than mechanisms. to declare void unfair "promises of gain." It is Looking forward, perhaps the most significant unclear what this statute covers, and the ambiguity issue for China with respect to IPRs and the WTO 60 China and the WTO is that country's commitment to enforce against represented a mix of state-owned enterprises infringing activities and prevent exports of pirated (SOEs), private Chinese enterprises, joint ventures and counterfeited goods. Because China remains a with international firms, and majority-owned sub- country with relatively low per-capita income and sidiaries of multinational enterprises. Most firms in technological development has not spread widely the last category are in high-technology sectors and through the economy, it is likely that significant have significant R&D programs in their home problems with enforcement will remain for some countries, although some undertake R&D in China time. Because the scale of activity in China is so as well. Thus, the sample is not representative of the large, weak enforcement against domestic infringe- bulk of Chinese industry at this time but is more ment would attract the concerns of patent, trade- focused on product and technology development. mark, and copyright holders. Thus, China may Overwhelmingly, enterprise managers believe become one focus of dispute resolution cases at the that the legal structure for IPRs in China has WTO as intellectual property interests from devel- improved markedly and is adequate. However, the oped countries demand stronger enforcement and majority think the enforcement environment protection. remains quite weak, while the rest find it to be weak Whether improvements in enforcement are but improving. Interestingly, Chinese enterprises liable to be in China's favor as a development issue tend to view the system as improving more rapidly depends on how economic agents respond to the than do foreign-owned enterprises and joint ven- changed incentives. As noted above, it is likely that tures. Many high-technology Chinese enterprises this factor will generate higher equilibrium inflows applaud the new legal climate, which allows them of technology transfer. However, the essential ques- scope for defending their intellectual property. tion is whether it will expand incentives for local Unquestionably, inadequate enforcement is the business development. The following section pro- main problem facing firms wishing to exploit intel- vides evidence that this expansion is likely to occur. lectual property in China. There can be long delays in enforcement actions and court rulings. Prior to the legislative changes in 2001, monetary penalties The IPRs Situation in China were small even in cases of significant infringe- Two sources of information about the current situ- ment, and there was little scope for criminal prose- ation regarding the use and adequacy of IPRs pro- cution of willful and ongoing violations. The vide more concrete perspective on the potential new laws increase maximum fines and clarify the impacts of IPRs reform. First, the results of inter- nature of criminal activity, but some interviewees views of public officials, university scholars, and thought these changes would be insufficient to enterprise managers conducted in 1998 and 2001 deter infringement. Enforcement actions may be paint a consistent picture that has not changed sig- taken arbitrarily and may lack transparency. The nificantly over the three-year period. In fact, a central government and certain regional and number of interviewees described the situation municipal governments are taking steps to reduce with respect to enforcement of IPRs to have deteri- these problems, and several enterprise managers in orated in that period. Second, analysis of recent 2001 positively commented on this change. patent and trademark statistics in China suggest There are several structural sources of weak that the use of formal IPRs is growing rapidly enforcement: but significant regional disparities exist. Overall the analysis suggests that the IPRs regime for inven- · Trademark infringement and illegal copying tion and innovation is improving in China but remain profitable and face little opposition, significant problems remain.7 especially in rural and inland regions. · Enterprises engaged in infringement often are important employers and sources of revenue for Discussion of Interview Findings local governments. Management officials and intellectual property · Low salaries for public officials may reduce managers of enterprises from several IPR-sensitive their effectiveness as enforcement agents, while industries were interviewed. The enterprises administrative programs may be underfunded. Intellectual Property Rights in the WTO Accession Package: Assessing China's Reforms 61 · Legal and technical expertise for administrative Defection of technical and managerial employ- and judicial operations is limited despite the ees remains a basic problem for both foreign- existence of special training programs in IPRs. owned and Chinese enterprises. In economic terms a balance is needed between promoting mobility of Among these problems, "regional protection- skilled labor, which raises diffusion and competi- ism" in IPRs is regarded as the most difficult to tion, and discouraging uncompensated losses of confront by enterprises suffering infringement. technical knowledge, which can reduce competi- There is little coordination among regional bureaus tion over time. Both foreign and domestic enter- of the Administration for Industry and Commerce prises attempt to manage the problem with tempo- (AIC). Moreover, the regional AICs have weak rary anti-disclosure clauses, but such contracts administrative powers and actions of municipal have been difficult to enforce in China and recent governments may supercede them. Municipal gov- changes in the law have not clarified this issue. ernment officials may well have priorities that take Most respondents agreed that the environment precedence over IPRs enforcement. for selling copyrighted materials is improving in Managers of both Chinese and foreign-affiliated China, although pirating of software, games, DVDs, enterprises expressed the view that weak enforce- and music remains common. While large foreign ment of IPRs results in widespread copyright and firms claim significant harm from such copying, it trademark infringement. A major problem is that is likely that relatively larger losses are suffered by trademark violations often target innovative Chinese entertainment and publishing interests. Chinese enterprises and thereby deter local business For its part, the Chinese software industry is grow- development. Examples were given of problems ing rapidly, largely because of a substantial base of facing Chinese-brand producers of such consumer skilled software engineers and managers. However, goods as medicines, soft drinks, processed foods, according to many interviewees, such firms concen- tobacco products, and clothing. Enterprises selling trate on developing small-scale programs that electronics products seemed particularly vulnera- attract less copying, such as business applications ble. Once brand recognition is achieved, domestic or limited-run games. This problem could delay the enterprises find their trademarks applied to unau- establishment of Chinese-developed software stan- thorized products of lower quality, damaging the dards and networking software. original enterprise's reputation. In some cases, Significant differences exist between Chinese and Chinese enterprises either had to give up on their foreign-owned businesses in their ability to deal trademarks and become licensees of better-known with trademark and other IPRs-related violations. enterprises or undertake extensive private and pub- Foreign companies have more resources to combat lic enforcement actions. It is impossible to know the infringement than domestic enterprises. Insofar as extent to which this problem hampers industrial an enforcement action is a significant expense for a development, but the impact could be significant. small or medium-sized Chinese operation, enforce- Weak enforcement also impedes efficient use of ment difficulties are biased against Chinese business patents and trade secrets. Patent infringement development. Furthermore, foreign companies, seems to be most common in utility models, which particularly Western ones, are more inclined to seek are easy to copy but are overwhelmingly owned by legal solutions to IPRs problems. Chinese enterprises. Several foreign enterprises also Foreign companies may undertake more defen- claimed to have lost patented technologies through sive actions in the presence of weak IPRs. Managers unfair means, such as the selling of design specifi- of most foreign enterprises indicated a reluctance cations and technical manuals by former employ- to locate R&D facilities in China, although this ees. According to one industry association, such is changing rapidly as the legal environment cases are becoming more common and increasingly improves. Nearly all indicated that in the past they targeted on sophisticated technologies. Interview- transferred technologies that are at least five years ers claimed that many foreign companies are behind global standards in the expectation that considering more carefully whether they wish to those technologies would be lost to local competi- transfer advanced technologies into the Chinese tion, or they brought in technologies that would be economy. obsolete quickly. Foreign enterprise managers are 62 China and the WTO often reluctant to license technologies, preferring Patent and Trademark Activity in China joint ventures and majority-owned subsidiaries in Despite these problems, data on patent and trade- which they can exercise greater control of propri- mark use indicate that both foreign and domestic etary secrets. Enterprises are unlikely to integrate enterprises are applying for more protection. fully their Chinese operations, splitting various Table 4.3 presents figures on applications to the production processes among facilities in order not State Intellectual Property Organization for all to reveal fully the underlying know-how. three types of patents from 1994­2000. Domestic Other defensive measures are used by both enterprises more than doubled their applications Chinese and foreign companies. One is to sell only for invention patents, while foreign applications to established customers that need assured quality, rose by 235 percent. From 1996 through 1999 for- such as hospitals, large enterprises, and public eign applications considerably exceeded domestic agencies. This acts as a barrier to the entry of small applications. A significant rise in Chinese domestic firms needing the associated products or inputs. A applications in 2000 virtually equalized the number second is to establish strict vertical supply and for that year, however. Chinese enterprises now distribution chains to permit monitoring of qual- apply for nearly as many invention patents, with ity. A third is to employ technical safeguards, such their higher inventive content, as do foreign as software locks and encrypted source codes that enterprises.8 must be decoded to operate software upgrades. In contrast, applications for utility models and It is impossible to know how these distortions design patents overwhelmingly are filed by Chinese associated with weak IPRs contribute to economic organizations. In both categories domestic applica- inefficiency in China, although the effects presum- tions rose far faster than foreign applications. Thus, ably are significant.If so,stronger IPRs over time will these rewards aimed at encouraging small-scale generate important static and dynamic efficiencies. invention seem to be having their desired effect on Public officials often raise concerns about the domestic innovation. potential impacts of stronger IPRs on prices and Table 4.4 shows data for patent grants and competition. Some officials also recognize that the ratio of cumulative grants to cumulative stronger IPRs need to be accompanied by other applications over the same period. Grants of inven- policy measures to build technological capacities tion patents to both domestic and foreign appli- and maintain competition. However, this recogni- cants increased rapidly, with the former nearly tion only recently has been translated into such catching up to the latter by 2000. The aggregate policy initiatives as the Standards Office within the grants ratios for invention patents are surprisingly State Industry Commission, which will establish low, perhaps reflecting long examination delays. uniform national standards for information net- Grant rates are much higher in utility models and works, including copyright provisions. TABLE 4.3 Patent Applications by Type and Nationality, 1994­2000 Invention Patents Utility Models Design Patents Total Patents Domestic Foreign Domestic Foreign Domestic Foreign Domestic Foreign 1994 11,191 7,876 45,188 323 11,428 1,729 67,807 9,928 1995 10,018 11,618 43,429 312 15,433 2,235 68,880 14,165 1996 11,471 17,046 49,341 263 21,395 3,219 82,207 20,528 1997 12,713 20,953 49,902 227 27,456 2,957 90,071 24,137 1998 13,726 22,234 51,220 177 31,287 3,345 96,233 25,756 1999 15,596 21,098 57,214 278 37,148 2,905 109,958 24,281 2000 25,346 26,401 68,461 354 46,532 3,588 140,339 30,343 Growth 126.50% 235.20% 51.50% 9.60% 307.20% 107.50% 107.00% 205.60% Total 100,061 127,226 364,755 1,934 190,679 19,978 655,495 149,138 Intellectual Property Rights in the WTO Accession Package: Assessing China's Reforms 63 TABLE 4.4 Patent Grants by Type and Nationality, 1994­2000 Invention Patents Utility Models Design Patents Total Patents Domestic Foreign Domestic Foreign Domestic Foreign Domestic Foreign 1994 1,659 2,224 32,611 208 5,507 1,088 39,877 3,520 1995 1,530 1,863 30,195 276 9,523 1,677 41,248 3,816 1996 1,383 1,593 26,961 210 11,381 2,252 39,725 4,055 1997 1,532 1,962 27,185 153 17,672 2,488 46,389 4,603 1998 1,655 3,078 33,717 185 26,006 3,248 61,378 6,511 1999 3,097 4,540 56,094 274 32,910 3,241 92,101 8,055 2000 6,177 6,506 54,407 336 34,652 3,267 95,236 10,109 Growth 272.30% 192.50% 66.80% 61.50% 529.20% 200.20% 138.80% 187.20% Total 17,033 21,766 261,170 1,642 137,651 17,261 415,954 40,669 Grants Ratio 17.00% 17.10% 71.60% 84.90% 72.20% 86.40% 63.50% 27.30% Source: State Intellectual Property Office 2000. TABLE 4.5 Bilateral Invention Patent Applications, 1994­98 1994 1996 1998 by China in China by China in China by China in China Country in Foreign by Foreign in Foreign by Foreign in Foreign by Foreign U.S. 190 8,105 246 14,892 436 25,634 Japan 137 3,742 145 7,212 373 11,301 Germany 172 2,094 210 3,631 624 6,599 U.K. 188 2,028 224 2,656 627 4,216 Australia 87 692 101 746 295 1,067 Korea, Rep. of 84 569 97 1,645 309 2,076 Brazil 85 20 100 39 290 104 India 21 8 54 6 15 21 design patents, which are easier to examine and in China during 1985­96 and 2000. Residents of carry shorter protection periods. Guangdong applied for over 21,000 patents in Tables 4.5 and 4.6 indicate trends in bilateral 2000, while people in Hebei applied for less than invention patenting activity between China and key 4,000. Better measures of inventive capacity are trading partners.9 Chinese patent applications given in the final two columns as applications per abroad rose sharply between 1996 and 1998 in all million people and applications per million yuan of countries listed except India, suggesting an increas- regional GDP. In these rankings Beijing is at the top ing international orientation of Chinese innovation. of the list, with far more applications per capita and However, China remains well behind most devel- per unit of output than any other province. This oped countries in terms of bilateral applications reflects both Beijing's status as a technology devel- flows, with the United States and Japan together oper and the fact that many patent registrations applying for some 37,000 patents in 1998 in that come through legal offices in the capital. Shanghai country. Despite the increase in Chinese applica- has the second highest applications per person but tions abroad, only Japan actually granted rising ranks sixth in applications per yuan of GDP. Fujian numbers of invention patents to Chinese inventors. and Hebei rank low in both categories. Table 4.7 provides a breakdown of total domestic The middle column ranks these regions in patent applications for the top 11 patenting regions terms of average income per capita. There are large 64 China and the WTO TABLE 4.6 Bilateral Invention Patent Grants, 1994­98 1994 1996 1998 to China by China to China by China to China by China Country by Foreign to Foreign by Foreign to Foreign by Foreign to Foreign U.S. 48 701 46 449 72 785 Japan 5 579 29 445 7 927 Germany 16 213 15 148 7 318 U.K. 21 94 29 66 12 133 Australia 4 24 5 20 14 36 Korea, Rep. of 0 48 0 49 2 149 Brazil 2 4 0 1 5 2 India 5 2 1 0 0 3 Note: Data include applications under Patent Cooperation Treaty. Source: WIPO, various years. TABLE 4.7 Patenting Indicators for Top Patenting Regions, 1985­96 and 2000 2000 Applications 2000 Applications Applications 2000 GDP per per Million of per Million Yuan of Region 1985­96 2000 Capita, Yuan Population GDP Guangdong 42,159 21,123 11,180.55 487.84 436.33 Shanghai 21,758 11,337 27,187.57 1,299.76 478.07 Beijing 54,348 10,344 17,936.32 3,932.56 2,192.51 Zhejiang 29,197 10,316 12,906.56 624.27 483.68 Shandong 37,082 10,019 9,408.97 408.44 434.09 Jiangsu 34,983 8,211 11,538.99 470.33 407.60 Liaoning 38,768 7,151 11,017.23 914.77 830.31 Sichuan 27,046 4,496 4,814.86 324.72 674.41 Fujian 11,027 4,211 11,293.86 317.69 281.29 Hunan 26,400 4,117 5,732.76 409.94 715.08 Hebei 20,584 3,848 7,545.91 305.22 404.48 Correlation with GDP per capita 0.54 0.78 Sources: The Patent Office, Annual Report, 2000; State Statistical Bureau, China Statistical Yearbook 2000; Author's calculations. regional disparities in income levels, ranging from and applications per million yuan of GDP is higher Sichuan at the bottom to Shanghai at the top. The (0.78). Accordingly, higher incomes are associated difference between them is a factor of 5.6, which is with greater innovation propensities, which in turn extraordinarily high for regions within a country. raise regional economic growth. It is interesting to correlate per-capita GDP with Data for trademark registrations not shown here the relative patent application figures; there is a tell a similar story. In particular, trademark applica- strong positive correlation (0.54) between GDP per tions have risen rapidly since 1994, especially those capita and patent applications per million people. through the Madrid Protocol. Far more domestic Thus, richer provinces apply for more patents marks are registered than foreign ones, but foreign (develop more products) per person than poor applications have increased at a faster pace. In provinces. The correlation between GDP per capita terms of regional performance, Guangdong had Intellectual Property Rights in the WTO Accession Package: Assessing China's Reforms 65 the largest absolute number of applications in IPRs limits incentives to develop products and 2000, followed by Zhejiang and Jiangsu. Scaled by brand names, especially on the part of small and population, however, Shanghai ranked first by a medium-sized domestic enterprises. This structural large margin, followed by Beijing, Zhejiang, and difficulty likely limits entry of new firms and the Guangdong. There is a very high correlation (0.81) development of entrepreneurial skills. It also between per-capita GDP and per-capita applica- restricts the ability of enterprises to market nation- tions, reflecting again that trademark applications ally and to take advantage of economies of scale,and rise with income levels. tends to reduce investment in quality improve- Thus, the use of patents and trademarks is rising ments. Over time, this situation could make it rapidly in China for several reasons: increasingly difficult to break into export markets for high-quality and high-technology goods. · Laws have been strengthened and fees reduced, Second, Chinese enterprises and research organ- encouraging more applications. izations are engaging in more innovation, as · As trademark and patent infringement have suggested by the patent and trademark statistics. increased, both domestic and foreign enterprises However, the country remains behind global stan- recognized the importance of establishing intel- dards in allocating resources to R&D and science lectual property protection, even in an environ- (Maskus, Dougherty, and Mertha 1998). Moreover, ment of weak but improving IPRs. interviewees noted that SOEs and state research · Chinese markets are getting deeper as income institutions face structural difficulties in commer- grows, despite the substantial barriers to inter- cializing the results of invention. This points out regional integration, And registration of IPRs is the importance of continuing to develop a technol- important for exploiting deeper markets. ogy innovation system that encourages innovative · Chinese research organizations and enterprises activity.10 The state has important roles to play in are engaged in more inventions, and Chinese promoting pre-competitive research and removing firms are undertaking more innovative activity. disincentives to commercialization.China has made progress toward these goals, with support programs in information technology, biotechnology, and Conclusions and Recommendations other important areas, along with efforts to raise the In recent years China has made significant progress flow of knowledge from institutes and universities on the legislative end of intellectual property rights, to producing enterprises. Nonetheless, ambiguities especially in preparation for its entry into the remain about effective ownership of intellectual WTO. The specific standards it has adopted across property rights. This is another reason that the new the range of intellectual property regimes are system of IPRs should be an important component largely consistent with what might be recom- of the evolving innovation system. mended for middle-income developing countries Third, stronger IPRs alone are not sufficient to with strong innovation potential. However, in some establish effective conditions for further technology dimensions the new Chinese standards may be development and growth. Rather, they must be overly protective for an economy that remains embedded in a broader set of complementary largely a net importer of new technology and infor- initiatives that maximize the potential for IPRs mation. At the same time, China continues to expe- to be pro-competitive over the long term. An rience severe enforcement problems. Enforcement important complement is development of human is likely to be problematic for the intermediate term capital through education in science, technology, because of structural difficulties with the system and and law, and acquisition of skills through training because costs of copying and counterfeiting remain in enterprises. Enterprises should be more willing small relative to prices of legitimate products. to undertake such training under an improved China is undergoing a long process of increasing IPRs regime. Both directly and indirectly, then, sophistication in technology use and development. effective IPRs can help Chinese enterprises raise Three important problems arise with IPRs that their technological capabilities, which is critical for may influence the pace and characteristics of this adaptation of foreign technologies and innovation transformation. First, inadequate enforcement of of new products. 66 China and the WTO Another supporting factor is to ensure that com- Notes petition on domestic markets is sufficient to prevent 1. See "Draft Protocol on the Accession of China," July 10, stronger IPRs from becoming a damaging source of 2001. market power. Further enterprise reform and dereg- 2. For policies taken before 2003, the description in these paragraphs relies on information from an interview with a sen- ulation are important in this context. Over time the ior judicial official in Beijing and on Lehman, Lee and Xu, China liberalization commitments made in the WTO will Intellectual Property Newsletter: Special Issue, 2001. provide important competition as well. 3. For extensive reviews see Evenson and Westphal (1997), Finally, China like other countries, has the right Maskus (2000a), and Primo Braga, and others (1998). 4. UNCTAD (1996) presents estimates of such costs in to safeguard its interests in competition and social several developing countries. objectives through effective regulation of IPRs as 5. Prices of copyrighted goods have fallen sharply in Taiwan those rights become stronger. Thus, the govern- since the aggressive crackdown on counterfeiting in the mid- 1990s, in part because of additional competition from legitimate ment should further analyze the appropriate form local developers. of pricing regulations and compulsory licensing 6. The European Union has moved recently toward the in its drug procurement programs as medicines provision of patent protection. 7. It is conceivable that the increase in Chinese applications receive stronger protection. come primarily from joint venture partners of foreign enter- Moreover, an opportunity arises for China to prises but unfortunately the available data do not make this consider what form of competition regime it will distinction. implement as it shifts further toward the market. 8. Because these data were taken from WIPO, they include applications made under the Patent Cooperation Treaty. Currently China tries to maintain competition 9. Dahlman and Aubert (2001) discuss this in detail. through centralized regulation of market structure, 10. See Maskus (2000a) for more detailed discussion. ownership, and innovation, a system that will become increasingly incompatible with needs for References technological change. A shift toward antimonopoly regulation of such IPRs abuses as monopoly pric- Caves, Richard E. 2000. Creative Industries. Cambridge: Harvard University Press. ing, restrictive licensing arrangements, and refusals Coe, David T., Elhanan Helpman, and Alexander Whoffmaister. to deal is important. China also may need to employ 1997. "North-South R&D Spillovers." The Economic Journal compulsory licenses for this purpose. To be effec- 107: 134­49. tive, such regulation needs to be well defined, non- Dahlman, Carl J., and Jean-Eric Aubert. 2001. China and the Knowledge Economy: Seizing the 21st Century. Washington, discriminatory, and professionally applied by the D.C.: World Bank. competition authorities and courts. This points Dougherty, Sean M. 1997. "The Role of Foreign Technology in again to the need for building legal expertise in IPRs Improving Chinese Productivity." MIT Science and Technol- ogy Initiative, Beijing. over the long term. Evenson, Robert E., and Larry E. Westphal. 1997. "Technological The authorities in China expect that the stronger Change and Technology Strategy." In Handbook of Develop- IPRs regime set in place by new legislation and ment Economics: Volume 3. Amsterdam: North-Holland. Jiang, L. 1996. "Technological Innovation in Business Strategy." increased efforts at enforcement will support Science and Technology International 1: 60­3. dynamic gains in technology acquisition and inno- Lacroix, Sumner, and Denise Eby Konan. 2002. "Intellectual vation. This outcome seems achievable in those Property Rights in China: The Changing Political Economy regions and sectors that are technologically of Chinese-American Interests." The World Economy 25: 759­88. dynamic. At the same time, however, substantial Lanjouw, Jean O. 1997. "The Introduction of Pharmaceutical differences in incomes, education, and entrepre- Product Patents in India: `Heartless Exploitation of the Poor neurship persist among regions, while much of the and Suffering'?" Discussion paper no. 775. New Haven Connecticut, Economic Growth Center: Yale University. country remains poor. Developing a comprehensive Lee, Jeong-Yeon, and Edwin Mansfield. 1996."Intellectual Prop- approach to regulating the use of IPRs that helps erty Protection and U.S. Foreign Direct Investment." Review bring these poorer regions and groups more fully of Economics and Statistics 28: 181­6. Ma, Chi, and Gao Chang Lin. 1998. "Technological Innovation into the modern commercial system will pose a in China's Manufacturing," State Science and Technology significant challenge for China. In the long run, Commission, Beijing, manuscript. however, the technological dynamism that should Mansfield, Edwin. 1995. "Intellectual Property Protection, be facilitated by the new IPRs regime will greatly Direct Investment, and Technology Transfer." Discussion paper 27. Washington: International Finance Corporation. assist the country in its efforts to spread widely the Maskus, Keith E. 1998. "The International Regulation of Intel- gains from competition. lectual Property." Weltwirtschaftliches Archiv 123: 186­208. Intellectual Property Rights in the WTO Accession Package: Assessing China's Reforms 67 ------. 2000a. Intellectual Property Rights in the Global Economy. Smith, Pamela J. 1999. "Are Weak Patent Rights a Barrier to U.S. Washington D.C.: Institute for International Economics. Exports?" Journal of International Economics 48: 151­77. ------. 2000. "Intellectual Property Rights in Lebanon." In B. UNCTAD (United Nations Conference on Trade and Develop- Hoekman and J. Zarrouk, eds. Catching Up with the Compe- ment). 1996. The TRIPS Agreement and Developing Coun- tition. Ann Arbor: University of Michigan Press. tries, Geneva: UNCTAD. Maskus, Keith E., and Christine McDaniel. 1999. "The Impacts Watal, Jayashree. 1999. "Pharmaceutical Patents, Prices and of the Japanese Patent System on Post-War Productivity Welfare Losses: A Simulation Study of Policy Options for Growth." Japan and the World Economy 11: 557­74. India under the WTO TRIPS Agreement."Washington, D.C.: Maskus, Keith E., Sean Dougherty, and Andrew Mertha. 1998. Institute for International Economics. Photocopy. "Intellectual Property Rights and Economic Development in World Bank. 2001. Global Economic Prospects 2002: Making Trade China," manuscript. Work for the World's Poor. Washington, D.C.: World Bank. Maskus, Keith E., and Mohan Penubarti. 1995. "How Trade- WTO (World Trade Organization). 2003. "Transitional Review Related Are Intellectual Property Rights?" Journal of Interna- Mechanism of China: Communication to the Council for tional Economics 39: 227­48. Trade-Related Aspects of Intellectual Property Rights," Potter, Pitman B. 2001. The Chinese Legal System: Globalization IP/C/W/415. and Local Legal Culture. London: Routledge. Yang, Guifang, and Keith E. Maskus. 2001."Intellectual Property Primo Braga, Carlos A., Carsten Fink, and Claudia Paz Rights and Licensing: An Econometric Investigation." Sepulveda. 1998."Intellectual Property Rights and Economic Weltwirtschaftliches Archiv 137, 1, 58­79. Development." Manuscript. Washington, D.C.: World Bank. Scotchmer, Suzanne. 1991. "Standing on the Shoulders of Giants: Cumulative Research and the Patent Law." Journal of Economic Perspectives 5, Winter, 29­42. 5 China's Agricultural Development and Policy Readjustment after Its WTO Accession Chen Xiwen At the time of its accession to the World Trade to 15.9 percent, 4.6 percent lower than in 1995. The Organization (WTO), China was following the reduced share of agriculture in the national econ- ninth Five-Year Plan for the Development of the omy, which is a normal development in a period of National Economy and Society. The Plan required accelerated industrialization, does not alter the China to accomplish two major goals in its national important position of agriculture in the national economic and social development: to establish the economy. The agricultural growth was remarkably system of the socialist market economy at the pri- higher than the population growth; annual popula- mary stage and to attain a higher standard of living. tion growth on average was 0.91 percent, having Accomplishing these two goals would provide an provided a reliable guarantee for improvements in important foundation for reform and development the standard of living, especially diet and nutrition. and for the accomplishment of the third step, the The agricultural and rural economic structures strategic goal of modernization. While remarkable have improved gradually.In agricultural production, achievements had been made in agriculture and the share of forestry, animal husbandry, and fishery rural economy during the ninth Five-Year Plan, sig- continued to increase compared to crop farming. nificant problems had developed that could not be Promoted by technical progress, the capacity of ignored. crop farming to ensure provision of farm produce and support agricultural restructuring is increasing (table 5.1). Agriculture and Rural Development At the same time, the percentage of farmers' during the Ninth Five-Year Plan incomes derived from secondary and tertiary indus- tries has been increasing. In 1995, the average net During this period, the gross value of agricultural income from secondary and tertiary industries production in China grew from RMB1199.3 billion accounted for 32.65 percent of the net productive in 1995 to RMB1421.2 billion in 2000, an average incomes; this percentage increased to 46.65 percent annual increase of 3.5 percent in constant prices, in 2000. The value-added of township and village 0.5 percent lower than annual growth during the enterprises (TVEs) was equivalent to 121.7 percent eighth Five-Year Plan. In 2000, the share of agricul- of the gross value of agricultural production in ture in Gross Domestic Product (GDP) decreased 1995 and increased to 191.1 percent in 2000. 69 70 China and the WTO TABLE 5.1 Changes in Farm Output key link" was stressed for years to alter the situation Value Shares (percent) of food shortage, but the problem was not resolved. The grain rationing system was terminated as late Crop Animal as 1992, producing improvements in the food Year Farming Forestry Husbandry Fishery supply situation. Beginning in 1995, China had 1995 58.43 3.49 29.72 8.36 enjoyed good harvests in grain production for five 2000 55.68 3.76 29.67 10.89 consecutive years. In the first four years of the Plan period, the annual grain production surpassed Source: China Statistical Yearbook 2001. 500 million tons, leading to a continuing situation of supply exceeding the demand. The average standard of living of farmers con- Four causes contributed to the significant growth tinued to improve. In the ninth Five-Year Plan, even in grain production during the ninth Five-Year though the growth of farmers' income slowed Plan period: down year by year, the per capita net income increased from RMB1,577.7 in 1995 to RMB2,253.4 · The basic rural policies were improved continu- in 2000, an average annual increase of 4.7 percent ously. The household contract responsibility sys- after adjustment for price changes. This growth ex- tem featuring a combination of centralization ceeded the target of 4 percent set in the outline of and decentralization played an important role the ninth Five-Year Plan. Farmers' standard of liv- in expanding grain production. In November ing also continued to improve. In 1995, spending 1993, the central government explicitly extended on food made up 58.62 percent of per capita living the land contract for another 30 years after expenses of farmers; this percentage was reduced to the expiration of the original 15-year contract 49.13 percent in 2000. Compared with 1995, elec- period. This policy gave farmers a sense of sta- tric fans owned by every 100 farmers' households bility in terms of land contract rights and stimu- increased by 34 sets, black-and-white televisions by lated their initiative to increase investment 38 sets, color televisions by 32 sets, refrigerators by in farmland. Many farmers have been digging 7 sets, washing machines by 12 sets, and motor motor-pumped wells and irrigation ditches on cycles by 17 sets. The per capita housing space the contracted land and purchasing sprinkler increased from 21 to 24.82 square meters. irrigation devices to develop water-saving irriga- The most remarkable achievements were the tion. The area of irrigated farmland increased by significant expansion of the production capacity of 68.55 million mu or 9.3 percent. grain and other major farm produce and the histor- · Increases in the contract purchase price for grain ical transition from the long-standing short supply provided an incentive for farmers to increase pro- of major agricultural products to balance in aggre- duction. In 1994, consumer prices rose. To make gate and surplus in years of good harvests. up for the effect of these rises on farmers, the The achievements in grain production made government raised the contract purchase price during the ninth Five-Year Plan period are of his- for grain by 40 percent; the price was further torical significance to the development process of increased to 42 percent in 1996. the Chinese economy and society. The pressure of · Scientific and technological progress has been population growth on demand for food has been responsible for nearly 40 percent of the agricul- a significant economic and social problem con- tural growth in China. The extension of several fronting China for close to 400 years. In the late key technologies, in particular, played an impor- Ming Dynasty and the early Qing Dynasty, China's tant role in increasing grain production. The population was less than 100 million and had 800 "seed project" introduced a large number of million mu of farmland. By 2000, the total popula- improved varieties and eliminated poor varieties tion had grown to 1.266 billion, while the farmland that did not suit market needs. The "White area only expanded to 1.924 billion mu. The popu- Revolution" extended mulching technology in lation grew over 11-fold, while the farmland area the cold and dry areas in the northern part of increased by a factor of little more than two. After China to advance sowing time and prolong the the country's founding in 1949,"taking grain as the period of crop production while improving soil China's Agricultural Development and Policy Readjustment after Its WTO Accession 71 moisture conservation and repressing weed 4.3 percent in 1998, 3.8 percent in 1999, and growth. Water-saving irrigation technologies 2.1 percent in 2000. have been developed, and demonstration coun- The decreasing growth in farmers' net income in ties where these technologies are being applied recent years is only a symptomatic manifestation of have been designated. The flood irrigation prac- the problem. The severity of the problem lies pri- tice has been replaced by wide application of marily in the sustained decrease of income from spraying, trickle, and micro-irrigation methods agricultural production. Table 5.2 shows the com- to meet the needs of crop growth with less water. position of and changes in per capita net income of · Favorable climatic conditions produced adequate farmers from 1997 to 2000. rainfall that alleviated long-standing drought It can be clearly seen from Table 5.2 that, while conditions and facilitated increases in grain pro- the productive net income of farmers in 2000 duction. The most serious threat to agricultural increased by RMB142 over 1997, the net income production in China traditionally has been from agricultural production decreased by 132, a drought. reduction of 10.4 percent. Between 1998 and 2000, the average farmer's income from agricultural Generally, the agricultural policies during this production declined progressively, decreasing by period were stable and explicit, and many of the RMB30 between 1998 and 1997, RMB57 between scientific achievements facilitated growth in grain 1999 and 1998, and further by RMB44 between production. However, factors such as major rise in 1999 and 2000. According to the statistics in the prices and favorable climatic conditions are incon- agricultural survey, 59 percent of the rural house- sistent, and reliance on a confluence of such favor- holds in China are "pure farmer households" (i.e., able factors in the future would be imprudent. the family members are mainly engaged in farm- Accordingly, the long-term development of grain ing), and another 18 percent are "farmer house- production remains an issue meriting ongoing seri- holds with combined occupations" (i.e., the ous attention. number of family members engaged in farming is more than those engaged in nonfarm occupations). These two types of households with farming as The Biggest Problem in China's Rural their main source of income account for 78 percent Economy: Raising Farmers' Incomes of households in rural China. Accordingly, due to The decreasing growth in farmers' income was the the progressively reduced income from agriculture most outstanding problem during the ninth Five- in the last three years, the per capita net income of Year Plan period. Changes in the patterns of supply the households with farming as the main source of and demand produced an excess of supply for most income actually declined. farm produce, leading to declines in prices and new It was under just such circumstances that part difficulties in raising farmers' incomes. In 1996, of the agricultural provinces and the western farmers' net income increased by 9 percent, but the provinces with underdeveloped nonfarm industries rate of growth declined to 4.6 percent in 1997, faced a severe reduction in the per capita income of TABLE 5.2 Sources of Farmers' Incomes, 1997­2000 Productive Net Income from Net Income from Net Income from Net Income Primary Industry Secondary Industry Tertiary Industry Year (Yuan/Person) (Yuan/Person) (Yuan/Person) (Yuan/Person) 1997 1987 1,268 438 281 1998 2040 1,237 499 303 1999 2079 1,180 564 334 2000 2130 1,136 598 395 2000 vs. 97 +142 -132 +160 +113 72 China and the WTO TABLE 5.3 Declines in Per Capita Farm Incomes in Six Key Provinces Yuan Yuan Yuan Yuan Yuan Yuan Reduction in Shanxi Liaoning Jilin Heilongjiang Gansu Xinjiang income in 1999 86 79 123 87 36 127 Reduction in Guangxi Liaoning Jilin Heilongjiang Shaanxi Ningxia income in 2000 184 145 238 18 12 30 TABLE 5.4 Changes in Farm Household Spending, 1997­2000 Household Total Operation Living Spending Spending on Spending Spending Expenses on Food Garments Year (Yuan/Person) (Yuan/Person) (Yuan/Person) (Yuan/Person) (Yuan/Person) 1997 2,537 706 1617 890 109 1998 2,457 652 1590 850 98 1999 2,390 600 1577 829 92 2000 2,652 654 1670 821 96 farmers. In 1999 and 2000, farmers in six provinces There were successive declines in 1998 and 1999 and regions experienced decreases in their per in per capita spending on farm household opera- capita net income over the previous year. Table 5.3 tions, with a cumulative 15 percent reduction shows the provinces and amounts of the reduction between 1997 and 1999 (a reduction of RMB106). in income. The increase in per capita spending on farmer Per capita income decreased for two consecutive household operations in 2000 constituted some- years in the provinces of Liaoning, Jilin, and thing of a recovery, but it still left spending over Heilongjiang. The reduction in Jilin was as much as 7 percent lower than in 1997. On the other hand, RMB361 in the same period, and per capita income this increase was indirect, the result of increases in in the province was 15 percent lower in 2000 than drought relief expenses and the rise in prices for in 1998. fuels. The net income from farming declined for years The changes in the per capita consumer spend- running, while the tax burden on farmers' house- ing of farmers have similar characteristics: succes- holds mainly included taxes on agricultural and sive declines in 1998 and 1999, with spending in animal products, taxes on special products, and 1999 at 2.5 percent (RMB40) below the level in a contract land deduction. These factors have 1997. Although the per capita living expenditures deepened the feeling of households in which farm- increased by RMB93 in 2000 compared with the ing is the main source of income that "income can previous year, the change was to a large extent not go up and burdens can not come down." This passive. First, the per capita spending on food con- situation has forced some farmers to seek work tinued to decrease after the successive decrease for from other sources to offset payment of agricul- two years, while per capita spending on garments tural taxes and fees with income; more farmers increased slightly but was still lower than the levels have had to reduce their spending on production in 1997 and 1998. Second, much of the spending and living expenses. Since 1998, the per capita increase resulted from the rise in service prices, spending on operation and living expenses has such as medical expenses that increased by RMB18 decreased or stopped increasing. Table 5.4 shows or 25 percent, transport by RMB24 or 35.5 percent, the changes in this regard. and education and entertainment devices and China's Agricultural Development and Policy Readjustment after Its WTO Accession 73 TABLE 5.5 Share of Farmers in Consumer WTO Accession and the New Spending at County and Lower Stage of Development of Chinese Levels (percent) Agriculture and Rural Economy Year Percent The Initiation of the New Stage and its Central Tasks 1996 40 In October 1998, the Third Plenary Session of the 1997 39 15th Central Committee of the Communist Party 1998 39 of China (CPC) adopted the Decisions of the CPC 1999 39 Central Committee on Major Issues of Agriculture 2000 38 and Rural Work. The judgment of balance in aggre- gate and surplus in harvest years for grain and other major agricultural products was made in the services by RMB18 or 11 percent. The above expen- Decisions. Based on this judgment, the central gov- diture increases, which were basically passive, ernment stated at the National Conference on increased by RMB60 or 65 percent of the total Rural Work at the end of 1998 that the develop- increase in living expenses. Therefore, even though ment of Chinese agriculture and rural economy farmers' per capita spending on living had had entered a new stage. The main basis for the new increased, the amount of the spending that could stage thinking is the major change in the pattern of be used to improve their quality of life was very agricultural supply and demand. One of the key limited. features of the situation is that the quantitative The data on per capita expenditures on produc- contradictions in the supply of and demand for tion and living of farmers are only national aver- agricultural products have been relieved. Another ages, and, very often, average figures conceal many is that past agricultural production had been disparities. The situation was more serious in terms restricted by natural resources; presently, however, of the reduction in spending for most of the farmer restrictions of market demand on production have households that mainly depend on farming for become more acute. What farmers are worrying incomes. This is one of the major reasons for the about is not whether enough products can be pro- decline in farmers' share of rural consumption evi- duced, but whether the products can be sold at dent in table 5.5. good prices. Precisely because of such changes, the Thus it can be seen that the decrease and stagna- central government explicitly directed at the end of tion in farmers' spending has actually produced 1999 that the central task for agriculture and rural negative impacts on the expansion of the domestic economy in the new stage should be implementa- market and healthy development of the national tion of strategic restructuring. economy. As the central government pointed out at the Central Working Conference on Rural Issues, The Crux of the Strategic Restructuring the issue of increasing tangible benefits to farmers of Agriculture and Rural Economy is central to the continuation of the progress achieved in agriculture productivity. If the purchas- Since China adopted reform and open economic ing power of farmers does not continue to improve, policies, Chinese agriculture has experienced the policy for expanding domestic demand will not restructuring. The starting point of the strategic achieve the desired results. If farmers' living stan- restructuring lies in the adjustment of variety and dards cannot be improved, the risk of increased quality instead of quantity. The adjustment in the social instability in the rural areas will increase. If area of crop sowing is no more than which crop the ability of agriculture to save and accumulate should have more sowing area and which should cannot be constantly strengthened, it will be diffi- have less. This kind of adjustment was effective in cult for China's agricultural products to compete the past, is effective at the present, and will be effec- in the global marketplace. If Chinese agriculture tive in the future. However, adjustment in quantity experiences major setbacks, the development of has major limitations. For example, judging from the whole national economy and society will be the existing production level and consumer demand affected adversely. for grain in China, the total sowing area for grain 74 China and the WTO crops should be maintained at 1.65 billion mu to conference in Anhui Province on the reform exper- 1.7 billion mu, with only about 50 million mu as a iment in rural taxes and fees to explore fundamen- margin for adjustment. tal solutions to the burdens on farmers. At the turn Strategic restructuring differs substantially. of the century, the central government held a series Adjustment can be made on every piece of land of major conferences on rural restructuring, within more than 2.3 billion mu of sowing area, increasing farmers' income, and alleviating burdens such as replacing the unmarketable poor quality on farmers. varieties with high quality marketable ones. Adjust- Securing income growth for farmers is by no ment can be made not only on cultivated land, but means a simple matter. The most important also in forest products, fruits, animal products, and immediate cause of the slow increase in farmers' aquatic products. Accordingly, adjustments empha- income is the current difficulty in selling agricul- sizing optimizing varieties, improving quality, and tural products and the declines in prices. However, efficiency can be made everywhere. Only by such problems of product sales and price declines are strategic restructuring focused on quality and effi- not only evident in agriculture, but in most other ciency can the overall quality of Chinese agriculture industries as well. Therefore, the current difficul- be improved to meet the requirements and chal- ties facing farmers' income are to a great extent lenges posed by WTO accession and economic glob- related directly to the economic cycles at both alization, to ensure improvement in the people's home and abroad. Achievement of sustained living standards, and to accelerate the process of growth in farmers' income is a complicated project agricultural modernization in China. of system engineering. One or two specific policies cannot be expected to provide fundamental solu- tions. The key is to make great efforts in all aspects The Fundamental Objective of the Strategic of agricultural and rural work. This also depends Restructuring to a large degree on the improvement in the The new stage of strategic restructuring was initi- national economy as well as the world economy so ated early in 2001 when the central government as to create a broader market for the agricultural specified that the fundamental objective of the development. strategic restructuring is to ensure increases in farmers' income. After entering the new stage and The Basic Direction of the Strategic Restructuring basically solving the quantitative contradiction of Agriculture and the Rural Economy in agricultural supply and demand, increasing farmers' income has increasingly become a priority As directed by the Chinese government, the central of the Chinese central government. In November task of the new stage of agricultural development is 2000, Secretary General Jiang Zemin and Premier to implement strategic restructuring of agricultural Zhu Rongji stressed at the National Economic production. This restructuring is strategic for two Conference that increasing farmers' income should reasons. The first is to put the long-standing prac- be a key goal in the entire economic work program. tice of pursuing quantitative growth in agricultural The government has paid close attention to issues products on the track of stressing optimized variety concerning agriculture, rural areas, and farmers, and improved quality and efficiency. The second is with the focus on increasing farmers' income. The that the ongoing development of Chinese agricul- fundamental objective set at the Central Working ture needs to continue to consider the constraints Conference on Rural Affairs at the beginning of on the resource side and to pay more attention to 2001 was to achieve growth in farmers' income. constraints from demand side. To achieve this, the In mid January 2001, the central government con- regulation of agricultural production by market vened a National Conference on Agricultural forces should be expanded. Science and Technology, at which a group of agri- It can clearly be seen that WTO accession will cultural scientists was commended and the Outline accelerate agricultural restructuring and adjust- for the Development of Agricultural Science and ment in agricultural policies. The agricultural Technology in the Next 10 Years was adopted. In restructuring consistent with the above goals will February 2001, the State Council held a working require changes in the relevant policies. Although China's Agricultural Development and Policy Readjustment after Its WTO Accession 75 these policy adjustments primarily resulted from 2. Expand processing of agricultural products. major changes in the patterns of agricultural supply Agricultural processing will become more and more and demand in China, they are fundamentally important in agricultural development in China. consistent with the direction of China's WTO com- With continuous increases in people's incomes and mitments. Even if no WTO accession issues are pre- living standards, the Engel coefficients and income sented, such adjustments in agricultural policies elasticity of demand can be expected to fall, which would be required. could pose a grave challenge to the continued devel- The restructuring of agriculture and rural econ- opment of agriculture. omy, revealed at the end of the ninth Five-Year Currently, the main factor restricting the growth Plan, has five major orientations: of farmers' income is the decrease of income from agricultural production. This is related to the stage 1. Optimize varieties of farm produce and of social and economic development in China. improve the overall product quality. Optimization Profound changes have taken place in the current of varieties and improvement in quality and effi- consumption structure of residents. One is the ciency should be the main orientation of adjust- decrease of the Engel coefficient and the associated ment for all agricultural production, including crop decline in the income elasticity of demand for food. farming, animal husbandry, fishery, forestry, and In 1991, the Engel coefficients of China's urban and fruit production. With the issue of quantity in agri- rural residents were 54 percent and 58 percent, cultural supply and demand resolved, optimizing respectively. These declined to 38 percent and varieties and improving quality will constitute an 48 percent in 2001. objective requirement for the current stage of agri- The decline in the propensity of households to cultural development facing supply and demand spend their income on food is evident in Table 5.6. changes. It is preferable in agricultural production Between 1996 and 2001, per capita disposable to have smaller quantities of higher quality prod- income increased by RMB2 021 and consumer ucts to avoid wasting the human, financial, and expenditure increased by RMB1,390, while food material resources of farmers. To this end, the cen- expenditure only increased by RMB109, even tral government announced the termination of though expenditures on food eaten outside the purchasing at protective prices for long-grained home increased by RMB128. Thus, the urban resi- nonglutinous rice, northeastern spring wheat, dent expenditures on grain, oil, meat, and vegeta- southern winter wheat, and other unmarketable bles decreased by RMB131. varieties, starting in 2000. The purpose of this Only by expanding farm produce processing reform was to guide farmers to replace inferior and encouraging the consumption of higher varieties with superior ones. quality, more nutritious, and safer food can the TABLE 5.6 Changes in Urban and Rural Resident Income and Spending, 1996­2001 2001 (Yuan) 1996 (Yuan) Per capita disposable income 6,860 4839 Per capita consumer expenditure (total) 5,309 3919 Food 2,014 1905 Grain 188 272 Oil 59 69 Meat, poultry, and products 413 439 Aquatic products 152 132 Vegetables 194 207 Tobacco 104 84 Liquor and beverage 104 85 Fresh and dried fruits 131 118 Dining out 314 186 76 China and the WTO continued expansion of agricultural markets be giving play to local comparative advantage in agri- achieved. Food processing can not only increase the culture. China has a large population and little value of agricultural products, but it also can guide land. Most of the land-intensive agricultural prod- consumption and develop new markets to make it ucts, such as grain, cotton, and edible oil are rela- possible to continuously increase farmers' income. tively high-cost in China. Most labor-intensive Two concepts should be altered in terms of devel- agricultural products, such as animal and poultry oping agricultural processing industry. First, food products, aquatic products, and horticultural prod- processing should not be regarded as an industry ucts (fruits, vegetables, flowers, and bonsai), have a with low-technology content and without the need comparative advantage in international markets. for substantial investment. On the international Currently, there is an urgent need for Chinese agri- market, brand-name and marketable processed culture to foster strengths and circumvent weak- foods all have high-technology content. These nesses and to give full play to China's comparative products simply can not be produced without the advantage in international market competition. To application of advanced technology. Second, it tap the comparative advantage of Chinese agricul- should not be believed that unsaleable fresh prod- ture on the international market, it is first necessary ucts can be processed into high-quality processed to bring into full play the agricultural comparative products. The raw materials for food processing advantages of different domestic regions. In this should be special-purpose farm produce, such as way, it will be possible to create a pattern of mutual special wheat and special grapes, which are distinc- benefit and complementarity between regions to tive from those directly used for food. Therefore, improve the competitiveness of Chinese agricul- adjustments should be made in the varieties of tural products on international markets and to agricultural products to achieve better results from promote the development of China's agriculture food processing. as a whole. 3. Bring into play the local comparative advan- 4. While continuing to support the develop- tages in agriculture. In the past, widespread short- ment of TVEs, make positive and steady efforts to ages of agricultural products left localities with advance urbanization. In the second half of the no choice but to pursue higher degrees of self- ninth Five-Year Plan period, the per capita net sufficiency. Now that the supply and demand situa- income of farmers increased, although income tion has changed, local authorities should be from farming declined. This increase depended liberalized from such traditional thinking and almost entirely on income growth from nonfarm bring into play local comparative advantage in sectors. This shows that the fundamental way to agriculture. Particularly, the coastal regions and enrich farmers is to reduce the number of farmers. suburbs of large and medium cities should make Agricultural resources are limited and efforts use of capital and technology-intensive techniques. should be made to transfer rural labor and popula- These localities should properly reduce their grain tion to allow the farmers remaining on the land to production and develop higher value-added animal expand their scale of production. The township husbandry, fishery, and horticulture so as to give and village enterprises (TVEs) in the countryside more market space to the main grain-producing have provided 130 million job opportunities to regions in a pattern of mutual benefit and relative farmers, and their role in alleviating the unem- complementarity. At the same time, the various ployment problem can hardly be replaced. TVEs localities should be encouraged to develop their should be guided to conduct restructuring and own agriculture with distinctive features. institutional innovation to promote their further To promote restructuring of the regional distri- development and to make a greater contribution bution of agriculture, the central government to increasing the employment and income of decided in 2001 to liberalize grain markets in the farmers. main coastal grain-consuming regions, including Another major source of income growth of Beijing, Tianjin, Shanghai, Jiangsu, Zhejiang, farmers during the ninth Five-Year Plan was tran- Fujian, Guangdong, and Hainan. Particularly, in the sient employment of farmers in towns and cities. face of the challenges and opportunities after WTO This part of income has provided an increasing accession, much more attention should be paid to share in the net income of farmers. According to an China's Agricultural Development and Policy Readjustment after Its WTO Accession 77 estimation by Sichuan, 6 million farmers in the towns. Medium and small cities should be devel- province seek transient employment and remit oped, the regional center cities should have distinc- nearly RMB30 billion each year. The transient tive features, and large cities should bring into play employment of farmers, which should be encour- their leading role and growth pole functions. aged, should be managed to keep it orderly. How- However, measured with the goal of transferring ever, many large and medium cities have set up the rural population, developing small towns is various restrictions and rules on employing farm- more realistic. Presently, large and medium cities ers, and many of these rules and regulations are frequently provide job opportunities for transient both unreasonable and ineffective. On one hand, farmers and opportunities for them to accumulate these transient farmer workers should be guided to capital, but it is very difficult for these farmers to flow in an orderly manner and to abide by relevant settle in these cities. After a certain period of capital laws and regulations. On the other hand, the vari- accumulation, rural residents typically leave the ous rules and regulations restricting farmers seek- large cities, but not all of them return to rural areas. ing jobs in urban areas should be reviewed so as to Some of them settle in small towns. This process of alleviate the burden on the transient farmers. In the resettlement will provide practical opportunities process of economic development in China, the for the development of small towns. However the transient employment of farmers will certainly be a development of small towns should not rush head- long-term process. A large number of farmers will long into mass action and "blossom" everywhere, be transferred to nonfarm employment, but the or the results will be counterproductive. There are transition to settlements in towns will occur over about 45,000 towns and townships in China, in time and will require an extended period of tran- which 19,000 are governmentally nominated town- sient employment. Conscientious research should ships. Considerations of how many small towns be conducted, and effective policies should be should be established should proceed from realistic adopted to guide and manage such a large scale and assessments of China's actual conditions and long-term population flow. national strength. Towns with better basic condi- Developing small towns is an important way to tions and development potential should be given transfer rural population and accelerate the process priority in construction. The most important pre- of urbanization in China. In the decisions adopted condition for the development of small towns is a at the Third Plenary Session of the 15th Central prosperous local economy. Committee of the Chinese Party Congress, it was Construction of small towns differs from that specified that developing small towns would be a of residential quarters. A residential district can major strategy for the development of the rural accommodate several thousand residents and has economy and society. sound living environments and complete living There are two schools of thought on the specific facilities; however, most of the residents are not ways in which urbanization should be advanced. employed within the district, that is, residential dis- The first places the emphasis on large and medium tricts have no function to create job opportunities. cities. Through a host of analyses and researches, The basic function of small towns is to attract the many scholars think that large cities can generate transferred farmers and provide job opportunities. higher economic benefits and provide a large num- Accordingly, emphasis should be laid on giving play ber of job opportunities. The second school of to functions of towns; the basic function is to create thought stresses the effect of developing small a prosperous local economy and to provide towns. Its advocates believe that the basic goal of employment. The most basic judgment on whether urbanization at the current stage is to speed up the conditions exist for constructing a town is whether transfer of rural surplus labor and rural popula- a planned town can become a source of economic tion, and these advocates view developing small growth. Only by making this judgment can we towns as a more effective means of achieving this achieve the goal of accelerated transfer of rural sur- goal. plus labor and population, provide better condi- The path to urbanization should be one with tions for agricultural and rural development, and Chinese characteristics, featuring coordinated provide more opportunities for income growth of development of large and medium cities and small farmers. 78 China and the WTO 5. Intensify the construction of ecological farmers to produce high-quality, low-residue, environment and implement sustainable devel- and safe farm produce opment. In ecologically vulnerable regions such as · adjust the existing supportive and protective those along the upper reaches of major rivers policies for agriculture and establish a domestic and arid areas, the government should provide protective system for agriculture that conforms financial aid for farmers to return grain plots to to WTO rules forestry to conduct ecological construction includ- · accelerate the innovation of agricultural man- ing re-vegetation and soil and water conservation. agement system; develop an integrated agri- Starting in 1999, experiments on returning grain cultural management system with "company plots to forestry were conducted in some provinces plus farmer households" as the main form to and regions. By the end of 2001, 11 million mu of improve farm organization and farmers' access grain plots had been returned to forestry nation- to markets wide. In 2002, the central government officially · further reform and perfect the circulation sys- adopted policies for returning grain plots to tem for grain and other major agricultural prod- forestry. The government will subsidize 300 jin of ucts to reduce circulation expenses and improve grain to farmers in the southern part of China and competitiveness 200 jin to farmers in the north for every 1 mu of · further restructure the rural financial system grain plots returned to forestry, plus RMB50 for and explore the establishment of an agricultural each mu incurred in sapling and grass seed insurance system that suits the national condi- expenses and RMB20 in living expenses. Under this tions in China policy, implementation of returning grain plots to · restructure the agricultural science and tech- forestry not only can recover balance of ecology nology research and extension systems and effectively, but it also can offer opportunities for intensify international cooperation and technol- farmers to receive direct benefits. By the end of ogy import to accelerate technical progress in 2002, the total area of returning grain plots to agriculture forestry reached 34 million mu and is expected to · reform the household registration system in small reach 100 million mu during the whole tenth Five- towns and encourage the rural population that Year Plan. has met necessary conditions to settle in small towns. Policy Orientations for Promoting Restructuring of Agriculture and Remaining Policy Issues the Rural Economy A number of policy issues needs to be examined in To promote the strategic restructuring of agricul- more detail in the future. Some of the most impor- ture and rural economy and to improve the over- tant of these issues are as follows: all quality and international competitiveness of Chinese agriculture, the government is accelerating · Correct analysis of price issues for staple agricul- a series of policy readjustments to meet the needs tural products, particularly grain, soybeans, and of current agricultural development. The main cotton. The relationship among producer prices, readjustments include the following: wholesale prices, and retail prices should be dis- tinguished, and the actual gap between prices · accelerate the establishment of market informa- for China's staple agricultural products and tion systems for agricultural products and international market prices should be analyzed provide farmers with timely, comprehensive, objectively. Monopoly still exists in the circula- accurate, and authoritative market information tion of staple agricultural products. Therefore, it on agricultural products is absolutely possible to reduce the circulation · accelerate the establishment of quality, security, expenses of China's staple agricultural products and standards systems, and improve the means by further reforming the agricultural circulation of inspection and testing for agricultural prod- system to reduce the gap between the Chinese ucts, particularly food products, guiding the and international market prices. China's Agricultural Development and Policy Readjustment after Its WTO Accession 79 · The comparative advantages of Chinese agricul- the international agricultural markets. Both tural products; labor costs give China a clear com- research and preparations for the new round of parative advantage in many animal and aquatic WTO agricultural negotiations are inadequate. products. However, given the aggravated "green The central government should set up special barriers" in international agricultural trade, low bodies to intensify research on the relevant labor costs are not enough to constitute interna- policy recommendations. tional competitiveness. The key is to find ways to produce agricultural products that meet References international quality and safety standards. At the same time, the question remains whether other "The Outline of the Ninth Five-Year Plan (1996­2000) for National Economic and Social Development and the Long- WTO members will open their markets to range Objectives to the Year 2010." Delivered at the Fifth Chinese agricultural products. Plenary Session of the 14th Central Committee of the Com- · Reforming the foreign trade system for agricul- munist Party of China. Xinhua News Agency, September 28, 1995. tural products. The key is to properly resolve National Bureau of Statistics. Statistical Yearbook of China 2002. the problems in giving nonstate firms that Beijing. China Statistics Press. have met the necessary conditions of foreign National Bureau of Statistics. Statistical Yearbook of China 2001. trade rights for importing and exporting agri- Beijing. China Statistics Press. "The Decisions of the CPC Central Committee on Major Issues cultural products. of Agriculture and Rural Work." Delivered at the Third Ple- · Both tariff concessions and tariff-rate quotas for nary Session of the 15th Central Committee of the Commu- imports of agricultural products are hard to bear nist Party of China. People's Daily. October 15, 1998. The Outline for the Development of Agricultural Science and in the transitional period. However, it is crucial Technology in the Next 10 Years: 2001­2010. People's Daily. to study how to deal with the pressures from May 24, 2001. 6 The Nature of Distortions to Agricultural Incentives in China and Implications of WTO Accession Jikun Huang, Scott Rozelle, and Min Chang The initiation of economic liberalization and struc- efficiency gains from decollectivization, the country tural change in 1978 has produced substantial still enjoyed agricultural growth rates that have out- growth in China's economy. The annual growth rate paced the rise in population. New opportunities in of gross domestic product (GDP) was 8.5 percent the off-farm sectors have allowed farm families to from 1979­84 and 9.7 percent from 1985­95. shift part of their household labor out of the agri- Despite the Asian financial crisis, the GDP growth cultural sector into higher paying off-farm jobs. continued at 8.2 percent annually from 1996 to 2000. Despite the healthy expansion of agriculture, the Although reform has penetrated the entire econ- sector still faces serious challenges. According to omy since the early 1980s, most of the successive the World Bank (2000), more than 100 million peo- transformations began with, and to some extent ple fell below the poverty line in the late 1990s, were dependent on, growth in the agricultural sec- earning less than one dollar (U.S.) per day in pur- tor (Nyberg and Rozelle 1999).After 1978, decollec- chasing power parity (PPP) terms. Some regions tivization, price increases, and the relaxation of of the nation remain highly dependent on crop local trade restrictions on most agricultural prod- production, including farmers in some Northeast ucts accompanied the substantial growth of the provinces (maize and soybeans) and the North agricultural economy from 1978 to 1984. Grain China Plain (wheat and maize) (Ministry of Agri- production increased by 4.7 percent per year. Even culture 2001). In recent years, expanding supplies higher growth was enjoyed in horticulture, live- and increased liberalization have pushed real agri- stock, and aquatic products. Although agricultural cultural prices to their lowest levels in history (Park growth decelerated after 1985 after the one-time and others 2002). With the retreat of the state The authors would like to thank the research assistance of Yuping Xie. This work could not have been done without him. We also acknowledge the helpful comments and suggestions of Kym Anderson, Fred Crook, Tom Hertel, Elena Ianchovichina, Will Martin, Francis Tuan, and participants in the World Bank-sponsored meeting on the impact of WTO accession on China. We acknowledge the financial support of the World Bank's Trade and Rural Development units. 81 82 China and the WTO occurring in many sectors of the economy, such as than any other group on the income earned from rural health care and provision of welfare services cropping, they are the ones most likely to be (Nyberg and Rozelle 1999), large numbers of peo- affected by the WTO's liberalizing measures (Chen ple remain poor and vulnerable to even relatively and Ravallion 2004). In conducting our case study, minor income shocks. we focus on the effects of the WTO accession on Agriculture has been at the center of discussion agricultural prices even though the other effects on of China's entry into the World Trade Organiza- the rural population are likely to be at least as tion (WTO), due in part to the vulnerability of important as cropping (Zhao and Sicular 2002). We parts of the rural economy and in part to the do not want to quantify the total welfare effect; importance of agriculture in the political economy instead, we focus primarily on the qualitative of a number of developed nations with whom effects on China's farmers to illustrate how to China negotiated its accession to the WTO. How- approach assessing the effects of trade liberaliza- ever, the likely shifts in China's future agricultural tion on agricultural prices. policy and their impacts are not well understood, The rest of the chapter is organized as follows. and debates on the future of China's agriculture First, after providing a brief context for our analysis continue. Some argue that the impact of WTO and discussing our data, we present measures of accession on China's agriculture will be substan- NRPs for a set of commodities for China. Next we tial, adversely affecting hundreds of millions of discuss how these distortions to agricultural prices farmers (Carter and Estrin 2001; Li and others should be expected to change as China implements 1999). Others believe that, although some impacts its WTO obligations and gains access (or not) to the will be negative and even severe in specific areas, promises that were made to China. The next sec- the overall effect of accession on agriculture will be tion analyzes the transmission of prices through modest (Anderson and Peng 1998). In part, the the economy, and the final section discusses the confusion about the ultimate impact of WTO implication of the findings. accession on agriculture can be traced to a general lack of understanding of the policy changes that Gradual Opening and Remaining accession will engender (Martin 2002). However, Distortions in perhaps an even greater way, the lack of clarity of the debate can be traced to a lack of under- Although agriculture has been at the center of standing of the fundamental facts about the nature China's negotiations over its WTO entry, the likely of the distortions to China's economy on the eve of shifts in future agricultural policy and its impacts its WTO entry. are not well understood. Debates on future food In this chapter, we provide an approach to help security are growing. The fundamental confusion researchers and trade officials better understand about the ultimate impact of WTO accession in how trade liberalization will affect agricultural part can be traced to a widespread lack of under- prices and how the price changes will be experi- standing of the policy changes that may result from enced in different parts of the nation. First, we China's WTO accession (Martin 2002). Tradition- describe a new way of collecting data that can be ally, analysts have focused on four sets of trade poli- used to design more accurate, disaggregated meas- cies, measures that are most frequently used by ures of protection (specifically, nominal rates of other countries to protect their agricultural sectors. protection NRPs). Then, we use price determina- In examining the previous work (e.g., CARD 2001; tion and market integration analyses to study the Tuan and Cheng 1999; OECD 2001), we find that nature of domestic markets to begin to understand almost all of the discussion is directed at tariffs, how price shifts at the borders that arise from trade quotas and licensing, state trading, and traditional liberalization measures affect producers in different non-tariff barriers (NTBs). It is implicitly assumed parts of the country. that the WTO agreement is focused solely on these We report the results of a case study of the policies, that these policies are responsible for most impact of China's WTO accession on agricultural if not all of the protection that China enjoyed prior prices. Since many of China's poor rural house- to accession, and that accession represents China's holds live in remote regions and tend to rely more initial assault on protection at the border. In fact, The Nature of Distortions to Agricultural Incentives in China and Implications of WTO Accession 83 while at one time these policies were the source of exports in the 1990s, approximately 20 percent of high distortions, some of the worst distortions total agricultural exports, is less than half of what it caused by these policies have already disappeared was in the early 1980s (Huang and Chen 1999). By after nearly two decades of reform in the external the late 1990s horticultural products and animal and economy. The experience of China's agricultural aquatic products accounted for about 80 percent of tariff policy illustrates the gradual but dramatic agricultural exports. These trends are even more changes that China has experienced in the past two evident when reorganizing the trade data grouping decades. In the late 1970s and early 1980s, the them on the basis of factor intensity (table 6.2). The domestic wholesale price of China's four major commodities, converted at the official exchange rate far exceeded the world price measured at TABLE 6.1 Changes in Nominal Rates China's border (table 6.1). For example, in 1978, of Protection Over Time of China's rice price was 10 percent above the world China's Major Agricultural market price. The nation's wheat and maize prices Commodities, 1978­2000a exceeded the world price by approximately 90 per- Nominal Rates of Protection (Percent) cent. However, over the next two decades, the NRP Rice Wheat Maize Soybean for rice fell to 6 percent and to about 30 percent for wheat and maize. Although the NRP for soy- 1978­79 10 89 92 40 beans fell similarly between the late 1980s and the 1980­84 9 58 46 44 mid 1990s, the rate rose in the late 1990s before 1985­89 4 52 37 39 falling to less than 20 percent in 2001 (table 6.1; Xie 1990­94 7 30 12 26 2002). During this period, the intervention by state 1995­97 1 19 20 19 traders and the use of NTBs also fell gradually 1998­00 6 26 32 49 (Martin 2002). 1998 6 22 40 37 Falling protection and changes in international 1999 9 30 33 67 trade and domestic marketing policies have resulted 2000 2 26 23 44 in dramatically shifting trade patterns. Disaggre- a gated, crop-specific trade trends show how exports Nominal rates of protection (NRPs) measured as difference (percentage) between average border and imports increasingly are moving in a direction prices and average domestic wholesale (market) that is more consistent with China's comparative prices. advantage. For example, the proportion of grain Source: Huang 2001. TABLE 6.2 China's Agricultural Trade (US$ millions) by Factor Intensity, 1985­97 Land Intensive Labor Intensive Labor/Capital Products Products Intensive Products Value Share Value Share Value Share Year (US$ Million) (Percent) (US$ Million) (Percent) (US$ Million) (Percent) Agricultural exports 1985 2,119 36.4 2,199 37.8 1,497 25.7 1990 1,689 17.7 4,971 52.1 2,881 30.2 1995 875 6.0 7,095 48.4 6,704 45.7 1997 2,158 14.1 6,538 42.6 6,642 43.3 Agricultural imports 1985 1,072 43.8 680 27.8 695 28.4 1990 4,032 71.9 642 11.5 935 16.7 1995 6,575 54.5 3,278 27.2 2,216 18.4 1997 4,644 47.3 2,179 22.2 2,987 30.5 Source: Huang and Chen 1999. 84 China and the WTO imports of land-intensive bulk commodities,such as re-export also are to be assessed the value-added grains, oilseeds, and sugar crops, have risen (or net tax.Although rates vary, the typical value-added tax exports have fallen); while exports of higher-valued, rate ranges from 13 to 17 percent. more labor-intensive products, such as horticultural For a variety of political and tax collection rea- and animal (including aquaculture) products, have sons, in the early stages of the implementation of risen (table 6.2). the tax, authorities decided to exempt farmers when they sold their products into the market. Traders that purchased grain, for example, from a Nontraditional Sources of Trade Liberalization farmer in his home or in a local market would not or Protection have to pay the value-added tax. When the good Based on the preceding discussion, two facts was resold in a downstream wholesale market, the become clear. First, distortions have declined sig- value-added tax was assessed, but the trader only nificantly in the past 20 years. Considering this fact, owed the tax on the amount of the marketing the current episode of policy reform that accompa- margin, or the difference between the procurement nied China's accession to the WTO should be con- price and sale price. Xie (2002) and Rozelle (2003) sidered an extension of past efforts. Second, much found that in China's competitive marketing regime, of the falling protection has come from decentraliz- the marketing margins (the difference between the ing authority and relaxing licensing procedures for buying and selling prices) are extremely low. For some crops (e.g., moving oil and oil seed imports shipments from markets that are 500 to 1,500 kilo- away from state trading firms), reducing the scope meters from the port, traders report margins that of NTBs, relaxing real tariff rates at the border, range from 10 to 20 percent. Assuming an average and changing quotas (Huang and Chen 1999). It is of 15 percent, the real tax rates on domestic agri- perhaps for these reasons that much research on cultural goods are only 15 percent of those on China's entry into the WTO focuses on the policies imported goods. responsible for much of the earlier progress and Such a tax system can provide some of China's studying these policy tools might be beneficial. farmers with significant protection. For example, in Undoubtedly, changes in China's tariff regimes, the case of soybeans, there is only a 3-percent tariff state trading system, and matrix of NTBs will play a on imports. In recent years, however, many traders continuing role in creating or eliminating distor- have been given the right to import soybeans. Theo- tions in China's agriculture. However, partly retically, then, the international price of soybeans because many of the gains from traditional trade upon arrival at China's borders should differ by only reforms have already been experienced, China may 3 percent. However, as soybeans enter the country, be able to affect further trade liberalization via soybean importers also must pay a 13-percent tax to other, less discussed policies. Even if traditional meet their value-added tax obligation. Domestic policies are still important, the gains from these soybeans, in contrast, are taxed, on average, at less other policy reforms may be as important as those than 3 percent. Through this means, then, China's from traditional trade reform. soybean producers have an added 10 percent of For example, China also has used its taxation price protection.1 policy to protect its agriculture, especially in certain China also has used export subsidies in recent sectors, such as soybeans, that have been most liber- years to increase exports of some commodities, alized. In the early 1990s, leaders radically revised thereby increasing protection by raising the price China's fiscal system, making it much more reliant of domestic commodities (table 6.3). Maize and for revenue generation on a value-added tax system cotton are the two crops that have received the most (Nyberg and Rozelle 1999). The theory of the tax is substantial export subsidies. In general, the subsi- that it is assessed on value added in all goods during dies are paid out of the national budget on approval their manufacture and sales process from the time of the State Council. Only approved state trading the raw material comes out of the ground until it organizations are authorized to export with the reaches the consumer. National regulations state subsidies. During interviews in the field during that imported goods that are not for immediate 2001, we found that maize exporters, especially The Nature of Distortions to Agricultural Incentives in China and Implications of WTO Accession 85 TABLE 6.3 Subsidies and Tax Rebates for however, tax policies help exporters compete on an Exports of Selected Agricultural equal basis in international markets. Based on the Commodities in China, 2001 trade ministry's estimate of the average value- added tax paid on the products exported by meat Rebate of Value-Added Tax traders, when a meat exporter executes a contract, for Exported the company can receive a rebate equal to the esti- Export Agricultural mated value of the value-added tax (table 6.4). Subsidies Commodities Commodity (Percent) (Percent) For example, pork and beef exporters receive a rebate equal to 5.2 percent of the value of their Rice 1 0 transaction (rows 4 and 5). Poultry exporters Cotton 10 0 received 13 percent rebates (row 6). When interna- Maize 34 0 tional prices for meat commodities are higher than Pork 0 5.2 those inside China, rebates encourage domestic Beef 0 5.2 Chicken 0 13 producers and traders to export, since the perceived price gap between the domestic and international price will be greater (which, of course, will also help Source: Authors' survey. raise China's domestic price as exports rise). In summary, then, as China enters the WTO, those in Northeast China, received subsidies that those officials interested in liberalizing China's trade averaged 34 percent of export price (row 3). For will face a number of challenges. Alternatively, example, one trader said that for each ton of maize China also has a number of instruments that it has that his company exported in 2001, it received back been using and may continue to use (legally or not) 378 yuan per ton (45.7 US$ per ton) after produc- in managing its domestic economy. In addition to ing an export bill of sale with the export sales price. traditional trade policies, tariffs, quotas and licens- With a sale price of 104 US$ per ton, the trading ing, state trading, and NTBs, China has protected company received a subsidy of 44 percent, 10 per- and/or has the potential to protect its agriculture centage points above the average. The total pay- with a number of other policy measures. In partic- ment received (export earnings plus subsidy) was ular, our analysis has shown that taxation policy 1,240 yuan per ton, which was about 90 yuan may still be a tool that China could try to use to higher than what could have been earned in the protect or further open its agricultural sector. It domestic market (1,150 yuan). We also discovered may also use other policies, such as sanitary and that cotton exporters received fairly large subsidies phytosanitary (SPS) regulations. In the past, it also when they exported raw cotton, up to 10 percent or has used export subsidies and rebates (when used more (table 6.3). Finally, in several isolated cases, for commodities on which the domestic value- rice exporters reported that they received small added tax is not assessed) to create wedges between subsidies (though no more than 5 percent for any the domestic and international prices of importable single trade) from municipal and prefectural gov- commodities and to decrease the domestic price rel- ernments, a subsidy that we only documented in ative to the world price of exportable goods. While south China (although it should be noted that we export rebates can still be used legally (assuming the did not have the opportunity to interview many taxes that are being rebated have been paid domesti- rice traders in north China--that is, ones that cally), some instruments will not be able to be used might be exporting japonica varieties into the in the future (such as, export subsidies). northeast Asia market). Most of the rice traders we spoke with received no subsidy for their exports, New Estimates of China's Nominal meaning the average subsidy almost certainly was Rates of Protection less than 1 percent (table 6.3). Because of the higher numbers of meat exporters In this section, we estimate a new set of NRPs on and the private or commercialized public nature the eve of China's accession to the WTO. These of the companies, subsidies are not provided; estimates will attempt to overcome some of the 86 China and the WTO previous problems faced by researchers. In particu- for a commodity. For example, it is difficult to lar, we try to understand in a more disaggregated provide one single NRP of wheat in China, one of way the part of certain markets (in terms of vari- the world's largest importers of wheat (table 6.4). eties or commodity type) that China is protecting. Traders reported that the price of very high quality Such an analysis should help us more accurately wheat from North America was 20 to 50 percent assess what the impacts will be after China imple- higher in the domestic markets of China's major ments it WTO obligations. To do so, we first explain ports than when it was on a ship in China's port how we collected our data. Next, we look at the dis- ready to be brought into the country (table 6.4). aggregated results. Finally, to make the information More precisely, the average trader told us that if a more useful to policy makers and other researchers, ton of Canadian Number 3 hard white wheat were we create a series of more aggregate NRPs. The brought in and auctioned off in October 2001, the aggregation of our disaggregated NRPs into a single competitive bid price would have been 20.5 percent crop-specific figure allows us to assess how our higher than the international price on a cost, insur- methods would have changed had we used tradi- ance, and freight (CIF) basis. The same traders told tional methods of estimating NRPs. Appendix A us that the competitive bid price of high quality summarizes some of the difficulties that practition- U.S. wheat would have been 50 percent higher. ers face when trying to measure NRP for China's Hence, based on this price gap, one would have to agriculture. assume that China's protection price is high, and if To overcome the shortcomings of previous NRP it were to open its markets completely, wheat prices studies,we conducted a set of interviews and surveys would plummet and import volume soar. with the stated goal of precisely identifying the However, traders were quick to point out that differences in prices at a precise point of time and a they did not think that even with open markets particular location between an imported good on China's wheat price would fall anywhere near one side of the border (outside China) and a domes- 50 percent even if there was no effect on the world tic good of identical quality on the other side (inside price; they were not considering the impact of China). Similarly, we also wanted to identify the China's imports on the world price.4 According to same price gap between exportable domestic goods our interviews, the market for baking-quality as they leave the country and the same goods from wheat, the main use for hard white wheat from other countries that are being traded in interna- North America, is actually relatively small in China, tional markets.Conducted in 2001,the enumeration at most only several million metric tons (MMTs). team was in the field more than three months, from We also were told that few users in China outside August to November. The team visited seven coastal those who demanded flour for making cakes, pas- cities--Guangzhou, Shenzhen, Ningbo, Shanghai, tries, and high-quality breads would use this type of Lianyungang, Qinghuangdao, and Dalian--and the wheat and that only a small group of farmers and two more inland cities of Beijing and Changchun. In processors inside China could supply this type of each port, a number of "sampling frames" was used wheat. If this is in fact the case, this would mean to select a sample of domestic traders, importers that even in a world free of any trade restrictions, and exporters, wholesalers, grain and oilseed users, imports would come into China until demand was trade regulators, agents, and other grain and fiber filled and the domestic price for that variety fell to officials. More than 100 people were interviewed;2 international levels. less than 10 percent of those contacted refused to be Alternatively, it could be the case that all pro- interviewed.3 duction of that particular variety would shift out of China if all of China's farmers abandoned them because they could not make a profit. In such an Dissagregated NRPs for Selected Agricultural extreme case, with few domestic supplies and with Commodities in China little or no substitution of the baking-quality wheat The results of our analysis clearly illustrate the for other domestic uses, the price impact on most problems with a strategy of NRP estimation that domestic producers would be small. Growers of the attempts to come up with a single rate of protection high-quality wheat would lose; they would have to The Nature of Distortions to Agricultural Incentives in China and Implications of WTO Accession 87 TABLE 6.4 Disaggregated Nominal Rates of Protection for Selected Grains in China, October 2001 Comparable Domestic Border Prices Price (US$/Ton) Variety or Quality Yuan/Ton US$/Ton C.I.F F.O.B NPR (%) Estimated at official exchange ratea Rice Weighted average 3 Thai jasmine rice 3,690 446 380 17 High-quality japonica 2,930 354 398 11 Medium-quality indica 1,519 184 185 0.5 Wheat Weighted average 12 US DNS 2,350 284 190 49 Canadian #3 1,800 218 181 20 Australian soft 1,625 196 175 12 US hard red 1,550 187 169 11 UK 1,350 163 145 12 China high quality 1,350 163 145 12 China medium quality 1,250 151 140 8 China low quality 1,100 133 133 0.1 Soybean Common variety 1,950 236 205 15 Maize Common variety 1,150 139 105 32 C.I.F. Cost, insurance, and freight F.O.B. Free on board aThe estimated official exchange rate is 8.28. Source: Authors' survey. keep growing at a lower price, switch to another China's domestic market, however, this wheat was wheat variety, or change cropping patterns. Since considered to be high-quality Chinese wheat. Inter- the quantities of such grain are so small, however, estingly, evidence that medium-quality wheat on the overall impact would be minimal. international markets is the same as high-quality While not as extreme as the case for North wheat supplied by China's farmers is found in the American baking-quality wheat, traders reported answer to the question that we asked our intervie- that arbitrage possibilities in other markets wees: if China's higher-quality wheat were sold (table 6.4). With a remarkable degree of consis- on international markets, how much loss would tency, the CIF price of medium-quality wheat a trader incur? Our survey found that this rate, imports from Australia, England, and the Pacific 10 percent, was almost exactly the same as the Northwest of the United States (hard red) was premium importers would make from bringing reported to be 10 percent lower than the price that in medium-quality grain from the international they believed the same wheat would command in market. China's domestic market. Used for more common Finally, although there have been no imports of bread, cheaper pastries, industrial uses, and high- low- or lower-medium quality wheat from inter- quality noodles, interviewees believed that this national markets, it appears that China's medium- market accounted for approximately 10 to 15 per- quality wheat, by far the biggest part of China's cent of China's wheat demand. However, unlike the production (estimated to be more than 60 per- case of the highest-quality baking wheat, produc- cent) is only marginally protected (table 6.4). Our tion was greater in China. In fact, in 2001 domestic survey found that traders believed if China's producers supplied most of the wheat of this quality medium-quality wheat was sold on the interna- into this segment of China's wheat market. In tional market in late 2001, it would sell at a discount 88 China and the WTO TABLE 6.5 Average Nominal Protection weighted by their area shares. On average, the prices Rates for Major Imports and of all varieties of domestically produced wheat sold Exports in China, October 2001 in the domestic markets of China's major port (and inland) cities are 12 percent above the average CIF Nominal Major Imports Domestic Price Rate of Protection price of all types of imported wheat varieties. Rice, and Exports (Yuan per Ton) (Percent) on the other hand, is implicitly taxed at 3 percent. The aggregate figures, although helpful (and per- Imports Wheata haps needed for analysis that is only disaggregated 1,250 12 Maize 1,150 32 to the crop level, are less interesting and provide Soybeans 1,950 15 much less insight into which groups of farmers in Cotton 9,500 17 which areas are producing which varieties will be Sugar 2,612 40 hurt or helped if trade liberalization reduces the Exports distortions. Ricea 1,954 3 However, to the extent that certain commodities Porka 11,442 30 have less intracrop quality differences, the aggregate Beefa 13,743 10 measures have more inherent interest. For example, Poultrya 9,904 17 maize, soybeans, cotton, and sugar have far less Fresh fruits 5,472 4 quality differences among varieties than rice and wheat. This is partly due to the fact that maize and aAverage Nominal Rates of Protection are created by soybeans are rarely consumed directly (as are rice summing the NRP rates of individual varieties weighting and wheat, making them more sensitive to human with the sown area (production) share. tastes and preferences).Instead,maize and soybeans Source: Authors' survey. are used mostly as feed or are processed. As a consequence, in our analysis we only examine of about 8 percent. Another way to interpret this aggregate crop NRPs for maize, soybeans, cotton, result is that if international traders can ship this and sugar.5 quality of wheat to China, it would command a Our findings show not only that significantly premium of 8 percent. It is likely that imports of positive rates of protection exist for a number of wheat will increase after WTO accession because China's major field crops, but also that they vary of the persistent price gap. The effect, however, over the nation and according to the position in appears to be less than 10 percent. China's lowest which China finds itself as a net importer or as a quality of wheat (about 10 to 15 percent of its net exporter. Maize prices, according to exporters, harvest) is at the world's feed wheat price (table were more than 30 percent, on average, above 6.5). China did export some feed wheat into world prices. In other words, they would have lost international markets in 2001 (mostly to Asia, more than 30 percent of the value of their ship- according to an interviewee). Similar differences ment if the government had not paid a subsidy. in the size of the price gap among varieties of a Protection rates for maize as an import differed. single grain are found for rice, though not for For example, traders in the northeast told our sur- wheat and maize, which are more homogeneous vey team that if they were not exporting and for- products. eign maize was to come into China, the importer could gain 21 percent since the price of imported maize is 21 percent lower than the domestic maize New NRPs for China price. Our interviews in south China, however, Although differences exist among major types found that the price gap between imported maize of any individual agricultural commodity, by (CIF) and maize being traded in the domestic weighting them by their sown area (for crops) and market in and around Guangzhou was more than production (for meats) shares, a set of by crop 30 to 40 percent. Aggregated across areas on the aggregate NPRs can be created (table 6.5). Wheat, basis of meat consumption shares, we estimate for example, has an NRP of 12 percent (row 1) that China's maize NRP was 32 percent in 2001 when the individual NRPs from table 6.4 are (table 6.5). The Nature of Distortions to Agricultural Incentives in China and Implications of WTO Accession 89 TABLE 6.6 Percentage of Market Pairs NRP of sugar remained about 40 percent through- that Test Positive for Being out the year. Integrated based on Dickey- Fuller Test in Rural China, 1988 to 2000 Assessing the New Methodology Commodity 1989­95 1996­2000 Since one of our objectives was to use a new data source and method for aggregating NRPs data to (Percent of Market Pairs) generate crop specific NRPs, it would be interesting Maize 28 89 to analyze what would have happened had we not Soybeans 28 68 used this time- and data-intensive survey method- Japonica Rice 25 60 (Yellow River ology. To conduct such an experiment we used the Valley) same methodology, data sources, and assumptions Indica Rice 25 47 that many people use for calculating NRPs to (Yangtse Valley calculate an NRP for China in 2001.6 Although the and South China) two approaches give almost the same answers for some commodities, such as soybeans and maize Note: Results for two periods from same data (though soybeans were still overstated, in part set. For results from 1989­95 for maize and rice, see Rozelle and others (2000). Rice results because of difference in prices over the entire year-- are for the whole country in 1989­95. Results China's domestic prices fell sharply over the year, from soybeans for 1989 to 1995 and all results suggesting that the NRP in late 2001 was lower than from 1996 to 2000 are by authors. it was in early 2001), the answers vary considerably for other commodities. For example, the national average price for wheat in 2001 reported from the Interviewees also reported that despite the large Ministry of Agriculture reporting system was volume of increase of soybean imports in recent 1,113 yuan per ton. The average price of imports years, a difference remains between the CIF and calculated by dividing total import value by total domestic price in the port (table 6.6). The average import quantity was 1,393 yuan per ton. In other difference between the domestic price and the words, the domestic price of wheat using these international price was 15 percent. In one sense, the sources of data about prices is 21 percent below the fact that there is a remaining price gap is remarkable CIF price of imports.From this standard methodol- given that China imported almost 15 MMT of ogy, one would come to the conclusion that wheat, soybeans in 2001, the official tariff is only 3 percent, rather than being protected (by 12 percent--see and the commodity is freely traded without secur- tables 6.4 and 6.5), was actually being taxed by trad- ing a license or quota allocation. On the other hand, ing policies.Yet, as we have seen the main reason for the remaining price gap reminds us that there may generating a negative rate of protection is that be other reasons for distortions beyond tariffs and China is importing almost exclusively very high- state trading. grade, baking-quality wheat, while its domestic Our results also demonstrate that cotton and consumers use mostly medium- and lower-quality sugar were fairly highly protected in October 2001 wheat. The wrong conclusion is reached when one (table 6.5). The case of cotton, however, is an exam- uses the specialty prices for imports as an interna- ple of how fast the NRP can changeacross time.The tional reference price for types of wheat that are NRP was measured at 17 percent in October 2001. much lower quality and are lower priced. When our team returned for follow-up work at the The same problem is found for rice. Because end of November, the domestic price of cotton had China imports only high-quality jasmine rice from fallen from 9,500 yuan per ton in October to less Thailand, the international price of rice (3,908 yuan than 8,000 yuan per ton. With this fall, the NRP per ton--that is calculated by total import value went to less than zero. However, later in the year, the divided by total import quantity) appears to be more international price of cotton also dipped, a fact that than 150 percent higher than the average domestic would lead to a higher NRP. Being less variable in price (1464 yuan per ton). In fact, as shown in 2001 in both China's and international markets, the table 6.5, China's average price protection (tax) rate, 90 China and the WTO calculated on a variety-by-variety basis, is almost the country and should not be expected to have zero ( 3). highly adverse impacts on the poor who are largely Thus, according to this illustrative example, we located in inland areas far from major urban cen- can see the necessity of approaching the estimation ters. While being isolated from negative external of NRPs in a more careful way for some commodi- shocks is a plus, there is also a cost. Those living ties. Using the traditional approaches work fairly in poor, isolated areas also would not benefit well for commodities that are relatively homoge- from price rises and opportunities to export, and nous in their quality characteristics, such as maize they are potentially vulnerable to price shocks as and soybeans. We have seen for the case of wheat regional production and consumption demand and rice for China in 2001, however, that compar- change. However, if markets exist that link together ing average prices inside and outside of the nation distant regions with the coast and price changes in can lead to misleading results. Based on this exam- one part of the economy quickly ripple through the ple, one might conjecture that traditional estimates economy, even though imports are infused into of NRP for some products, such as sugar and edible (and exports flow out of) areas concentrated oils, may be fairly reliable. Those for meat products, around a few large coastal cities, they could have cotton, and horticulture crops, however, could be ramifications for poor households thousands of misleading. kilometers away. To the extent that there are high transaction costs inside China and to the extent that certain WTO Effects Away from the Border domestic markets are isolated from others in the While important in determining the size of the country--especially those inland areas isolated shock at the border, the broader magnitude of the from port regions where imports land--the impact effect of the WTO agreement on China's farmers of WTO policies may not be evenly distributed. In depends not only on the size of the distortion, but previous work done on China's agricultural mar- also on the size of the area across which it will kets (e.g., Park and others 2002), it was found that, be felt. This second factor, in turn, is a function of in general, China's markets were becoming fairly the size and nature of China's market. In fact, there integrated by the mid-1990s. However, this con- are at least three factors: clusion should be qualified. First, although mar- kets improved greatly during the early 1990s, the · policy safeguards that limit market forces from analysis still found large parts of the country, fully equilibrating domestic and international especially poorer areas, were not completely inte- prices grated into national markets in all years. More- · household responses by which households are over, the study's dataset is dated. Since the final able to move into the production of higher year of the available data, more than seven years profitability commodities and away from those have passed. It is unclear whether markets have that experience price falls matured since that time or whether the actions · high transaction costs that possibly can serve to taken by leaders have led to greater fragmentation buffer the effects of liberalization policies on (Nyberg and Rozelle 1999). Surprisingly, given the those who live in rural areas in China. fragile nature of reforming China's agricultural markets, almost no recent work addresses these The policy safeguards are discussed elsewhere in questions. this volume. The effects of household responses are discussed in Taylor (1998) and Huang and Rozelle Assessing the Determination of Price and Market (2002). In this section, we focus on the nature of Integration in China markets. Ultimately, the distributional impacts of WTO To assess the state of the integrated and developed will depend on the nature of China's markets. If markets in rural China in the late 1990s and 2000, large areas of the country are isolated from coastal we first describe the data. Second, we test for inte- markets where imports land, then the effects of gration and conduct direct tests of how well prices WTO may be circumscribed to restricted parts of in different markets move together and if prices are The Nature of Distortions to Agricultural Incentives in China and Implications of WTO Accession 91 integrated between the market town and China's residuals, et, for the second step. The second step villages. Finally, we measure the degree of price then tests for the stationarity of the residuals from transmission. equation (6.1) using the augmented Dickey-Fuller test: Data The data come from a unique price dataset collected by China's State Market Administration N (6.2) Bureau (SMAB). Nearly 50 sample sites from 15 of et = et -1 + et -j +t j=2 China's provinces report prices of agricultural commodities every 10 days. The prices are the aver- If the test statistic on the coefficient is less (i.e., age price of transactions that day in the local rural more negative) than the relevant critical value from periodic market. The Ministry of Agriculture the Dickey-Fuller (D-F) table, the null hypothesis assembles the data in Beijing, making them avail- may be rejected and the two series are said to be able to researchers and policymakers. cointegrated of order (1,1). When the series are We examine rice, maize, and soybean prices cointegrated this way, it implies that the two mar- from 1996 to 2000 (except for maize that was only kets from which the price series come are inte- available only through 1998). The three crops are grated. The absolute value of the test statistics produced and consumed in nearly every province should be greater than 3 at 10% significant level. In in China. Rice price data are available for 31 mar- our paper, we are conducting only the unit root kets. Because of quality differences among rice vari- test, where j equals zero, since the error term from eties in different regions of China, we look at price equation (6.1) is an AR(1) process. integration between markets within four regions, The results of the cointegration analysis support South China, the Yangtse Valley, the North China both our descriptive findings and the conclusions Plain and Northeast China. For the provinces of the determinants of commodity price analysis in included in the sample, rice prices are available for the previous page, especially when they are com- over 90 percent of the time periods. Prices for pared to the findings of research on market inte- maize and soybean data are available for 13 and 20 gration in the late 1980s and early 1990s (table 6.6). markets, respectively.7 Product homogeneity makes In middle part of the reform era (1988­1995), a it possible to include a broader geographic range of time when markets were starting to emerge, buyers and sellers in a single analysis, and we are between 20 to 25 percent of markets showed signs able to assess the integration of markets spread out the prices were moving together during the study over thousands of kilometers. We compare these periods and sub-periods. According to their find- results to results from 1988 to 1995 that were pro- ings, although there were many market pairs in duced with the same data and published in Park which prices did not move together, between the and others (2002). late 1980s and mid-1990s, there was evidence of rising integration. Integration Tests In this section we use more for- Using the results from the early 1990s as a base mal tests of market integration. To do so, we apply line, our current analysis shows that during the late the Engle-Granger cointegration approach to test 1990s, China's markets continued along their pre- for the integration of China's grain markets. Two or vious path of maturation; markets in China, espe- more price series are cointegrated (even if each is cially those for maize, are remarkably integrated. individually nonstationary) if a linear combination In the late 1990s, examining the co-movement of of the variables (e.g., the differences of the prices) is prices between pairs of markets in our sample, we stationary. Following Engle and Granger (1987), we see a large increase in the number of integrated apply a two-step residual-based test. The first step markets. In the case of maize, for example, in uses the ordinary least square (OLS) regression of 89 percent of the cases, prices in one market move one price series on another: at the same time as in another (table 6.6). This is up Pt = + t + Pt + et , i j (6.1) from only 28 percent of the time in the early 1990s. The number of pairs of markets for soybeans, where t is the common trend of the two price- japonica, rice, and indica rice shows similar series and where et is the error term. The main rea- increases (table 6.6). The integration of these son for running the first step is that it provides the markets is notable because, in many cases, the pairs 92 China and the WTO of markets are separated by more than a 1000 kilo- markets are not integrated to the village level; in con- meters. For example, we find prices in almost all trast, if the local supply shock does not affect the years to be integrated between markets in Shaanxi price, villagers are price takers and markets will be and Guangdong provinces and between those in thought to be integrated. Sichuan province and southern Jiangsu. The data for this study were collected in a ran- Despite the significant progress in terms of inte- domly selected, nearly nationally representative grations, our results also show pairs of markets sample of 60 villages in six provinces of rural during different years that are not integrated. For China the China National Rural Survey (CNRS). example, in one third of the cases, japonica rice To accurately reflect varying income distributions prices moved in one market but did not in another. within each province, one county was randomly The case of indica rice trade is even more notable. selected from within each income quintile for the In the case of more than half the time (and places), province, as measured by the gross value of indus- prices do not move together in China's indica trial output. Two villages were randomly selected rice producing and consuming regions. One within each county. The survey teams used village explanation for such a result is that there is rosters and our own counts to randomly choose some kind of institution (policy or infrastructure/ twenty households, both those with their residency communication) breakdown that is creating permits (hukou) in the village and those without. China's fragmentation, as shown in Park and others A total of 1,199 households was surveyed. The (2002). It is also the case that since every province CNRS project team gathered detailed information in China has rice production and consumption, if on both the production and marketing behavior of during a certain year in a certain area, supply in all of the farmers in the sample and the character- that region is just equal to demand and price differ- istics of each village and its relationship to the entials between regions stay within the band nearest regional market. From each respondent in between regional "export" and "import" prices, the survey in each village, we identified the price moderate price movements in another area may and timing of the sale for each commodity. From not necessarily induce a flow into or out of the these data, we construct an average village price region that is in equilibrium. for each month in yuan per kilogram. In a com- Even with the nontrivial number of cases in munity questionnaire, we know how far the vil- the late 1990s in which market prices in pairs of lage's center is from the nearest paved road and the markets do not move together, based on each of distance to the county market in kilometers. the market performance analyses, one must con- Finally, for each crop that the farmer cultivated, we clude that the impacts of WTO on China's agricul- know whether the farmer's crop suffered a shock, ture will be experienced across wide regions of the recording both the incidence and the percentage nation from coastal to inland areas. However, this is by which the yield fell. We do not include any vari- only half of the story. The discussion of table 6.6 able that controls for the presence of a community demonstrates a remarkable degree of integration buffer stock system, primarily because such an between markets on the coast and those inland, institution is almost never observed in modern such an analysis is still not sufficient to ensure China.8 In addition, sales among farmers within a that many of China's villages will be affected by village are rare (according to our data, less than the shocks that hit the coast and are transmitted 5 percent of sales). County dummies hold quality inland. constant. To do so, in this part of this section we examine Our regression analysis clearly shows markets the extent to which villages are integrated into in China are integrated down to the village level regional markets. Our test of integration will essen- (table 6.7). The signs and level of significance of the tially identify if farmers in China's villages are price coefficients on variables, such as the distance that a takers or are villages isolated, making prices deter- village is from the market, demonstrate that the minedbylocalsupplyanddemand.Inbriefestterms, farther a village is from a market, the lower the if variables that affect local supply significantly price the farmer receives, which is the expected affect prices, we will assume villages are isolated and result. More importantly for our purposes, the The Nature of Distortions to Agricultural Incentives in China and Implications of WTO Accession 93 TABLE 6.7 Soybean, Corn, and Wheat analysis, we stack the price data from various mar- Village Price Regression, 2000 kets for the last three years (1998­2000) and regress the price of the inland market (in logs in time Explanatory Soybean Corn Wheat Variable Price Price Price period t) on the price at the port market (in logs in time period t) and three lags (t-1; t-2; and t-3-- Distance to the 0.029 0.00064 0.0095 table 6.8). The sum of the coefficients on the port nearest country market variables provide an intuitive measure for the total (2.37)** ( 1.63)* (3.24)** transmission of price shocks in percentage terms. The price transmission coefficients range from 42 Village-level 0.04 0.12 0.081 shock to to 51 percent for rice (the lower range being a production measure of the price impact that does not include ( 0.17) ( 1.34) ( 1.02) coefficients with t-ratios under 1.580 (table 6.8). Other variables timing of sales/net purchase According to these measures, about half of the price not shown or seller change at the port is transmitted to the inland mar- ket. The transmission coefficients for maize range from 51 to 57 percent (table 6.8). This means that if t-ratios of the coefficients of the village supply shock variables are all less in absolute value terms TABLE 6.8 Transmission Coefficients for than 1.35, signifying that the output of the local Rice and Maize Measuring the village's crops do not affect the local price. One Percentage of Price Shock at implication of this result is that factors outside the Port that Is Transmitted to village affect the price that farmers receive, making Inland Markets in China, them price takers. In other words, farmers, even in 1996­2000a China's remote villages, are linked to the markets of Maize Rice its main commodities. Standard Vector 51­57 42­51 Autogression Model (VAR)b Price Transmission Coefficients VAR with corrections 49 10­13 While integrated markets mean that inland markets for autocorrelation. will experience price changes in the direction of the Vector Autogression 20­35 12­25 price movements at the ports (for those crops that Model with Impulse are imported and exported), frictions in the mar- Response Simulationd keting system may shield inland producers from some of the effects.9 Moreover, despite the rise of aThe transmission coefficient is interpreted at price integration in China's domestic markets dur- the average proportion of a price shock in the port market that is experienced by the markets ing the reform period, there are still a significant inland. number of market pairs during certain years that bStandard Vector Autoregression Analysis stacks do not move together. Hence, when assessing the the price data from various markets for the last impact of WTO-induced price shocks at the border three years (1998 to 2000) and regresses the price of the inland market (in logs in time on farmers inland, we need to examine the degree period t) on the price at the port market (in of transmission of these effects. logs in time period t) and three lags (t-1; t-2; To examine the proportion of the price changes and t-3). The sum of the coefficients on the port variables is a measure of the total transmis- that would be experienced in regions away from the sion of price shocks in terms of proportions. port, we conducted a series of analyses to try to cThis uses the same model as in row 1, but also measure the extent of the change of prices inland includes lags of the dependent variable for t-1, for a percentage change at the port. We do the t-2 and t-3. Simulation analysis proceeds by shocking the price at the border and following analysis for the two major crops for which we have the price from port to inland, holding the own complete data series: rice and maize. In the first market's price generation constant. 94 China and the WTO maize prices fell 10 percent at the border due to on agricultural prices. Moreover while trade will the import of an amount up to China's quota also affect other subsectors of the rural economy, limits, the price of maize inland would fall around this study's focus on the agricultural sector showed 5 percent. that there will be an impact. However, like other The analysis in the above paragraph, however, effects, those in the agricultural sector may not all does not account for the fact that the error term in be negative. Our findings, based on new methods to the equation could be subject to autocorrelation. collect data and create NRPs, show that indeed for The corrected transmission coefficients fall only some crops WTO accession will likely lead to a fall modestly for maize, declining to 49 percent in prices and a rise in imports. Maize and cotton (table 6.8). In other words, even after accounting may be most affected. It is possible that soybeans for autocorrelation, if the price at the border (or and sugar could be significantly affected in the port) changes by 10 percent, the inland price falls longer run. There are also commodities in which by nearly 5 percent. After the same treatment in the China has considerable comparative advantage-- rice equations, however, the price transmission for example, rice, meats, and horticultural products coefficient falls sharply (to around only 10 percent and, hence, WTO accession could provide benefits (table 6.8)). According to this result, we find that to those engaged in these activities. The prospect of rice markets are subject to much more friction increased imports of feed grains (e.g., maize and than those for maize, a result that may result from soybeans) at lower prices means that livestock pro- the inherent differences between rice (a commod- ducers could become even more competitive. ity with a wide range of qualities) and maize (a The extent to which prices fall from rising more homogeneous commodity). In other words, imports or rise from rising exports in part depends when we are observing a price shock in the port on how China executes its WTO obligations. market, it may be caused by the new inflow or Although there may be room for footdragging shortage of a particular type of rice. For example, (which could delay that negative effects), the nature when a certain type of rice in Guangzhou suffers a of the agreement also provides many means to limit shock (e.g., the harvest in one of the production the downside effects. Likewise, China's benefits are bases is greatly reduced, although the price in going to depend on how well trading partners another production base that produces that variety honor their commitment and provide better access may move, the price in an area that does not pro- to global markets. We suggest that rather than foot- duce that variety may not change. In the case of dragging on its own reforms, China would be better maize, however, its homogeneous nature means to use some of its capital and goodwill to fight that most of the price shock in one market is measures in its trading partners, such as Japan's passed through by traders. safeguards against mushrooms. Here, China has A third analysis confirms the finding of the more already had a huge, unheralded win by getting traditional price transmission model (table 6.8). Japan to move from its original way of conducting Using a impulse response analysis (which basically bilateral trade with China (based on blatantly uses the price transmission model and adds a set of WTO­illegal measures to safeguards that are trans- lagged own prices the price in the inland market in parent and temporary. China also should realize period t-1, t-2 and t-3), we find that a 10 percent that, as a WTO member, it now has a means to file a price change in the port price of maize changes the protest if it believes it is being treated unfairly. inland price by up to 35 percent and up to 28 per- When countries use the WTO enforcement organi- cent in the case of rice. zations to fight unauthorized trade barriers, com- plainants tend to have a high success rate. In gen- eral, to gain the most in the long run from this Conclusions agricultural agreement, both China and its partners The purpose of our study is to examine the effects need to endeavor to live up to their agreements. of China's accession to the WTO on the agricultural We also found that unlike the case of Mexico, it sector. Although there are many possible effects, we appears as if most of China's markets may be well- have focused on the impact of trade policy changes integrated into the economy. This is good news and The Nature of Distortions to Agricultural Incentives in China and Implications of WTO Accession 95 bad news for poor farmers. The good news is that NPR of a commodity is simply the difference of they can benefit from falling input prices and rising these two numbers. export opportunities. The bad news is that unlike a However, more careful observation shows the large number of maize farmers in Mexico who were search for a single number may be one of the main not affected by the North American Free Trade reasons different analysts can come to so many dif- Agreement reduction in maize import restriction, if ferent conclusions. In fact, there are many prices for our results are correct for large parts of China, its wheat in China. Prices vary across time within a farmers will be affected. The problem, although it is year. Prices vary across regions within a time a short-run one, may be that it is this group of rural period. When calculating the NPR, does one look at households that are most dependent on agriculture the price of corn in a Guangzhou feedlot or the and least able to be flexible. As a consequence, our price of corn sitting in storage in a farmer's home- findings should be taken as a warning to govern- made silo in Northeast China? Moreover, rice is not ment leaders that they need to be concerned about rice is not rice. There are many different varieties the welfare of these susceptible groups. and types, all of which command different prices at Our results also generate findings that show the different places at different times during the year. In close relationship that exists among the degree of fact, for some commodities, such as rice, China is integration of an economy, the size of the price exporting one type at the same time it is importing effect, and the amount that fraction of the TRQ another. The same sets of issues face analysts when that will be imported. If China's markets are really they attempt to choose a price series (or more diffi- so integrated, and leaders do not artificially delay cult yet, the single price) to represent the interna- the ability of traders to execute TRQs, our findings tional price. Which price should an analyst choose? suggest that the price effects may not be too large Should it be FOB or CIF? Should it be the average because they will be spread out across a large area annual price or a price during one particular of the country. However, if the price effects for a period? If there are many different types of given quantity of imports are not large, the vol- imported varieties, which type should be chosen? ume of imports may be larger than predicted by In part because previous studies have not dealt some and the bindings may be more likely to take with these issues (at least explicitly), it is unsurpris- effect. We do not expect in any circumstance that ing that different research efforts have generated imports will ever exceed the limits put on by the different estimates of NPRs. For example, Tuan and TRQs. Cheng (1999) estimated quite high and variable nominal rates of protection for agricultural com- modities. Their estimates for wheat, maize, and Appendix: Challenges and Issues soybeans in 1997 were 62, 15, and 140 percent in Measuring Nominal Protection respectively. On the other hand, Carter and Estrin Rates (2001) find generally negative price distortions. Although measuring the difference in price Huang (2001) provides sets of estimates that show between an economy's domestic price and the some products are highly protected and in other international price, the wide range of estimates of cases there are negative rates of protection. NPRs that exist for China demonstrate it is not a straightforward process. In fact, a number of issues complicate NPR measurement.First,confusion may Notes stem from the way analysts have asked their ques- tion about NPRs. Policy makers and researchers 1. Some scholars in China have also pointed out that since have sought to summarize the impact of various part of the value of agricultural commodity production uses inputs on which the value added tax has been assessed, the"real" commodities with a single number. Trade modelers tax rate on agricultural commodities is actually higher.Although need a single number to make their analytical certainly this is the case, the maximum that could be added frameworks tractable. People want to know what is would only be an additional 2 to 4 percentage points (15 percent times the share of the inputs that were taxed--about 10 to the price of wheat in China and compare that to the 30 percent--depending on the commodity, the technology, and world price of wheat. With this information, the region of production). 96 China and the WTO 2. Because of the absence of a single central authority that 8. We do, however, in an alternative specification (not shown manages grain flows, the enumeration team chose its sample in a here) include a variable that controls for the average amount of number of ways. In each location, we first visited the local grain private grain storage in the village. It is not significant and its bureau and obtained access to a list of all grain bureaus, the firms inclusion does not affect the results. that they were running on a commercial basis, and their sub- 9. It is also possible that there could be transmission elastici- sidiaries. We interviewed an official in the grain marketing divi- ties greater than unity. Producers in an exporting region may sion and transportation division. We also chose three firms that find their price moves proportionately more than in the port. were owned directly by the grain bureau and three that were For example, consider the case if the CIF price was 100 and the affiliated with the grain bureau. In several cities, the grain bureau interior price was 50. If the marketing margin were additive, the had a list of large grain trading and grain using firms (e.g., mills rise in the world price of 10 percent. and feed lots). In others, this was obtained from the market administration bureau. Five firms were chosen on the basis that they were private and had yearly sales that exceeded one million yuan. We interviewed at least two flour or rice mills and feed References mills in each location. Finally, we visited the wholesale market and randomly chose five stalls to interview. The team also visited Anderson, K., and C. Y. Peng. 1998. "Feeding and Fueling China a number of other entities, such as the grain reserve, the local in the 21st Century." World Development 26(8): 1413­29. COFCO agency, and supermarket chains. In some cases, the CARD. 2001."Effect of WTO on Agriculture."Paper Presented at managers of these entities knew the grain trade business well the 2001 annual meetings of the American Agricultural Eco- enough to answer our questions; in other cases they did not. nomics Association, August 5­8. Chicago, IL. 3. During the interviews, a survey instrument was filled out Carter, C. A., and A. Estrin. 2001."China's Trade Integration and documenting the scope of the interviewee's participation in Impacts on Factor Markets." Mimeo. Department of Agri- China's domestic and international food and fiber trade. We cultural and Resource Economics, University of California, were particularly concerned with understanding the transac- Davis. tions that the interviewees were involved with or knew about Chen, Shaohua, and Martin Ravallion. 2004."Welfare Impacts of that concerned imported or exported grains, fiber, meat and China's Accession to the WTO." In D. Bhattasali, S. Li, and W. other goods. The survey recorded the characteristics of the com- Martin, eds., China and the WTO: Accession, Policy Reform, modities that were involved in trade in the immediate marketing and Poverty Reduction Strategies. Washington, D.C.: World area during the fall of 2001. Enumerators then asked the inter- Bank and Oxford University Press. viewee a series of questions about commodities about which the Huang, Jikun. 2001. "Agricultural Policy and Food Security." traders were most familiar. For imported commodities, intervie- Working paper, Center for Chinese Agricultural Policy, wees first told the enumerators the international CIF price of the Chinese Academy of Sciences. Beijing. good. Second, the interviewee then told enumerators what the Huang, Jikun, and C. Chen. 1999. "Effects of Trade Liberaliza- good would sell for if auctioned of in a competitive auction. In tion on Agriculture in China: Institutional and Structural other words, we elicited a series of price gaps for a carefully Aspects." Working Paper 42, United Nations ESCAP CGPRT defined set of goods. Since, on average, each interviewee had Centre, Bogor, Indonesia. information about a number of commodities, we had several Huang, Jikun, and Scott Rozelle. 2002. "Market Distortions and hundred observations. A similar set of questions was asked the Impact of WTO on China's Rural Economy." Working about exportable goods, including maize, rice, cotton, and meat Paper, University of California, Davis, CA. products. Li, S., F. Zhai, and Z. Wang. 1999. "The Global and Domestic 4. The assertion that consumption will be limited is based on Impact of China Joining the World Trade Organization, A the assumption that tastes will not change. It could be, however, Project Report." Development Research Center, the State that when the changes in consumption due to rising incomes are Council, China. added to those due to changing tastes as higher quality products Martin, W. 2002. "Implication of Reform and WTO Accession become cheaper, the scope for much greater penetration of the for China's Agricultural Policies." Economies in Transition domestic market clearly seems possible. After years or centuries (forthcoming). of consuming certain foods, the Chinese might find as others Ministry of Agriculture. 2001. China Statistical Yearbook. have that better quality bread or noodles are desirable options, Beijing: China Agricultural Press. in some cases offering nutritional advantages as well. Nyberg, A., and S. Rozelle. 1999. Accelerating China's Rural 5. We should stress, however, our survey was conducted the Transformation. World Bank, Washington, D.C. same way. In most cases, interviewees told us that there were not OECD. 2001. "Implications of Trade and Investment Liberaliza- a lot of quality differences among maize varieties. Moreover, tion for China's Agriculture." Working Paper, Organization there was only a slight (2 to 3 percent) of price difference for Economic Cooperation and Development, Paris. between imported and domestic soybeans from quality. Park, A., H. Jin, Scott Rozelle, and Jikun Huang. 2002. "Market 6. These are computed by comparing the domestic wholesale Emergence and Transition: Transition Costs, Arbitrage, and price with the average implicit price of trade. For the importable Autarky in China's Grain Market," American Journal of Agri- (exportable) it is total value of import (export) divided by total cultural Economics. Forthcoming. volume of import (export). Rozelle, Scott. 2003. "After Accession to the WTO: Corn Trade 7. Since we use data over time, we need to convert prices to a Within China and Between China and the Rest of the real basis. Nominal prices from our data set are deflated using World." A Report written for the U.S. Grains Council, monthly consumer price indices calculated and reported by Beijing, China. the China National Statistical Bureau. Deflation facilitates Rozelle, Scott, Albert Park, Jikun Huang, and Hehui Jin. 2000. transaction cost comparisons across time and allows us to "Bureaucrat to Entrepreneur: The Changing Role of the disregard transaction cost increases within periods associated State in China's Transitional Commodity Economy," Eco- with inflation. nomic Development and Cultural Change 48, 2: 227­52. The Nature of Distortions to Agricultural Incentives in China and Implications of WTO Accession 97 Taylor, J. Edward. 1998."Trade Liberalization and the Impact on Center for Chinese Agricultural Policy, Chinese Academy of Small Holders in Rural Mexico," Working Paper, Depart- Sciences, Beijing. ment of Agricultural and Resource Economics, University of World Bank. 2000. "An Assessment of China's Poverty Policy." California, Davis. Working Paper, World Bank, Washington, D.C. Tuan F.C., and Cheng, G. 1999. "A Review of China's Agricul- Zhao, Yaohui, and Terry Sicular. 2002. "Affects of WTO Acces- tural Trade Policy," paper prepared for the IATRC Summer sion on China: Labor Market Responses." Working Paper, Meeting. June 25­26. San Francisco. Center for Chinese Economic Reform, Peking University, Xie, Yuping. 2002. "WTO and China's Agricultural Trade and Beijing. Impact on Domestic Markets," Unpublished Masters Thesis. Part II Economic impacts of Accession 7 The Impacts of WTO Accession on Chinese Agriculture and Rur al Poverty Kym Anderson, Jikun Huang, and Elena Ianchovichina After 15 years of negotiations, China acceded to the in rural poverty (reported in Kanbur and Zhang World Trade Organization (WTO) at the end of 2001). So high is the level of concern that at its first 2001. During those negotiations, China was contin- WTO ministerial meeting in Cancún in September ually opening up and reforming its economy, and 2003, China joined a new group of more than 20 further reforms will be introduced over the next developing countries that demanded that member few years to fulfill its legal obligations under its countries of the Organisation for Economic Co- WTO Protocol of Accession. Such reform necessar- operation and Development (OECD) reform their ily involves structural adjustments by households, agriculture sectors without requiring developing firms, and bureaucracies, and although the econ- countries to lower their farm trade barriers. omy as a whole can gain substantially from those The policy changes that China still must make to adjustments, losses and even hardship can result for fulfill its WTO obligations include major changes some households unless complementary domestic in its farm trade policies by 2005--that is, cuts in policies are in place to facilitate adjustment or protection that appear to be far greater, and faster, compensate losers. This situation underscores the than those imposed on any other developing coun- importance of first analyzing the likely distribu- try in the Uruguay Round Agreement on Agricul- tional consequences of the reforms themselves and ture. With one-quarter of rural households in then considering what complementary policies are China living on less than $1 a day in 1999 (versus needed to provide adequate safety nets for potential 1 percent of urban households) and more than losers. three-quarters of all poor Chinese people living in Of particular concern in China's case is that the rural areas, concerns about the impact of WTO reforms may cause farm incomes to fall, exacerbat- accession on rural poverty are understandable. Yet ing the rise since the mid-1980s in farm­nonfarm reduced protectionism also may boost output and and inland­coastal aggregate income inequality and exports of some labor-intensive farm products in possibly reversing the decline since the late 1970s which China still has a comparative advantage.1 In Our thanks are extended to the World Bank and Australian Research Council for their financial support and to Will Martin, Tom Hertel, Scott Rozelle, and three referees for helpful comments on earlier versions of this chapter. The usual disclaimer applies. 101 102 China and the WTO addition, farm households will be affected indi- types of farm households. The chapter concludes by rectly by many of the other commitments China drawing out implications for Chinese policymakers has made in its WTO Accession Protocol. Especially wishing to preempt any increases in food insecurity important will be the arrangements for phasing out or rural poverty. the "voluntary" export restraints on China's textile and clothing trade and the reductions in protection The Setting of the motor vehicle and parts industry. Those changes, together with the promised increase in To put prospective changes in perspective, it is a wide range of agricultural imports, will allow helpful to review rural developments over the past China to exploit more fully its strong comparative 25 years and then the policies that have affected advantage in unskilled labor-intensive products-- those changes. both farm and nonfarm. An assessment of the impacts on farmer incomes Rural Developments since the Late 1970s and on rural areas of the remaining reforms required to meet China's commitments to WTO China's decision in December 1978 to open up its membership is best conducted by seeing those economy acted as a major stimulus to economic impacts in the context of ongoing economic growth growth: the pre-reform rate of growth in the per and structural change. This chapter therefore capita gross domestic product (GDP) of 3.1 percent begins with a brief summary of rural developments a year more than doubled, and it has remained since the initial reforms began in the late 1970s, of above 7 percent for the past two decades (see last recent policies affecting rural households, and row of table 7.1). Rapid economic growth is nor- of the pertinent reforms still to be delivered as part mally accompanied by a relative decline in the farm of China's WTO commitments. With that back- sector, but in China that decline was initially tem- ground, the chapter then provides some indication pered by introducing the farm household responsi- of the likely effects of the reforms to be imple- bility system (which led to replacing collective mented between 2002 and 2007 on the welfare of farms with individually managed holdings) and by different factor markets and thus different types of raising the prices received by farmers. So began the farm households. Because even the direction, let process of moving away from taxing agriculture alone the magnitude, of some of the effects cannot relative to other sectors--a process followed by be discerned from theory (McCulloch, Winters, and most of the advanced economies in the early stages Cirera 2001), we use the numerical simulation of their industrialization (Anderson and others model known as GTAP (named after Purdue Uni- 1986; Lindert 1991). versity's Global Trade Analysis Project) to address Agriculture grew nearly as rapidly as industry these issues. Our experimental design, which is from 1979 to 1984. However, the one-off efficiency detailed in the companion paper by Ianchovichina effects of moving to the household responsibil- and Martin (2004), goes beyond previous studies ity system and raising relative prices for farm prod- (see the survey by Gilbert and Wahl 2002) in several ucts were mostly reaped by the mid-1980s, after respects: (a) it focuses just on the liberalization which agriculture grew at only one-third the pace required to meet China's WTO accession commit- of industry and less than half that of the services ments from 2002 (by projecting the GTAP database sector as industrialization boomed on the eastern to 2001 with China's prior reforms and all coun- seaboard (table 7.1). Meanwhile, the employment, tries' Uruguay Round commitments implemented); output, and exports of rural township and village (b) it incorporates new estimates of nominal rates enterprises also boomed (see the first three of agricultural protection as of 2001 and at the end columns of table 7.2). Despite that migration of of accession; (c) it takes account of existing duty farm workers to rural industrial and service activi- drawbacks so that it avoids overstating the extent of ties (not to mention the temporary unauthorized reform; (d) it binds the government budget by migration to urban jobs such as those in construc- adding a uniform consumption tax to offset the loss tion), the average area per farm worker has fallen of tariff revenue; (d) it takes account of differences steadily since the late 1970s, while the share of farm between farm and nonfarm unskilled labor; and household income from nonfarm sources has risen (e) it examines the real income effects on different (see last two columns of table 7.2). Table 7.3 shows The Impacts of WTO Accession on Chinese Agriculture and Rural Poverty 103 TABLE 7.1 Growth Rates of China's Economy, 1970­2000 (percent per year) Reform Period Pre-Reform 1970­78 1979­84 1985­95 1996­2000 Gross domestic product a 4.9 8.8 9.7 7.9 Agriculture 2.7 7.1 4.0 3.4 Industry 6.8 8.2 12.8 9.6 Services n.a. 11.6 9.7 8.3 Food production volume Grain 2.8 4.7 1.7 0.03 Oilseed crops 2.1 14.9 4.4 5.6 Fruit and vegetables 6.6 7.2 12.7 8.6 Red meat 4.4 9.1 8.8 6.5 Fish 5.0 7.9 13.7 10.2 Value of output of nonfarm n.a. 12.3 24.1 14.0 rural enterprises Population 1.80 1.15 1.43 0.9 Per capita GDP 3.1 7.7 8.3 6.9 n.a. Not applicable. aFigures for GDP in 1970­78 are the growth rate of national income in real terms. Growth rates are computed using the regression method. Growth rates of individual and groups of commodities are based on volume of production data; sectoral growth rates refer to value added in real terms. Sources: National Bureau of Statistics of China, various issues; Ministry of Agriculture, Agricultural Yearbook of China, various issues. TABLE 7.2 Farm and Rural Enterprise (RE) Developments in China, 1980­99 RE's Share Share of Farm of Total RE's Share RE's Share of Household Income Rural Labor of Total GDP Total Exports Farmland Size from Nonfarm (%) (%) (%) (Hectares/Farm) Sources (%) 1980 9 4 0 0.56 17 1985 19 9 15 0.51 25 1990 23 14 43 0.43 26 1995 34 25 48 0.41 37 1999 35 30 48 0.40 47 Sources: National Bureau of Statistics of China, a, b; various issues. the slowdown in the decline in agriculture's shares less than 20 percent in the late 1970s to 30 percent of GDP and employment in the 1980s and its sub- by the late 1990s (table 7.3). Within the crop sub- sequent acceleration in the 1990s. sector, fruit and vegetable production grew two to Income growth has boosted the demand for three times as fast as grain output (table 7.1). The foods that are high in protein and nutrients relative prices and marketing of grain and oilseed products to those high in carbohydrates, which in turn has have continued to be highly regulated, whereas the stimulated major structural changes within agri- markets for horticultural, livestock, and fish prod- culture as farmers respond to the changes in ucts have been greatly liberalized. This liberaliza- domestic demand. For example, livestock and fish tion has accentuated the growth in the output of increased their share of agricultural output from the latter group relative to grain and oilseed output 104 China and the WTO TABLE 7.3 Changing Structure of China's Economy, 1970­2000 (percent, based on current prices) 1970 1980 1985 1990 1995 2000 Share of GDP Agriculture 40 30 28 27 20 16 Industry 46 49 43 42 49 51 Services 13 21 29 31 31 33 Share of employment Agriculture 81 69 62 60 52 49 Industry 10 18 21 21 23 23 Services 9 13 17 19 25 28 Share of agricultural output Crops 82 76 69 65 58 56 Livestock 14 18 22 26 30 30 Fish 2 2 3 5 8 10 Forestry 2 4 5 4 3 4 Share of population 83 81 76 72 71 64 that is rural Sources: National Bureau of Statistics of China, a, b; various issues. TABLE 7.4 China's Grain Production, Trade, and Consumption, 1980­2000 (million metric tons) 1980­89 1990­94 1996­2000 Production 332 396 389 Net imports 8 1 3 Change in stocks 1 11 12 Consumption 339 384 374 Food--urban 35 42 41 Food--rural 177 190 177 Feed 64 86 91 Other (seed, industrial 63 66 65 use, waste) Source: Huang 2001, based on China National Statistics Bureau publications and the Center for Chinese Agricultural Policy's database. since the mid-1980s (table 7.1). Meanwhile, the almost completely by rising domestic production, so direct consumption of grain by both rural and that the trend level of grain self-sufficiency has urban households has virtually ceased growing remained close to 100 percent. Table 7.5 shows that (table 7.4)--a consequence not only of incomes there are nevertheless considerable changes from rising but also of population growth slowing to less year to year in grain exports and imports. It also than 1 percent a year and of cuts in the implicit shows that, overall, China has remained a net consumption subsidy for food grains. exporter of food and feed, with meat, fish, and fruits The use of grain for animal feed continues and vegetables providing most of the growth in net to grow. To date, that increase has been supplied export earnings.2 The Impacts of WTO Accession on Chinese Agriculture and Rural Poverty 105 TABLE 7.5 Structure of China's Food and Feed Trade, 1980­99 (millions of US$) SITC 1980 1985 1990 1995 1999 Exports Live animals 00 384 304 430 473 374 Meat 01 361 448 791 1,349 1,054 Dairy products 02 71 57 55 61 71 Fish 03 380 283 1,370 2,875 2,969 Grains 04 423 1,065 614 281 1,273 Fruits and 05 746 825 1,759 3,399 3,150 vegetables Sugar 06 221 79 317 321 214 Coffee and tea 07 328 435 534 523 561 Animal feeds 08 58 241 623 351 239 Other foods 09 49 66 107 290 541 Oilseeds 22 -- -- -- 522 373 Vegetable oils 4 -- -- -- 454 132 Total food 3,021 3,803 6,600 10,899 10,951 Imports Live animals 00 5 18 14 18 22 Meat 01 1 6 54 97 503 Dairy products 02 5 31 81 60 160 Fish 03 13 44 102 609 890 Grains 04 2,458 982 2,353 3,631 574 Fruits and 05 48 52 83 185 384 vegetables Sugar 06 316 274 390 935 183 Coffee and tea 07 56 40 30 74 72 Animal feeds 08 14 83 182 423 620 Other foods 09 2 23 46 92 182 Oilseeds 22 -- -- -- 110 1,531 Vegetable oils 4 -- -- -- 2,596 1,352 Total food 2,918 1,553 3,335 8,828 6,474 Net exports Live animals 00 379 286 416 455 352 Meat 01 360 442 737 1,252 551 Dairy products 02 66 26 26 1 89 Fish 03 367 239 1,268 2,266 2,079 Grains 04 2,035 83 1,939 3,350 663 Fruits and 05 698 773 1,676 3,214 2,766 vegetables Sugar 06 95 195 73 614 31 Coffee and tea 07 272 395 504 449 489 Animal feeds 08 44 158 441 72 381 Other foods 09 47 43 61 198 359 Oilseeds 22 -- -- -- 412 1,158 Vegetable oils 4 -- -- -- 2,142 1,220 Total food 103 2,250 3,265 2,071 4,477 -- Not available. SITC = Standard International Trade Classification. Source: Mathews 2001, based on UN COMTRADE statistics. 106 China and the WTO Recent and Prospective Policies Affecting country's border)? Table 7.6 shows new estimates by Rural Areas Huang, Rozelle, and Chang (2004) of nominal pro- As in most developing countries,3 agriculture in tection rates for key agricultural commodities for some recent years.It suggests that rice,meat,fish,and China was squeezed at the early stages of industrial- fruits and vegetables have been priced at less than ization when gross fiscal contributions to the sector border prices, while other grains, oilseeds, sugar, were more than outweighed by implicit taxation in cotton, and milk have been priced at 20­40 per- the form of depressed prices for farm products, neg- cent above border prices. Maize and cotton also lect of public infrastructure in rural areas relative to enjoyed export subsidies in 2001, amounting to one- that in urban areas, and capital outflows via the third and one-tenth of f.o.b. prices, respectively. financial system (Huang and Ma 1998). Then, agri- Table 7.6 also shows China's commitments in its culture experienced the introduction of price and WTO Protocol of Accession: the tariff-rate quota other market reforms associated with China's policy will apply to grains, sugar, and cotton whose out-of- shift from a socialist to a market-oriented economy, quota tariffs are quite high, but, otherwise, after the starting with nonstrategic commodities such as veg- phase-in period the tariffs range between just 1 and etables, fruits, fish, livestock, and oil and sugar crops. 15 percent--representing substantial liberalizations The early reforms were aimed at raising farm-level over 2001 levels. As well, producers of major crops prices and gradually deregulating the market. As the may continue to be affected by commodity-specific right to private trading was extended to include sur- policies of government procurement of a portion of plus output of all categories of agricultural products the crop at lower-than-market prices as in the past after contractual obligations to the state were ful- (see Sicular 1988) or at higher-than-market prices, filled, the foundations of the state marketing system as in 1998 (see Huang 1998). began to be undermined (Rozelle and others 1997). What will those reforms mean for agricultural Despite periodic stop-go cycles in the reform trade? Many analysts expected China to become process, the proportion of retail commodities sold ever more dependent on agricultural imports in the at market prices has continued to rise. According course of the economy's rapid industrialization to Lardy (2002), the share for agriculture was just over the past two-plus decades. Some extremists 6 percent in 1978, but had risen to 40 percent by (e.g., Brown 1995) even suggested that China could 1985, 79 percent by 1995, and 83 percent by 1999. seriously deprive other developing countries of What have these policies meant for the nominal food. Yet, as reported earlier in this chapter, growth rates of agricultural protection in China (the per- in net food imports has not yet happened, at least centage by which domestic prices exceed prices at the TABLE 7.6 Nominal Rates of Protection (NRP, Applied Tariff or Tariff Equivalent), Agricultural Products: China, 1995­2007 (percent) 2007 Assumed NRP 2007 Out-of-Quota in 2007 1995 1997 2001 In-Quota Tariff Tariff (Core Scenario) Rice 5 5 3 1 65 3 Wheat 25 17 12 1 65 12 Coarse grains 20 28 32 1 65 32 Fruits and vegetables 10 8 4 11 11 4 Oilseeds 30 28 20 3 3 3 Sugar 44 42 40 15 50 20 Cotton 20 17 17 1 40 20 Meats 20 19 15 12 12 15 Milk 30 30 30 11 11 11 Source: Based on research subsequently reported in summary form in Huang, Rozelle, and Chang 2004. The Impacts of WTO Accession on Chinese Agriculture and Rural Poverty 107 not in a sustained way, and China has continued to Over the 1990s the average scheduled tariff rates be a net exporter of meat, fish, and fruits and veg- for manufacturing initially exceeded those for agri- etables. Indeed, on occasion in the latter 1990s culture, but later fell more than those for agricul- China also was a net exporter of grain and cotton. ture. By 2010 the manufacturing average will be How much of that situation was attributable to well below that for agriculture (a simple average of government policies that constrained domestic 9 percent versus 16 percent for agriculture--see demand, including occasional export subsidies, is a WTO 2003: Appendix Table IIB). However, these moot point. numbers do not give a true indication of the extent In its WTO Protocol of Accession (WTO 2001), of change in protection that is taking place, because China agreed to have no agricultural export subsi- in the 1990s many manufactures were entering dies and to limit its domestic support for farmers to China at reduced or zero tariffs via duty drawbacks 8.5 percent of the value of production (compared to encourage foreign investment in the process- with 10 percent for other developing countries). ing of imported intermediate goods for subse- The import market access commitments China has quent export. Some agricultural products also have made to WTO members look substantial on paper. entered at less than the scheduled rate, including Tariff-rate quotas (TRQs) will be retained only on through smuggling. wheat, rice, maize, edible oils, sugar, cotton, and What all this information means for incentives wool--the domestic production of which in aggre- for each industry is difficult to discern precisely, but gate makes up only about one-sixth of China's agri- it is better information than has been available to cultural GDP. The quota volumes are to grow over date for analyzing empirically the economic effects 2003­05 at annual rates ranging from 5 to 19 per- of the reforms associated with China's WTO acces- cent. China has also agreed to weaken the monopo- sion, including the impacts on factor rewards and lies previously held by state trading enterprises, prices from which inferences about income distrib- except for tobacco. Even though some state trading utional effects can be made. We make these infer- enterprises will continue to operate, there will be ences bearing in mind the marked differences in some competition from private firms in the per capita incomes between eastern, central, and importing and exporting of farm products, at least western provinces, and between urban and rural within the tariff-rate quotas. areas (as shown in tables 7.7 and 7.8). Farmers and the rural sector more broadly will also be affected by China's commitment to pro- Applying the GTAP Model vide improved and WTO-bound market access for industrial products. Mineral and manufacturing Version 5 of the computable general equilibrium tariffs will be bound and generally reduced on a model of the global economy known as GTAP is broad basis,with many tariffs falling to 10 percent or used here.5 As an economy-wide model, GTAP less. Tariffs were to be cut on accession, and further describes both the vertical and horizontal linkages cuts will be phased in by 2005 (with just a few excep- between all product markets within the model's tions). Furthermore, for industrial products China individual countries and regions as well as between will reduce significantly its nontariff measures and countries and regions via their bilateral trade flows. will eliminate all quotas, tendering, and import For present purposes, the 1997 database is aggre- licensing on nonfarm merchandise by no later than gated to 25 sectors and 20 regions and projected 2005. Quotas on Chinese imports of automobiles forward first to 2001 and then to 2007, using and parts will grow by 15 percent annually, from World Bank projections of population, income, a level of about US$6 billion4 in 2000, and these and endowments of productive factors (agricul- quotas will be eliminated by 2005. For textiles and tural land, other natural resources, unskilled labor, clothing, however, the current "voluntary" export skilled labor, and other capital). The initial base restraints will be phased out by the end of 2004, case assumes that China retains its protection poli- although importing countries will be able to reim- cies as of 1995 and that Taiwan (China) retains its pose quotas under a special textile safeguard mecha- protection as of 1997, but that all other countries nism until the end of 2008. China also has made fully implement their Uruguay Round obligations commitments to open up its services markets. on schedule before 2005. China's trade policy 108 China and the WTO TABLE 7.7 Rural Income, Employment, and Food Output by Region: China, 2000 Western Central Eastern Provinces Provinces Provinces Rural income per capita (yuan) 1,557 2,030 2,994 Share of population that is rural (%) 81 75 71 Share of rural labor force in agriculture (%) 75 73 61 Share of farm household income from off-farm (%): Wages 20 22 36 Total 38 35 58 Share of national agricultural GDP (%) 17 34 49 Share of regional food output (%) from: Rice 8 11 6 Wheat 5 5 4 Maize 5 5 3 Fruits and vegetables 11 10 14 Oilseeds 4 6 3 Sugar 0 0 1 Cotton 3 2 1 Dairy products 2 1 1 Other livestock products 20 19 17 Fish products 2 6 20 Other food, agricultural, and forest products 40 35 30 Total 100 100 100 Sources: Center for Chinese Agricultural Policy's CAPSIM Model; China National Statistics Bureau, China Rural Statistical Yearbook 2001. TABLE 7.8 Income and Its Distribution: Rural and Urban China, 1980­99 Income Disparity Real Per Capita Income Index Urban/Rural Year Rural Average Poorest 20 Percent Gini Coefficient Income Ratio 1980 100 100 0.24 3.4 1985 175 157 0.23 2.2 1990 183 154 0.31 2.3 1995 239 179 0.34 2.8 1999 330 224 0.32 2.9 Sources: China National Statistics Bureau (CNSB), China Statistical Yearbook 1989­2000; CNSB's rural household income and expenditure surveys. changes between 1995 and 2001 are assumed to the additional reform commitments to be imple- have been in anticipation of the requirements of, mented after 2001, relative to the revised base case and thus part of, China's WTO accession. These in which China's reforms only up to 2001 are in changes are analyzed in detail in Ianchovichina and place and there are no further reforms to 2007. For Martin (2004), together with the effects of imple- key agricultural import policies, these remaining menting over the next few years the remainder of reform commitments are assumed to shift the China's commitments as recorded in its WTO Pro- nominal rates of protection (NRPs) from column tocol of Accession.6 In this chapter we focus just on 3 to column 6 of table 7.6. As well, the export The Impacts of WTO Accession on Chinese Agriculture and Rural Poverty 109 subsidies in place in 2001 (34 percent for maize, Ravallion (2004). It keeps the aggregate trade bal- 10 percent for cotton) are eliminated, and we ance and government tax revenue as a fixed share of assume no new farm production subsidies are GDP (with little change in net international capital introduced.7 The choices of new agricultural NRPs flows), and holds employment constant so that fall into three categories: wages adjust endogenously. · no change if they were negative in 2001 (rice, What Do the Results Show? meats, fruits and vegetables · a move to part way between the in-quota and We begin with the bottom line of the main scenario out-of-quota tariffs if the TRQs become binding before revealing the details. on imports (wheat, coarse grains, sugar, cotton) · a move down to the new in-quota tariffs The Core WTO Accession Scenario (oilseeds, milk). The core empirical results suggest that WTO acces- The sensitivity of the results for the first two cate- sion will increase farm­nonfarm income inequal- gories is explored in the results section later in this ity. The main reason for this greater inequality is chapter. that the relative producer prices of farm products If this reform were to require a movement of are projected by the GTAP model to fall more than unskilled labor out of farm activities, three impedi- the prices of labor-intensive nonfarm products ments have to be kept in mind. The first is that after the completion of WTO accession reforms those farm workers would be less than perfect sub- (table 7.9). stitutes for those already in nonfarm pursuits. The removal of the "voluntary" export restraint Econometric work by Sicular and Zhao (2004) sug- on sales of textile products to the United States and gests that restraints on mobility could be approxi- European Union is not the only boost to unskilled mated via a constant elasticity of transformation labor off farms. There is, in addition, a lower cost (CET) function with an elasticity of transformation structure in unskilled, labor-intensive manufactur- of 1.32.8 Accordingly, we have incorporated that ing activities for three reasons: function in the GTAP model for China. The second impediment to off-farm migration is that urban · The lower demand for labor on farms lowers the social welfare benefits such as subsidies for housing, cost of unskilled labor in manufacturing. food, education, and health care are not available to · Import taxes on the intermediate inputs used in nonurban people, except by purchasing a residence those manufacturing activities are lower because permit or hukou (Zhao 1999). And the third imped- of the accession process. iment is that farm workers who permanently cut · The real exchange rate effect of the tariff reduc- their ties with agriculture may lose entitlement to tions lowers the cost of nontraded goods and returns from their family's land, and even the direct other factors used as inputs into manufacturing support and assistance of family members (Hussain production. 2004). These latter two impediments have con- tributed to the persistence of a large gap in farm As a result, the quantity of unskilled nonfarm labor versus nonfarm returns to unskilled labor. We demanded is greater (by 0.8 percent). However, model that gap as a "tax" wedge that raises the cost lower farm product prices mean the quantity of of labor to urban employers, with urban unskilled unskilled farm labor demanded is less (by 1.7 per- workers receiving the tax-inclusive wage. cent), resulting in a 0.7 percent fall in the real wage The closure adopted is a long-run one in which, for unskilled farm labor and a rise in real wages in addition to the earlier assumptions about for unskilled nonfarm labor of 1.2 percent, after unskilled labor, we assume agricultural land is adjusting for the change in the aggregate cost of mobile between industries within the agriculture living. Farmers are also made worse off by the lower sector, and skilled labor and capital are mobile demand for farmland, the return from which is within and between sectors. This closure contrasts 5.5 percent lower in 2007 after WTO accession with the short-run closure used by Chen and reforms. Meanwhile, the real wages of skilled labor 110 China and the WTO TABLE 7.9 Sectoral Volume Effects of China's WTO Accession Reforms (Core Case), 2002­07 Household Trade Balance Producer Consumer Output Employment Consumption Exports Imports (Millions of Prices Prices (%) (%) (%) (%) (%) U.S. Dollars) (%) (%) Rice 2.1 2.3 0.1 6.1 7.1 64 0.9 0.9 Wheat 2.0 2.3 0.0 18.9 10.1 174 1.7 0.4 Feed grains 2.3 2.6 0.1 77.8 2.4 596 1.9 1.9 Fruits and vegetables 3.4 3.7 0.1 14.6 6.3 214 1.9 0.1 Oilseeds 7.9 8.4 0.9 29.8 20.9 789 2.8 4.7 Sugar 6.5 7.4 0.6 13.9 24.1 73 1.9 3.1 Plant-based fibers 15.8 16.4 0.6 51.8 7.7 189 0.1 3.1 Livestock and meat 1.3 1.1 0.0 15.5 8.9 837 1.6 0.2 Dairy 2.0 2.4 0.0 13.5 23.8 143 1.5 0.2 Other food 5.9 6.4 0.4 11.4 62.6 3460 1.7 1.8 Beverages and tobacco 33.0 33.1 1.5 9.7 112.4 14,222 1.8 6.9 Extractive industries 1.0 1.3 0.2 7.5 4.4 2,088 0.7 1.2 Textiles 15.6 15.5 0.7 32.7 38.5 10,366 1.7 3.2 Apparel 57.3 56.1 0.5 105.8 30.9 49,690 0.5 1.9 Light manufacturing 3.7 3.7 0.0 5.9 6.8 1,786 0.9 0.0 Petrochemical industry 2.3 2.3 0.2 3.1 11.8 8,810 0.7 0.8 Metals 2.1 2.1 0.3 3.7 6.8 1,893 0.4 1.3 Automobiles 1.4 2.2 1.0 27.7 24.0 516 3.9 4.2 Electronics 0.6 0.4 0.5 6.7 6.8 453 1.3 1.7 Other manufactures 2.1 2.2 0.2 4.1 18.9 11,291 0.5 0.8 Trade and transport 0.0 0.0 0.3 0.8 0.4 493 0.2 1.6 Construction 0.9 0.9 0.4 2.7 17.5 436 0.2 1.7 Communication 0.5 0.5 0.4 0.5 10.9 56 0.1 1.9 Commercial services 2.0 2.0 0.5 0.4 35.4 1,749 0.2 1.9 Other services 1.7 1.8 0.3 1.4 33.6 1,525 0.1 1.6 Source: Authors' GTAP results. increase by 0.8 percent, and the rewards to nonfarm certainly between rural and urban areas--appears capital are 1.3 percent higher. poised to increase slightly unless remedial policy Together, these results suggest that the owners of action is forthcoming. For farm households nonfarm capital gain almost the same in propor- entirely dependent on earnings from agriculture tional terms as unskilled laborers in nonfarm (type A in table 7.10), income would fall 1.6 percent employment, but the latter do better than skilled on average. This decrease would differ little across workers. Therefore, on balance, income inequality the three regions identified in table 7.7, because may improve slightly among nonfarm households product shares for farm output--when fish prod- dependent mainly on labor income.9 ucts are ignored--are reasonably similar in the However, income distribution can be expected western, central, and eastern provinces.10 For farm to worsen slightly, just as between farm and non- households earning 30 percent of their income farm households, although the degree depends on from nonfarm unskilled work, however, that the proportion of farm household income earned income fall is only half as large (0.8 percent), and off the farm. With only 1.7 percent (or about 6 mil- for farm households earning 60 percent of their lion) of unskilled farm workers leaving agriculture income from nonfarm unskilled work, their for nonfarm work and with land returns depressed incomes would not decline at all (types B and C, by 5.5 percent in addition to farm labor returns respectively, in table 7.10, rows 7 and 8). being 0.7 percent lower, the gap between farm and The sectoral details of the GTAP results are nonfarm incomes even within rural areas--and summarized in table 7.9. The real consumer prices, The Impacts of WTO Accession on Chinese Agriculture and Rural Poverty 111 TABLE 7.10 Change in China's Real Factor Prices and National Economic Welfare Attributable to WTO Accession, 2001­07 (percent) Alternative 2: Alternative 1: Core Case Plus Also Core Accession Greater Agricultural Removing Negative Case Protection Cuts Agricultural Protection Farm unskilled wages 0.7 0.9 0.5 Rental price of land 5.5 6.4 4.7 Nonfarm unskilled wages 1.2 1.4 1.1 Skilled labor wages 0.8 1.0 1.5 Rental price of capital 1.3 1.5 1.5 Farm householda income­A 1.6 1.9 1.3 Farm householda income­B 0.8 0.9 0.6 Farm householda income­C 0.1 0.1 0.1 National economic welfare (increase in billions of 1997 US$) 9.56 9.57 9.87 aAccording to the GTAP database, 57 percent of farm income from agriculture comes from unskilled labor, 26 percent from land, and 17 percent from capital. In 1999 on average 51 percent of rural household income in China was earned outside agriculture, mostly from unskilled labor. Therefore, to illustrate the importance of those off-farm earnings, three types of farm households are shown in this table: type A, where it is assumed that nonfarm unskilled labor contributes 0 percent of farm household income; type B, 30 percent; and type C, 60 percent. It is assumed that all households have the same expenditure pattern, and so all changes are net of the consumer price index change. Source: Calculated from the authors' GTAP results. relative to the consumer price index (CPI) driven China's food, feed, and fiber self-sufficiency will be lower the most by WTO accession are those for reduced at least slightly by these reforms. But the motor vehicles, oilseeds, and sugar (and for bever- extent is really quite minor:the trade balance column ages and tobacco, although if China was using in table 7.9 suggests that for all agricultural and food import taxes on those items as a form of consump- products net imports would be greater because of the tion tax and their decline were to be matched by an remaining accession reforms by only $3.96 billion a increase in domestic sales taxation, those price year by 2007 (in 1997 U.S. dollars), which represents declines may not materialize). The prices for textile only 1 percent of total imports. products, and to a lesser extent clothing, are driven The results described here depend as always on lower as well. Among the farm products, consumer the assumptions in the model. To check the sensi- prices rise slightly for livestock products, somewhat tivity of some of those assumptions, two alternative more for grains, and significantly for cotton (plant- scenarios--one with greater agricultural protection based fibers). cuts and the other removing negative agricultural Producer prices (also shown relative to the CPI in protection--were run to compare their results with table 7.9) fall more than consumer prices because those in the base accession scenario. of a uniform consumption tax of 1.9 percent that is levied to compensate for the loss of import tariff rev- Alternative Scenario 1: Greater Agricultural enue. The prices of farm products fall more than Protection Cuts those of most other products except autos, and farm outputisdownforallbutcottonandmeat.Moreover, In this scenario the grain, sugar, and cotton feed grain exports shrink by three-quarters and cot- NRPs drop to the in-quota tariff levels shown in ton exports by half with the abolishment of export table 7.6. In the additional GTAP simulation run, subsidies.The difference in the effects on production the differences for factor rewards are not huge in and consumption shown in table 7.9 reveal that the aggregate, but they would be in the direction of 112 China and the WTO worsening income inequality: unskilled farm wages depending just on labor income (assuming urban would fall 0.9 instead of 0.7 percent, and, on the laborers are more skilled). However, some farm one hand, rewards to farmland would fall 6.4 households facing increased import competition instead of 5.5 percent, while, on the other, nonfarm may be worse off in this case, other things being wages would rise 1.4 instead of 1.2 percent for equal, if they are unskilled workers and 1.0 instead of 0.8 percent for skilled workers (table 7.10). These changes would · unable to send household members to jobs in attract only another million workers from farms. expanding industrial and service industries Even though agricultural incomes would be lower, · too poorly served with infrastructure to attract farm household income would not fall if at least such activities to their own region 60 percent of income comes from wages of non- · unable to diversify into producing farm goods farm unskilled labor (see row 8 of table 7.10). whose relative prices have risen Domestic production of grains, sugar, and cotton · lacking relatives who are able to repatriate non- would, however, be less, and domestic consump- farm earnings to them. tion would be greater, so that self-sufficiency in those products would be slightly lower. Even so, the Thus in the core scenario the incidence of rural net imports of all food and agricultural products nonfarm poverty will fall mainly because of the would be greater by only $1.5 billion a year by 2007 growth in wages for unskilled workers in rural non- ($5.43 billion instead of $3.96 billion). Such an farm activities, while poverty may well increase in import increase would be within the tariff-rate agriculture-based hinterland provinces a long way quotas for those items, with the possible exception from markets and in regions poorly served by of maize (depending on the extent to which other the infrastructure needed to attract investment in feed grains that are not TRQ-restricted, such as expanding activities such as textiles and clothing. barley, are substitutable for maize). National eco- The first alternative scenario shows that this sit- nomic welfare would be only very slightly greater in uation would be exacerbated slightly if the TRQ- this case as compared with the core scenario (see protected items (grains, sugar, and cotton) were bottom row of table 7.10). to become even less protected than we initially assumed. By way of contrast, the second alternative Alternative Scenario 2: Removing Negative scenario suggests that the situation could be ame- Agricultural Protection liorated slightly by removing the negative protec- tion affecting rice, meats, and fruits and vegetables. If the negative NRPs for rice, meats, and fruits and But both of these alternatives involve only small vegetables were to be raised to zero, the income dis- changes to the magnitudes of effects rather than tributionaleffectswouldgointhedirectionopposite altering the sign of those effects, and both add only to that in the first alternative scenario--that is, a small amount to the aggregate gains from trade toward less inequality between farm and nonfarm liberalization. households. The changes are not great though, even National self-sufficiency in farm products would though these products account for nearly 40 per- decrease slightly, particularly for feed grains and cent of the value of food and agricultural output cotton as demand for livestock products grows, in China. A comparison of columns 1, 2, and 3 of driven by the income gains from trade reform, and table 7.10 reveal they would involve about as much as the production of natural fiber-based textiles improvement in income distribution as the previous and clothing expands. But, overall, most of the alternative scenario would worsen it. This case declines in domestic agricultural production as involvesa3percentlargernationaleconomicwelfare a consequence of the remaining postaccession gain than the core case (bottom row of table 7.10). reforms are relatively tiny in magnitude, especially when compared with the growth in farm output Conclusions and Policy Implications that would occur as a result of normal economic Our initial analysis suggests that rural nonfarm growth (compare columns 1 and 5 of table 7.11). incomes will rise on average absolutely and possibly What should be done if some farmers' incomes even faster than urban incomes for households worsen relative to those of nonfarm households, The Impacts of WTO Accession on Chinese Agriculture and Rural Poverty 113 TABLE 7.11 Changes in China's Farm Output, Employment, and Trade Volumes: Without WTO Accession, 1995­2007, and Resulting from China's WTO Accession Reforms, 2002­07 (percent) Changes Without WTO Changes Due to WTO Accession, 1995­2007 Accession, 2002­07 Output Employment Exports Imports Output Employment Exports Imports Rice 63.8 11.5 134.7 8.8 2.1 2.3 6.1 7.1 Wheat 81.4 6.4 15.2 126.3 2.0 2.3 18.9 10.1 Feed grains 109.5 23.8 0.6 95.9 2.3 2.6 77.8 2.4 Fruits and 98.2 16.8 10.8 122.1 3.4 3.7 14.6 6.3 vegetables Oilseeds 100.9 18.4 36.0 151.7 7.9 8.4 29.8 20.9 Sugar 112.5 14.5 109.4 88.7 6.5 7.4 13.9 24.1 Plant-based fibers 137.2 41.1 8.5 146.1 15.8 16.4 51.8 7.7 Livestock and meat 121.9 25.6 12.8 135.3 1.3 1.1 15.5 8.9 Dairy products 122.5 18.8 60.5 100.3 2.0 2.4 13.5 23.8 Other food 110.8 1.5 76.8 58.5 5.9 6.4 11.4 62.6 Source: Authors' GTAP results. and if there is concern about the fall in agricultural Second, improvements in rural infrastructure self-sufficiency? Rather than argue against trade such as roads and rail mean that a larger share of reform, policymakers should seek the best ways to the price eventually received at the end of the mar- deal with those concerns (and with any transitory keting chain for farm products can be passed back unemployment that might follow reform). The to farmers. Such improvements also lower the bar- most efficient policy responses are likely to involve rier for off-farm work by members of farm house- investments in rural human capital, rural infra- holds, making it easier for them to take advantage structure, and agricultural research and develop- of expanding employment opportunities in rural ment (Fan, Zhang, and Zhang 2002); improve- townships. ments in the land tenure system and rural financial Third, agricultural research and development can markets; reductions in the informal taxes/levies on ease both urban and rural poverty (see Fan, Fang, farmers by local governments; and changes in grain and Zhang 2001; Hazell and Haddad 2001). A boost marketing. in agricultural productivity could significantly offset First, the government might consider further the 2­8 percent drop in agricultural production that investments in basic rural education and health is estimated in the core scenario to result from WTO services to reduce the adverse effects of trade accession. An important policy issue here is whether reform on the incidence of poverty and perceived China should deny itself the use of genetically mod- food security. Better education and health for farm- ified organisms in food production.11 ers' children not only boosts their farm productiv- Fourth, improvements in the land tenure system ity should they choose to stay on the farm after fin- would not only increase the incentive to invest ishing school, but also increases their capacity to more in land but would also enhance the collateral find more lucrative off-farm work and to adjust to of farm households. If accompanied by improve- nonagricultural employment and living (Schultz ments in rural financial markets, investments by 1975; Zhang, Huang, and Rozelle 2002). In addition farmers back into agriculture would rise. They to those longer-term benefits, there could also be would rise even further if returns were increased an immediate poverty-alleviating effect if the gov- via reductions in the informal taxes/levies imposed ernment were to cut basic school fees and make up by local governments on farmers.12 the shortfall with a bigger direct grant to rural pri- Finally, the government might reduce its regula- mary and middle schools. tion of grain marketing and, in particular, cease 114 China and the WTO compulsory procurement from farmers at less than enhances China's chances of expanding its access to market prices and reduce the provision of grain to agricultural more than other markets abroad in the urban consumers at less than market prices. De- future, that would be a positive benefit of WTO emphasizing the governors' grain responsibility accession for China's farmers and rural areas. system (provincial self-sufficiency) would allow Martin (forthcoming) points out that Chinese farm more exploitation of comparative advantage within exports face particularly high barriers abroad, so China as well. this potential benefit is nontrivial in principle If all these measures proved insufficient to sup- (although in practice it may be difficult to secure, port the incomes of the poorest farm households, especially if the main barriers are sanitary and short-term adjustment assistance via inframarginal phytosanitary measures). That proposition is tested (and therefore not output-inducing) producer by Yu and Frandsen (2002), also using the GTAP price subsidies could be provided as an efficient model. They find that reductions in barriers way to boost those households' incomes without to agricultural imports and in domestic support to boosting farm output in an equal but opposite way farmers in OECD countries reduce the extent to that used to tax farmers in earlier decades (see to which China's farm output would fall with WTO Shea 2003). Such an intervention could well be accession, and that in some cases they lead to deemed WTO-consistent because of its decoupled expansion rather than contraction of outputs. As a nature. In any case, if it just targets poor farmers it consequence, China's agricultural imports would is unlikely to ever exceed 8.5 percent of the value of fall slightly and its agricultural exports would be China's output of the product concerned (China's greater. These changes are reflected in table 7.12. de minimis exemption limit for product-specific It shows that not only would China's food self- support under Article 6.4 of the WTO's Agreement sufficiency be higher with than without agricul- on Agriculture). tural protection in the European Union, United Finally, now that China is in the WTO it can take States, and Japan, but that the difference is in most part in the new rounds of multilateral trade negoti- cases more than enough to offset the fall in self- ations and thereby seek increased market access for sufficiency that is projected to result from China's its exports of farm (and other) products abroad. WTO accession. Such reform in the OECD would Although this factor is not taken into account in clearly benefit farm households in China, providing the analysis described here, if WTO membership a further pro-poor consequence of trade reform. TABLE 7.12 Effect on Agricultural Self-Sufficiency of China's WTO Accession without and with Cuts in Agricultural Protection in OECD Countries (domestic production as a percentage of domestic consumption) China's WTO Accession with EU, United States, and Base Case (No China's WTO Accession Japan Removing Agricultural WTO Accession with No Change in Tariffs, Exports Subsidies, for China) OECD Policies and Domestic Support Rice 99.9 99.4 99.6 Wheat 93.0 88.2 89.1 Feed grains 98.1 98.1 99.6 Fruits and vegetables 99.7 99.7 99.9 Oilseeds 81.0 81.3 82.3 Sugar 77.2 77.4 82.9 Plant-based fibers 83.8 83.7 83.9 Ruminant meat 88.5 88.2 89.4 Nonruminant meat 97.7 97.8 98.9 Dairy products 81.8 80.0 84.9 Source: Yu and Frandsen 2002, table 4. The Impacts of WTO Accession on Chinese Agriculture and Rural Poverty 115 Notes first maize and oilseeds and then rice and wheat on China, India, and other countries, see Anderson and Jackson (2003). 1. The difficulties China has had in exporting food products 12. According to Lin and others (2002), such levies amount to, for example, Japan, the Republic of Korea, and the United to as much as one-quarter of the net incomes of poorer farmers Kingdom in recent years because of those countries' quarantine in the hinterlands. and sanitary and phytosanitary measures should ease after WTO accession, or at least be challengeable under WTO dispute settle- ment provisions. Because those measures are notoriously diffi- References cult to predict, they are ignored in our empirical analysis (as are prospective changes to other nontariff measures that may limit The word processed describes informally reproduced works that China's access to foreign markets for farm or nonfarm prod- may not be commonly available through libraries. ucts). Also ignored is the possibility that China itself may use Anderson, K. 1995."Lobbying Incentives and the Pattern of Pro- quarantine measures to limit its imports of farm products. tection in Rich and Poor Countries." Economic Development 2. However, soybean imports grew substantially when the and Cultural Change 43 (2): 401­23. government (in anticipation of WTO accession) lowered the Anderson, K., Y. Hayami, and [others]. 1986. The Political Econ- tariff on soybeans from 114 percent to 3 percent in 2000. omy of Agricultural Protection: East Asia in International Per- 3. See Sah and Stiglitz (1992) and Anderson (1995). spective. Boston, London, and Sydney: Allen and Unwin (also 4. All dollar amounts are current U.S. dollars. available in an expanded Chinese language edition translated 5. The GTAP model is a multiregional, static, applied general by F. Cai. Tianjin: People's Publishing House, 1996). equilibrium model based on neoclassical microeconomic the- Anderson, K., and L. A. Jackson. 2003. "Standards, Trade and ory, including full employment of all factors of production, con- Protection: The Case of GMOs." Paper presented at a semi- stant returns to scale, and perfect competition. For a discussion nar at the World Bank, October 2. Washington, D.C. of the prospect of greater unemployment during the accession Brown, L. R. 1995. Who Will Feed China? Wake-up Call for a period, see Zhai and Wang (2002). The numeraire is the world Small Planet. New York: Norton. price of exports. See Hertel (1997) for comprehensive documen- CDS Consulting Co. 2002. "Applied Tariffs in China, 2001." tation. The Version 5 database is described at www.gtap.org. Washington, D.C. 6. A particularly important feature of their analysis is the Chen, S., and M. Ravallion. 2004. "Household Welfare Impacts inclusion of China's duty exemptions in the base scenario, of China's Accession to the WTO." In D. Bhattasali, S. Li, and because, otherwise, the model would overstate the gains from W. Martin, eds., China and the WTO: Accession, Policy tariff reductions. Tariff cuts are from the 2001 applied rates to Reform, and Poverty Reduction Strategies. Washington, D.C.: the postaccession bound rates (or zero if the latter exceed the World Bank and Oxford University Press. former). In this application the aggregate trade balance and gov- Fan, S., C. Fang, and X. Zhang. 2001. "How Agricultural ernment tax revenue are both assumed to remain a fixed share of Research Affects Urban Poverty in Developing Countries: GDP. The 2001 trade data are from the UN's COMTRADE data- The Case of China." Environment and Production Technol- base, and the 2001 applied tariffs for China are from CDS Con- ogy Discussion Paper 83. International Food Policy Research sulting Co. (2002). Institute, Washington, D.C. 7. Three nonfarm reforms are worthy of mention. The "vol- Fan, S., L. Zhang, and X. Zhang. 2002. Growth, Inequality, and untary" export restraint on China's textile and clothing exports Poverty in China: The Role of Public Investments. IFPRI to the United States and European Union, expressed in the base Research Report 125. Washington, D.C.: International Food scenario as taxes on those exports, are removed; restructuring of Policy Research Institute. the motor vehicle and parts industry after WTO accession is Francois, J., and D. Spinanger. 2004."Regulated Efficiency, WTO modeled as a 20 percent productivity boost to vehicle assembly, Accession and the Motor Vehicle Sector in China." In D. following François and Spinanger (2004); and liberalization of Bhattasali, S. Li, and W. Martin, eds. China and the WTO: China's services trade also follows François and Spinanger Accession, Policy Reform, and Poverty Reduction Strategies. (2004). Washington, D.C.: World Bank and Oxford University Press. 8. In a subsequent analysis, Sicular and Zhao (2004) provide Gilbert, J., and T. Wahl. 2002. "Applied General Equilibrium estimates of 2.67 for the "push" elasticity of transformation from Assessments of Trade Liberalization in China." World Econ- changes in agricultural returns to changes in the supply of labor omy 25 (5): 697­731. out of agriculture, but only 0.6 from changes in nonagricultural Hazell, P., and L. Haddad. 2001. "Agricultural Research and returns. Ianchovichina and Martin (2004) conducted sensitivity Poverty Reduction." Food, Agriculture, and the Environ- analysis on the implications of changes in these parameters, and ment. Discussion Paper 34. International Food Policy they found the results are not altered greatly. Research Institute, Washington, D.C. 9. Wages of skilled workers might increase more than sug- Hertel, T. W., ed. 1997. Global Trade Analysis: Modeling and gested here because we do not capture the endogenous produc- Applications. Cambridge and New York: Cambridge Univer- tivity growth resulting from the substantial liberalization of sity Press. the service sectors. For a recent study that does incorporate Huang, J. 1998. "Evaluation of Current Three-Grain-Policy." that effect using a dynamic version of the GTAP model, see CCAP's Working Paper WP-00-C01. Center for Chinese Ianchovichina and Walmsley (2003). Agricultural Policy, Chinese Academy of Agricultural Sci- 10. However, in the northern and southern provinces of ences, Beijing. China the differences could be large because wheat, maize, soy- ------. 2001. "Agricultural Policy and Food Security in China." bean, and cotton are planted primarily in northern China, and Center for Chinese Agricultural Policy, Chinese Academy of rice, horticultural products, livestock, and fish are raised mostly Sciences, Beijing. Processed. in southern China. Huang, J., and H. Ma. 1998. "The 20-Year Reform and the Role 11. For an empirical analysis using GTAP of the impact of of Agriculture in China: Capital Flow from Rural to Urban the adoption or nonadoption of genetically modified varieties of and from Agriculture to Industry." Reform 5: 56­63. 116 China and the WTO Huang, J., S. Rozelle, and M. Chang. 2004. "The Nature of Dis- Ministry of Agriculture. China's Township and Village Enterprise tortions to Agricultural Incentives in China and Implica- Statistical Yearbook. Beijing: China Statistics Press. tions of WTO Accession." In D. Bhattasali, S. Li, and W. National Bureau of Statistics of China a. China Statistical Year- Martin, eds., China and the WTO: Accession, Policy Reform, book (various issues). Beijing: China Statistics Press. and Poverty Reduction Strategies. Washington, D.C.: World National Bureau of Statistics of China b. China Rural Statistical Bank and Oxford University Press. Yearbook. Beijing: China Statistics Press. Hussain, A. 2004. "Coping and Adapting to Job Losses and Rozelle, S., A. Park, J. Huang, and H. Jin. 1997. "Liberalization Declines in Farm Earnings." In D. Bhattasali, S. Li, and W. and Rural Market Integration in China." American Journal of Martin, eds., China and the WTO: Accession, Policy Reform, Agricultural Economics 79 (2): 635­42. and Poverty Reduction Strategies. Washington, D.C.: World Sah, R., and J. E. Stiglitz. 1992. Peasants versus City-Dwellers: Bank and Oxford University Press. Taxation and the Burden of Development. Oxford: Clarendon Ianchovichina, E., and W. Martin. 2004. "Economic Impacts of Press. China's Accession to the WTO." In D. Bhattasali, S. Li, and Schultz, T. W. 1975. "The Value of the Ability to Deal with Dise- W. Martin, eds., China and the WTO: Accession, Policy quilibria." Journal of Economic Literature 13 (3): 827­46. Reform, and Poverty Reduction Strategies. Washington, D.C.: Shea, Y. P. 2003. The Political Economy of China's Grain Policy World Bank and Oxford University Press. Reform. Ph.D. diss., University of Adelaide. Ianchovichina, E., and T. Walmsley. 2003. "Impact of China's Sicular, T. 1988. "Plan and Market in China's Agricultural Com- WTO Accession on East Asia." Policy Research Working merce." Journal of Political Economy 96 (2): 383­87. Paper 3109. World Bank, Washington, D.C. Sicular, T., and Y. Zhao. 2004. "Earnings and Labor Mobility in Kanbur, R., and X. Zhang. 2001."Fifty Years of Regional Inequal- Rural China: Implications for China's WTO Entry." In D. ity in China: A Journey through Evolution, Reform and Bhattasali, S. Li, and W. Martin, eds., China and the WTO: Openness." CEPR Discussion Paper No. 2887. Centre for Accession, Policy Reform, and Poverty Reduction Strategies. Economic Policy Research, London. Washington, D.C.: World Bank and Oxford University Press. Lardy, N. 2002. Integrating China in the Global Economy. WTO (World Trade Organization). 2001. Report of the Working Washington, D.C.: Brookings Institution Press. Party on the Accession of China., WT/MIN(01)/3. Geneva. Lin, Y. F., R. Tao, M. Liu and Q. Zhang. 2002."Rural Direct Taxa- ------. 2003. World Trade Report 2003. Geneva. tion and Government Regulation in China: Economic Yu, W., and S. E. Frandsen. 2002. "China's WTO Commitments Analysis and Policy Implications." Working Paper. China in Agriculture: Does the Impact Depend on OECD Agricul- Center for Economic Research, Beijing University. tural Policies?" Paper presented at the Fifth Conference on Lindert, P. 1991."Historical Patterns of Agricultural Protection." Global Economic Analysis, Taipei, June 5­7. In P. Timmer, ed. Agriculture and the State. Ithaca: Cornell Zhai, F., and Z. Wang. 2002. "WTO Accession, Rural Labor University Press. Migration, and Urban Unemployment in China." Urban Martin, W. Forthcoming. "Implication of Reform and WTO Studies 39 (12): 2199­217. Accession for China's Agricultural Policies." Economies in Zhang, L., J. Huang, and S. Rozelle. 2002. "Employment, Reces- Transition. sions, and the Role of Education in Rural China." China Eco- Mathews, A. 2001. "The Possible Impact of China's WTO Acces- nomic Review 114: 1­16. sion on the WTO Agricultural Negotiations." Trinity Eco- Zhao,Y. 1999."Leaving the Countryside: Rural-to-Urban Migra- nomic Papers No. 15. Department of Economics, Trinity tion Decision in China." American Economic Review 89 (2): College, Dublin. 281­86. McCulloch, N., L. A. Winters, and X. Cirera. 2001. Trade Liberal- ization and Poverty: A Handbook. London: Centre for Eco- nomic Policy Research. 8 The Services Dimension of China's Accession to the WTO Aaditya Mattoo The fulfillment of China's accession commitments cross-subsidization ends, and in the factor markets, will lead to one of the most dramatic episodes of as the scale of activity expands but surplus labor is services sector liberalization seen in any country. shed. It is, unfortunately, not easy to quantify these Over the space of some six years, one of the most effects. closed services markets has promised to become It is possible, however, to analyze a key determi- one of the most open. This chapter begins by nant of these changes: how the transition to more describing China's policy commitments, and it then open markets is managed. Already known are the asks whether the disciplines implied by these com- policy status quo and the destination as specified in mitments are desirable and, more important, how the commitments, but how much discretion does the process of implementing these commitments is China retain in its choice of policy? China's sched- best managed. ule of commitments specifies not only the terminal Successful reforms will lead to significant situation, but also how many of the modifications improvements in the services markets themselves of policy will be phased in. Even though changes in terms of prices, quality, product variety, and the in the degree of competition and the pattern of availability of new products.1 More efficient provi- ownership are largely prespecified, they can be sion of telecommunications, transport, and other accelerated--commitments in the General Agree- services will also advance the integration of the ment on Trade in Services (GATS) do not prevent a Chinese economy with that of the rest of the world country from liberalizing faster than it has said it and within the country itself--provided access to would. More important, China retains much free- improved services is not limited to the coastal dom in designing domestic regulation, and reforms enclaves. The poor are also likely to be affected--in in this area will crucially affect the magnitude and the product market, as efficiency increases but distribution of gains from liberalization. The views contained in this paper are those of the author and should not be attributed to the World Bank. This paper has benefited greatly from the comments of Carsten Fink and two anonymous referees. Antonia Carzeniga, Stijn Claessens, Caroline Freund, Christina Lund, Will Martin, Cristina Neagu, Randeep Rathindran, and Beata Smarzynska also made valuable contributions. 117 118 China and the WTO The chapter begins by describing China's acces- (Article XVII). These provisions apply only to sec- sion commitments. It then examines whether the tors explicitly included by a member in its schedule implied loss of policy discretion is desirable and of commitments--a "positive list" approach--and whether China's commitments promote good policy. these sectoral commitments, too, are subject to the Finally, the chapter identifies the priorities for com- limitations that the member has scheduled. GATS plementary reform and offers some conclusions. commitments are guarantees, and the absence of such guarantees need not mean that access to a par- China's Accession Commitments ticular market is denied in practice. The market access provision prohibits six types The GATS Framework and Services Liberalization of limitations, unless they have been inscribed by a The GATS covers all measures taken by members Member in its schedule. These are as follows: of the World Trade Organization (WTO) affecting trade in services and all service sectors.2 The agree- · limitations on the number of suppliers ment is unusual in taking a wide view of what con- · limitations on the total value of service transac- stitutes trade, and it defines trade in services as the tions or assets supply of a service through any of four modes. · limitations on the total number of service Mode 1, cross-border supply, is analogous to trade in operations or on the total quantity of service goods and arises when a service crosses a national output frontier, for example, the purchase of software or · limitations on the total number of natural insurance by a consumer from a supplier located persons (i.e. individuals rather than juridical abroad. Mode 2, consumption abroad, arises when persons) that may be employed the consumer travels to the territory of a service · measures that restrict or require specific types of supplier to purchase, for example, tourism, educa- legal entity or joint venture tion, or health services. Mode 3, commercial pres- · limitations on the participation of foreign ence, involves foreign direct investment--for exam- capital. ple, when a foreign bank or telecommunications firm establishes a branch or subsidiary in the terri- In scheduled sectors, the existence of any of these tory of a country. And mode 4, movement of indi- limitations has to be indicated with respect to each viduals, occurs when independent service providers of the four modes of supply described earlier. or employees of a multinational firm temporarily In the traditional manner followed by the Gen- move to another country. eral Agreement on Tariffs and Trade (GATT), Certain GATS obligations apply across the national treatment is defined under Article XVII as board, while others depend on the sector-specific treatment no less favorable than that accorded to commitments assumed by individual Members. like domestic services or service suppliers. In con- The most important of the general obligations are trast to the GATT approach, however, members transparency and the most-favored-nation (MFN) may inscribe limitations on national treatment in principle. The transparency obligation requires, their schedules with respect to each of the four among other things, that each member publish modes of supply, as in the market access provision. promptly "all relevant measures of general applica- tion" (that is, measures other than those involving Numerical Overview of Commitments: How only individual service suppliers) to trade in ser- Different Is China? vices. The MFN obligation prevents members from discriminating among their trading partners. The This section compares China's services commit- agreement, however, permits members to list tem- ments with those of other WTO members. It begins porary exemptions to MFN status. with a brief look at the state of access today, before The liberalizing content of the GATS depends China's liberalization commitments have been on the extent and nature of sector-specific commit- phased in, for a representative sample of sectors. ments assumed by individual members. The core Figure 8.1 compares China's commitments for provisions of the GATS in this context relate to each mode with those of industrial, developing, market access (Article XVI) and national treatment and acceding countries. On consumption abroad The Services Dimension of China's Accession to the WTO 119 FIGURE 8.1 Country Services Commitments by Mode Percent 100 90 80 70 60 50 40 30 20 10 0 IC DC AC CHN IC DC AC CHN IC DC AC CHN IC DC AC CHN Mode 1 Mode 2 Mode 3 Mode 4 IC industrial countries, DC developing countries, AC acceding countries, CHN China. Note: Calculated on the basis of a sample of 37 sectors deemed representative for various service sectors. See chapter appendix table 8.3 for details on China and WTO (1999) for details on other countries. The upper part of each bar represents partial commitments, the lower part full commitments. Source: WTO. (mode 2), commercial presence (mode 3), and the Overall for China, the coverage of market access presence of natural persons (mode 4), China has commitments (i.e., the unweighted average count) made at least partial commitments in all the sec- was 57.4 percent (table 8.1). This coverage is much tors, and on cross-border supply (mode 1) for over more extensive than the commitments offered in 80 percent of the sectors. This situation compares the Uruguay Round by any other group of countries favorably with the commitments of all other coun- (including high-income ones). The "average cover- try groups. However, the number of sectors with a age," a measure of coverage that better reflects the guarantee of full access in China's schedule today is extent of liberalization of services, was 38.1 percent less than those for the other country groups, with a for China--again showing more openness than that very significant difference for mode 3. of even the high-income countries. China's share of The picture changes quite drastically when completely liberal commitments (i.e., no restric- looking at the period after all the liberalization tions) of the maximum possible was 23.1 percent, has been phased in. Three indicators are computed which is much higher than that for any other group by Ianchovichina, Martin, and Wood (2001) for of developing countries but somewhat lower than both market access and national treatment that for high-income countries. China's commit- commitments: ments on national treatment are deeper and wider than those of all other country groups. · "unweighted average count," indicating the number of sector/mode-of-supply combina- A Closer Look, Sector by Sector tions where a commitment was made about the maximum number possible Typical Restrictions In some respects, China's · "average coverage," providing an arithmetic commitments resemble those made by other coun- weighted average of the weights (0 for no com- tries. For most sectors, modes 1 and 2 are either mitment, 1 for a full commitment, and 0.5 for all fully open or unbound and are not subject to partial commitments) allocated to each cell specific restrictions. Commitments on mode 4, · the "no-restriction commitments share" in a specified horizontally rather than sector by sector, country's total commitments as well as in the are also standard: entry is guaranteed only for man- maximum possible commitments. agers, executives, and specialists--who must be 120 China and the WTO TABLE 8.1 Coverage of Specific Commitments by China to Services in Trade (percent) Low- and Large High-Income Middle-Income Developing Countries Countries Nations China Market access Unweighted average count (sectors/modes 47.3 16.2 38.6 57.4 listed as a share of maximum possible) Average coverage (sectors/modes listed as a 35.9 10.3 22.9 38.1 share of maximum possible, weighted by openness or binding factors) Coverage/count (average coverage as a share 75.9 63.6 59.3 66.4 of the average count) No restrictions as a share of total offer 57.3 45.5 38.7 40.2 (unweighted count) No restrictions as a share of maximum possible 27.1 7.3 14.9 23.1 National treatment Unweighted average count (sectors/modes 47.3 16.2 38.8 57.4 listed as a share of maximum possible) Average coverage (sectors/modes listed as a 37.2 11.2 25.5 45.0 share of maximum possible, weighted by openness or binding factors) Coverage/count (average coverage as a share 78.6 69.1 66.1 78.4 of average count) No restrictions as a share of total offer 65.1 58.0 52.3 63.5 (unweighted count) No restrictions as a share of maximum possible 30.8 9.4 20.2 36.5 Memo item No restrictions on market access and national 24.8 6.9 14.3 29.8 treatment as a share of maximum possible Number of sectors committed 293.0 100.0 239.0 356.0 Source: Ianchovichina, Martin, and Wood 2001. Note: The breadth and depth of commitments by other countries are understated because their more recent commitments in telecommunications and financial services have not been taken fully into account. either intracorporate transferees or employees of (49 percent or less), to 50 percent ownership, to foreign invested enterprises--and for services sales- majority ownership, to full ownership. There is persons on exploratory business visits. No commit- no commitment to allowing establishment of ments are made on other categories of natural per- branches by foreign enterprises, except in spe- sons such as unskilled personnel or movement not cific sectors. linked to a commercial presence. · Geographic scope. Business activity may be It is for the commitments on commercial pres- allowed only in specified cities (e.g., Shanghai) ence that a range of restrictive measures emerge. or in special economic zones (SEZs). They are related to the following: · Business scope. Transactions may be permitted only with a subset of consumers or restricted in · Form of establishment. The typical restriction some other way. is the requirement to form a joint venture that · Regulatory requirements. Foreign firms may be is either an equity joint venture (EJV) or a required to possess a certain minimum amount contractual joint venture (CJV). Foreign owner- of assets and be established as a representative ship in EJVs is frequently restricted to speci- office for a certain period of time before com- fied levels, ranging from minority ownership mencing full business operations. The Services Dimension of China's Accession to the WTO 121 Interestingly, most restrictions pertain to market developed areas"; and there are meaningful com- access; relatively few limitations affect national mitments in urban planning and legal services. treatment.3 In fact, one of the striking aspects of By 2007 geographic and quantitative limitations China's schedules is the country's willingness to will be eliminated in legal services, and fully owned commit across modes and sectors to full national foreign subsidiaries can operate in accounting, tax- treatment for foreign providers. ation, architecture, engineering, and urban plan- ning services. But some restrictions will persist, Commitments: Past, Present, and Future How especially in legal and medical services. Foreign much has China's trade policy already changed, and firms are not allowed directly to participate in legal how much will it change over the next few years? activity in China and are only entitled to work on China's schedules of commitments provide a first legal affairs related to their home country or to source of information. China participated in the entrust work to Chinese firms on behalf of their Uruguay Round services negotiations and submit- clients. In medical services, hospitals cannot be ted a schedule under GATS in April 1994. Because fully foreign-owned and are still subject to quanti- China did not become a member of the WTO at tative limitations in line with China's assessed this time, this schedule did not have legal status. needs. More important, it is not clear how much this schedule reflected the actual openness of Chinese Computer and Related Services This is the one services markets at that time. Nevertheless, it does sector where some commitments have actually provide the basis for an interesting comparison become less liberal. In 1994 there were no restric- with the final schedule. tions on establishment in software implementa- China's schedule of commitments under GATS tion, systems and software consulting, and systems has had legal status ever since China became a analysis. Now establishment can only take place WTO member, and it is the outcome of the tough- through joint ventures, although foreign majority est services negotiations undertaken in the WTO. ownership is permitted. Cross-border delivery in Whereas most WTO members have merely bound these services remains unbound, but it is fully open the policy status quo or even less, China agreed to in all other computer and related services. No allow significant liberalization to be implemented future liberalization has been specified. either immediately or in the near future. One con- sequence is that the schedule describes the state of Telecommunications Much change is anticipated actual policy at the time of accession and over the in this sector over the next few years. Today foreign next few years. providers can provide value added and mobile ser- Relying on these two schedules, table 8.2 reveals vices in and between Shanghai, Guangzhou, and a rough picture of the state of "policy" at three Beijing through joint ventures, subject to stringent points of time: 1994; 2001, the date of accession; restrictions on foreign ownership. By 2004 fixed- and 2008, the date by which all liberalization com- line services can be provided on similar terms in mitments will have been phased in (i.e., seven years and between the same cities. By 2007 all geographic after accession). The earlier date shows how far restrictions will be eliminated and equity restric- China has already come, and the latter date shows tions will be relaxed. But even at the end of the how far it is committed to go. period majority foreign ownership will not be allowed in any area. Furthermore, there is no Professional Services Today, commitments on commitment to allow cross-border any of these professional services are far more liberal than in services.4 1994, and the sector will be highly liberalized over the next few years. Unlike in 1994, China is now Construction and Engineering Services Whereas committed to allowing the cross-border supply commercial presence was unbound in 1994, today it of professional services; quantitative limitations is allowed through joint ventures with foreign no longer apply to accountancy firms; taxation majority ownership permitted, but only in foreign- services can be offered outside of "economically invested construction projects. By 2004 full foreign 122 China and the WTO TABLE 8.2 China's Commitments: Past, Present, and Future Sector 1994 2001 2008 Professional services Legal services No commitments Modes 1 and 2: None CONTINUED RESTRICTIONS Mode 3: Only through one rep. ON BUSINESS SCOPE. office which is allowed to Mode 3: Geographic and engage in profit-making quantitative limitations activities, but only in specified were to be eliminated cities. Business scope restricted by 2002. to home country legal affairs for Chinese and China-based clients and to entrusting, on behalf of foreign clients, Chinese law firms to deal with Chinese legal affairs. Accounting, Modes 1 and 2: Unbound FULLY LIBERALIZED except that auditing, and Mode 3: Through branch partnerships and incorporated bookkeeping offices and CJVs subject to accounting firms are limited services limitations on minimum to CPAs licensed by Chinese size, aggregate number authorities. (15), and geographic scope (SEZs). Auditing reports are only valid if a Chinese CPA title is obtained. Taxation Mode 1: None Mode 2: None FULLY LIBERALIZED Mode 2: Unbound Mode 3: Only through CJVs, with Mode 3: None; wholly Mode 3: Through branch majority foreign ownership foreign-owned subsidiaries offices subject to limitations permitted. permitted on minimum size and by 2007. geographic scope (SEZs). Architecture and Mode 1: Unbound Mode 1: None for scheme FULLY LIBERALIZED engineering Mode 2: None design; otherwise, cooperation EXCEPT FOR MODE 1 Mode 3: Only through an EJV with Chinese professional RESTRICTIONS. or CJV. Registered in own organizations is required. Mode 3: Wholly foreign- country. Mode 3: Only through an EJV or owned subsidiaries CJV. Registered in own country permitted by 2006. and engaged in architecture/ engineering services in home country. Urban planning Mode 1: Unbound Mode 1: None for scheme FULLY LIBERALIZED (excluding Mode 2: None design; otherwise, cooperation EXCEPT FOR MODE 1 general urban Mode 3: Unbound with Chinese professional RESTRICTIONS. planning) organizations is required. Mode 3: Wholly foreign- Mode 3: Only through an EJV or owned subsidiaries CJV. permitted by 2006. Medical and Mode 1: Unbound Mode 1: None FULL FOREIGN OWNERSHIP dental services Mode 2: Unbound Mode 2: None NOT ALLOWED AND Mode 3: Only through an EJV Mode 3: Foreign majority NEEDS-BASED QUOTAS. or CJV with a quantitative ownership explicitly permitted. limitation based on a needs Not required to accept sole test and approval by the responsibility for foreign Ministry of Public Health exchange balance and profits and Ministry of Foreign and losses, but still subject to Trade and Economic quantitative limitations based Cooperation. CJV or EJV on a needs test. solely responsible for Mode 4: Licenses can be foreign exchange balance obtained from the Ministry of and profits and losses. Public Health, and a contract Majority of personnel is not required. must be Chinese. Mode 4: Foreign doctors can provide services for six months (may extend to a year) provided a license is obtained at provincial level and they are contracted by Chinese medical institutions. The Services Dimension of China's Accession to the WTO 123 TABLE 8.2 (Continued) Sector 1994 2001 2008 Computer and related services Consultancy Mode 1: Unbound Modes 1­3: FULLY LIBERALIZED services related Mode 2: None Mode 4: Qualifications: B.A. and to the Mode 3: None three years of experience. installation of Mode 4: Qualifications: computer B.A. and five years of hardware experience. Data processing and tabulation Time-sharing Software Mode 1: Unbound Mode 1: Unbound implementation Mode 2: None Mode 2: None Systems and Mode 3: None Mode 3: Only through JVs with software Mode 4: Qualifications: foreign majority ownership consulting B.A. and five years of permitted. Systems analysis experience. Mode 4: Qualifications: B.A. and three years of experience. Systems design Mode 1: Unbound Mode 1: None Programming Mode 2: None Mode 2: None Systems Mode 3: Through EJV only. Mode 3: Only through JVs with maintenance Mode 4: Qualifications: foreign majority ownership Data processing B.A. and five years of permitted. Input preparation experience. Mode 4: Qualifications: B.A. and three years of experience. Telecommunica- tions Value added No commitments Mode 1: Unclear By 2002 expansion in Mode 2: None geographic area and Mode 3: Through JVs with a foreign investment limit foreign investment limit of to 49 percent. 30 percent only in Shanghai, By 2003 no geographic Guangzhou, and Beijing. restriction and FOREIGN INVESTMENT LIMIT TO 50 PERCENT. Basic telecom- No commitments Mode 1: Unclear By 2002 expansion in munications: Mode 2: None geographic area and mobile voice Mode 3: Through JVs with a foreign investment limit and data foreign investment limit of to 35 percent. 25 percent only in and By 2004 FOREIGN between Shanghai, INVESTMENT LIMIT TO Guangzhou, and Beijing. 49 PERCENT. By 2006 no geographic restriction. Basic telecom- No commitments Mode 1: Unclear By 2004 through JVs with a munications: Mode 2: None foreign investment limit fixed-line Mode 3: Unbound of 25 percent only in services and between Shanghai, Guangzhou, and Beijing. By 2006 expansion in geographic area, and foreign investment limit to 35 percent. By 2007 no geographic restriction and FOREIGN INVESTMENT LIMIT TO 49 PERCENT. 124 China and the WTO TABLE 8.2 (Continued) Sector 1994 2001 2008 Construction Construction Mode 1: Unbound Mode 1: Unbound RESTRICTIONS ON and related Mode 2: None Mode 2: None BUSINESS SCOPE OF engineering Mode 3: Unbound Mode 3: Through JVs with foreign FULLY FOREIGN-OWNED majority ownership permitted ENTERPRISES. and only foreign-invested Mode 3: By 2004 fully construction projects. foreign-owned enterprises permitted but only in projects financed by foreign investment and/or grants, or by loans from international financial institutions, or those projects that are technically difficult for Chinese enterprises. Distribution Commission No commitments Mode 1: Unbound LIBERALIZED EXCEPT agents and Mode 2: None CROSS-BORDER DELIVERY wholesale Mode 3: Foreign-invested AND TWO PRODUCTS. trade, and a enterprises are permitted Mode 3: By 2002 through full range of to distribute their products JVs subject to restrictions subordinated manufactured in China. on products, to be services, phased out by 2006 including (except salt and tobacco). after-sales By 2003 foreign majority services. ownership allowed and no geographic or quantitative restrictions. Retailing and a No commitments Mode 1: Unbound except for CONTINUED full range of mail order. RESTRICTIONS ON LARGE subordinated Mode 2: None CHAIN STORES. services, Mode 3: Through JVs (not foreign Mode 3: By 2003 all including majority controlled) in five SEZs provincial capitals open. after-sales and eight cities subject to By 2004 no more services. quotas (e.g., four in Beijing geographic restrictions. and Shanghai), restrictions By 2006 no restrictions on products (not books, on products. Foreign newspapers, pharmaceuticals, majority control allowed pesticides, chemical except in chain stores fertilizers, etc.). with more than 30 outlets selling a range of products. Franchising No commitments Mode 1: None FULLY LIBERALIZED Mode 2: None BY 2004. Mode 3: Unbound Mode 3: By 2004 none. Educational and environmental services Educational Mode 1: Unbound Mode 1: None services, Mode 2: None Mode 2: None excluding Mode 3: Unbound Mode 3: Only through JVs with special Mode 4: Subject to licensing foreign majority ownership education from the State Bureau of permitted (national treatment: (e.g., military Foreign Experts (SBFE) unbound). and political) and State Education Mode 4: Subject to invitation or and national Commission (SEC), and employment by Chinese compulsory possession of M.A. and institution and possession of education professional title. B.A., two years of experience, and professional title. Environmental No commitments Mode 1: Unbound except for services consultation services. Mode 2: None Mode 3: Through JVs with foreign majority ownership permitted. The Services Dimension of China's Accession to the WTO 125 TABLE 8.2 (Continued) Sector 1994 2001 2008 Financial services Insurance Mode 1: Unbound Mode 1: Unbound except for By 2004, FULLY (except Mode 2: Unbound international maritime, aviation, LIBERALIZED EXCEPT statutory Mode 3: Through a branch or and transport insurance and 50 percent FOREIGN insurance) JV only in Shanghai, subject reinsurance, and certain types of OWNERSHIP LIMIT IN LIFE to minimum local and brokerage. INSURANCE. global asset and local Mode 2: None, but unbound for Mode 3: By 2003 no presence (as rep. office) brokerage. establishment restrictions requirements. Mode 3: Form of establishment: in non-life. non-life: through a branch or By 2004 no geographic JVs with 51 percent foreign restrictions. ownership; life: through JVs with By 2004 no restrictions on 50 percent foreign ownership. business scope Geographic limitation: only in By 2005 no cession five cities. requirement. Business scope: only selected forms of non­life insurance. Life only to individuals, not groups. Licenses: no quotas but subject to minimum asset and duration of establishment requirements. Upon accession, a 20 percent cession required of all lines of the primary risks for non-life, personal accident, and health insurance business with an appointed Chinese reinsurance company. Banking Mode 1: Unbound Mode 1: Unbound except for FULLY LIBERALIZED Mode 2: Unbound provision of data, advice, etc. BY 2006. Mode 3: Through a branch, Mode 2: None. Mode 3: Geographic subsidiary JV only in Mode 3: Geographic limitation: limitations phased out specified regions, subject to none for foreign currency gradually by 2006. minimum asset and local business, but local currency only Clients: local currency presence (as rep. office) in four cities. Interregional business with Chinese requirements; acceptance of supply of services permitted. enterprises by 2003 and deposits only from nonresi- Clients: only foreign currency all clients by 2006. dents in foreign currencies business. (with some exceptions) and Licenses: only prudential criteria. no loans to Chinese citizens. Securities No commitments Mode 1: Unbound except B share Mode 3: By 2004 49 percent business. foreign ownership in JVs Mode 2: None to conduct domestic Mode 3: Unbound, except rep. securities investment fund offices may become special management business, members of the Chinese Stock and through JVs with up Exchange (CSE)s, and through to 33 percent foreign JVs with up to 33 percent foreign ownership to underwrite A ownership to conduct domestic shares and underwrite and securities investment fund trade B and H shares, as management business. well as government and corporate debts, launching of funds. Tourism and travel-related services Hotels Mode 1: Unbound Mode 1: Unbound FULLY LIBERALIZED Mode 2: None Mode 2: None BY 2005. Mode 3: Through JVs subject Mode 3: Through JVs with foreign to needs test at central and majority ownership permitted. local levels. Travel agency No commitments Mode 1: Unbound FULLY LIBERALIZED and tour Mode 2: None BY 2007. operator Mode 3: Through JVs subject to geographic and business scope restrictions. 126 China and the WTO TABLE 8.2 (Continued) Sector 1994 2001 2008 Transport services MARITIME TRANSPORT International Mode 1: None Mode 1: None transport Mode 2: None Mode 2: None Mode 3: Unbound Mode 3: Through JVs subject to 49 percent foreign ownership limits to operate only a Chinese flag fleet Auxiliary Mode 1: Unbound Mode 1: Unbound services Mode 2: None Mode 2: None Mode 3: Through JVs only. Mode 3: Through JVs only with foreign majority ownership permitted. INTERNAL Mode 1: Only international As in 1994. wATERWAYS shipping in ports open to foreign vessels permitted. Mode 2: None Mode 3: Unbound AIR TRANSPORT Aircraft repair No commitments Mode 1: Unbound and Mode 2: None maintenance Mode 3: Through Chinese controlled JVs and subject to an economic needs test. Computer No commitments Mode 1: By connection with reservations Chinese computer reservation system, etc. Mode 2: None Mode 3: Unbound SPACE No commitments No commitments TRANSPORT RAIL TRANSPORT No commitments Mode 1: None FULLY LIBERALIZED BY 2007. Mode 2: None Mode 3: Majority Mode 3: Through JVs with a ownership by 2004. foreign ownership limit of 49 percent. ROAD TRANSPORT Mode 1: Unbound Mode 1: None FULLY LIBERALIZED BY 2004. (FREIGHT) Mode 2: None Mode 2: None Mode 3: Majority Mode 3: Through JVs subject Mode 3: Through JVs with a ownership by 2002. to an economic needs test. foreign ownership limit of 49 percent. SERVICES AUXILIARY TO ALL MODES OF TRANSPORT Commitments only for Mode 1: Unbound FULLY LIBERALIZED BY 2004. Storage and maritime transport, as Mode 2: None Mode 3: Majority warehousing above. Mode 3: Through JVs with a ownership by 2002. foreign ownership limit of 49 percent. Freight Commitments only for Mode 1: None FULLY LIBERALIZED BY 2005. forwarding maritime transport, as Mode 2: None Mode 3: Majority agency above. Mode 3: Through JVs with a ownership by 2002. services foreign ownership limit of 50 percent and subject to minimum capital requirements. JV joint venture; CJV contractual joint venture; EJV equity joint venture; SEZ special economic zone; CPA certified public accountant. Source: Compiled by the author from China's schedules of commitments under the GATS. The Services Dimension of China's Accession to the WTO 127 ownership will be permitted but subject to certain representative office). On accession, non­life insur- restrictions on business scope. ers are permitted to open a branch or joint venture with 51 percent foreign ownership, whereas life Distribution Services No commitments at all insurers are permitted 50 percent ownership of a were shown in the 1994 schedule. Substantial liber- joint venture with a partner of their choice. alization has already taken place, but restrictions Non­life insurers can provide "master policy" persist on establishment (only through joint insurance and insurance of large-scale commercial ventures), geographic scope (retailing in only five risks without geographic restrictions, and insur- special economic zones) and products sold (e.g., ance of enterprises abroad as well as property not books, newspapers, pharmaceuticals, and pesti- insurance, related liability insurance, and credit cides). By 2006 the sector will be largely open. Per- insurance of foreign-invested enterprises in five haps most strikingly, China has agreed to open up cities: Shanghai, Guangzhou, Dalian, Shenzhen, the whole logistical chain of related services, and Foshan. Life insurers are permitted to provide including inventory management; assembly, sort- individual (not group) insurance to foreigners and ing, and grading of bulk lots; breaking bulk lots and Chinese citizens in the same five cities. By 2004 all redistributing into smaller lots; delivery services; restrictions will disappear except the foreign own- refrigeration, storage, warehousing, and garage ership limit on life insurers. Licenses are to be services; sales promotion, marketing and advertis- awarded solely on the basis of prudential criteria, ing; and installation and after-sales services, includ- with no application of quantitative limitations or ing maintenance and repair and training services. economic needs tests. No WTO member has made such deep commit- Under the 1994 commitments, foreign banks ments in this sector. could only operate in specified regions, accept Some restrictions will remain. Salt and tobacco deposits only from nonresidents and only in for- are excluded from the scope of commitments on eign currencies (with some exceptions), and make commission agents and wholesalers. Furthermore, no loans to Chinese citizens. On accession, geo- majority foreign ownership will not be allowed in graphic and client limitations will be eliminated for retail chain stores that sell multiple products and the foreign currency business. Even though the different brands of products such as books, news- schedule states that on accession local currency papers, pharmaceuticals, and chemical fertilizers business will be allowed in four cities (Shanghai, and have more than 30 outlets. There are no com- Shenzhen, Tianjin, and Dalian), there seems to be a mitments on the cross-border supply of commis- binding restriction on clients that will only be sion agents' services and delivery of wholesale trade relaxed in two years. The entire banking sector will services. More interestingly, cross-border supply of be fully liberalized by 2006. As for insurance, retail services is allowed only through mail order, licenses are to be awarded solely on the basis of which presumably covers electronic commerce. prudential criteria, with no quantitative limitations or economic needs test applied. Educational Services Cross-border delivery was There were no commitments on securities in unbound in 1994, but it is now fully open. 1994.On accession,joint ventures with up to 33 per- Commercial presence was also unbound, but it now cent foreign ownership will be allowed to conduct can be established through joint ventures with for- business in domestic securities investment fund eign majority ownership permitted. National treat- management. By 2004 foreign ownership of such ment is not, however, guaranteed for foreign educa- ventures will be allowed to increase to 49 percent. tional institutions. Furthermore, joint ventures with up to 33 percent foreign ownership will be allowed to underwrite Financial Services The 1994 commitments speci- domestic equity issues and underwrite and trade in fied that insurance services could be supplied international equity and all corporate and govern- through a branch or joint venture only in Shanghai, ment debt issues. subject to conditions pertaining to minimum capi- tal and prior presence, globally (30 years as an Grandfather Provisions One of the key difficul- insurance company) and locally (three years as a ties in the accession negotiations arose because of 128 China and the WTO the presence in the Chinese insurance market of right to establish fully owned subsidiaries, then the firms that enjoyed better conditions than China latter can claim that the assurance contained in was prepared to guarantee to new entrants. In par- the footnote is not being respected.5 ticular, the U.S. life insurance firm AIG was estab- lished in Shanghai with full foreign ownership; Transport Services Liberal commitments had Allianz (German), Axa (French), and Manulife already been made in international maritime trans- (Canadian) were 51 percent foreign-owned. As port and certain supporting services, even in 1994. noted earlier, the best China was prepared to offer The major change is that the greater depth of com- new entrants was entry through joint ventures that mitments across the whole range of transport and were 50 percent foreign owned. Apparently this auxiliary services will significantly facilitate the asymmetry in itself was generally acceptable and provision of multimodal transport services.In road, led to the following grandfather provision in rail, and key auxiliary services, notably storage and China's schedule: warehousing and freight forwarding agency ser- vices, restrictive or no commitments will be The conditions of ownership, operation and replaced on accession by the requirement to enter scope of activities, as set out in the respective as joint ventures, and by 2007 there will be full lib- contractual or shareholder agreement or in a eralization. The exception will be hard rights in air license establishing or authorizing the opera- transport, which are excluded from the scope of the tion or supply of services by an existing foreign GATS. service supplier, will not be made more restric- tive than they exist as of the date of China's accession to the WTO. Do China's Commitments Promote Good Policy? The problem arose because of the implications for branching rights. The key element of China's The commitments just described imply a dramatic commitment was the guarantee that "internal loss of discretion in policymaking. China has prom- branching for an insurance firm will be permitted ised to give up over the next few years much of the consistent with the phase out of geographic restric- freedom it had to restrict new entry and foreign tions." In effect, AIG would be able to expand ownership, to discriminate between trading part- operations through branches of its fully owned ners and in favor of its own firms,and,more broadly, subsidiary in China, whereas other new firms could to change its mind.The central argument of this sec- only do so through branches of their 50 percent­ tion is that, for the most part, the discipline implied owned joint ventures. This situation was apparently by the commitments is desirable. However, the per- not acceptable to the European Union. The even- sistence of restrictions on foreign ownership may tual compromise was a rather convoluted footnote deprive China of the full benefits of foreign invest- in the schedule, which states: ment, and the scheduled sequence of liberalization across regions may accentuate regional inequalities. Any further authorization provided to foreign Moreover,as discussed in the next major section,the insurers after accession under more favorable size and distribution of the gains from liberalization conditions than those contained in this schedule will depend on how China uses the freedom it still (including the extension of grandfathered invest- has to strengthen its regulatory framework and ments through branching, sub-branching or any choose an appropriate sequence of reform. Given other legal form), will be made available to other the limited empirical evidence and the uniqueness foreign service suppliers which so requested. of China's experience, these issues are at this stage This solution is messy and could lead to a not much more than plausible hypotheses that dispute. If AIG is denied the right to branch from its define an agenda for future research. fully owned subsidiary, the United States can claim that the assurance contained in the grandfather Eliminating Discretion and internal branching commitments is not being respected. If AIG is allowed the right to establish In many areas the Chinese government has been branches and the European Union is denied the reluctant to liberalize immediately. Some of the The Services Dimension of China's Accession to the WTO 129 protection is probably a consequence of political has given up the right to offer preferential treat- economic pressures from vested interests. In some ment to domestic enterprises through any measure cases, the government evidently has felt the need to affecting the supply of services. Virtually no scope protect the incumbent suppliers from competition exists for discriminatory taxation, and discrimina- because of infant industry­type arguments, to tory subsidies can be awarded only in aviation, facilitate an "orderly exit," or to allow complemen- audiovisual, and medical services--and there, too, tary reforms sufficient time--all of which are argu- only to the extent that such subsidies already exist. ments for temporary protection. But protection The other instances of discriminatory provisions once granted can be difficult to remove. The failure are the following: all legal representatives are of infant industry policies in the past and the innu- required to be resident in China for no less than six merable examples of perpetual infancy (or senility) months each year; a majority of doctors in joint are attributable in part to the inability of the venture hospitals must be Chinese; the existing reg- Chinese government to commit itself to liberalize istered capital requirements for joint venture con- at some future date and thus to confront incum- struction enterprises are slightly different from bents with a credible deadline. The binding com- those of domestic enterprises; joint venture travel mitments under the GATS to provide market access agencies are not allowed to provide services to by a precise future date should play a valuable role Chinese traveling abroad; and foreign insurers are in overcoming the credibility problem. Failure to subject to a 20-percent cession requirement with a honor these commitments creates an obligation to Chinese reinsurance company, to be phased out in compensate those who are deprived of benefits, four years. making the commitment more credible than a Although the scope for explicit discrimination mere announcement of liberalizing intent in the has shrunk, the nature of measures affecting trade national context. The price in terms of the loss of in services does offer considerable scope for flexibility would seem to be worth paying. implicit discrimination. First, there is the possibil- It is also worth emphasizing that GATS commit- ity of variable treatment through domestic regula- ments represent only a ceiling on protection and tions such as licensing, recognition of qualifica- not a floor. China has given up the freedom to liber- tions, and various other technical regulations. Such alize more slowly than specified in its GATS com- discretion is likely to be constrained, but probably mitments, but it still has the freedom to move faster. not eliminated, by the assurance that licenses in financial services will be issued solely on the basis of prudential criteria and by the requirement for Eliminating Discrimination transparency of licensing criteria and decision Discrimination either between trading partners or processes in other areas. Furthermore, Article VII of in favor of domestic firms results in economically the GATS, which deals specifically with recognition costly distortions, and so its prohibition is generally agreements pertaining to educational qualifica- desirable. The two pillars of nondiscrimination tions, licenses, and such, strikes a delicate balance under the GATS are the MFN and national treat- by allowing such agreements, provided they are not ment obligations, but both allow exceptions to be used as a means of discrimination, and third coun- listed. China's MFN exemptions are relatively nar- tries have an opportunity to accede or demonstrate row. The only one listed applies to international equivalence. maritime transport: to permit cargo-sharing agree- Second, China has retained the freedom to ments only with certain countries and to allow impose explicit quantitative restrictions on the joint ventures and wholly owned shipping sub- number of providers in legal, medical, and retailing sidiaries to be formed on the basis of bilateral services. De facto quotas also may exist in some agreements. In any case, because China has com- areas, such as those imposed by the scarcity of radio mitted to full market access and national treatment spectrum needed for the provision of mobile on cross-border supply, there seems to be little telecommunications services and the scarcity of scope for discriminatory cargo sharing. space needed for department stores or airports There are also surprisingly few limitations on in a city. Apart from the general MFN obligation, national treatment, and so, for the most part, China no specific rules are in place to ensure the 130 China and the WTO nondiscriminatory allocation of quotas in services guaranteeing the rights of incumbents. Insofar as (WTO rules are, however, in place for goods). In they provide the benefits of security to investors who the past, this lack of rules was not a major issue are already present in the market rather than to new because commitments reflected the status quo, and investors, they may not do enough to make markets quotas, particularly those for service suppliers, were morecontestable.Newentrantsmayevenbeplacedat descriptions of the existing market structure. But a competitive disadvantage insofar as differences in when genuine liberalizing commitments are made, ownership and legal form affect firm performance. like those by China, the nondiscriminatory alloca- Forexample,largerequitysharesmaymakeiteasierto tion of quotas is bound to be an important issue. exerciseeffectivecontrolovertheoperationsof afirm The instinctive candidate for a nondiscriminatory and ensure efficient production, and the marginal rule is an auction, which also has the virtue of cost of providing a service through a branch may be transferring quota rents to the government. It lower than through a subsidiary.Fortunately,the fact remains to be seen whether the Chinese regulators that several foreign and domestic firms are already can resist the temptation to resort to other, more competing in the market may limit the impact of discretionary methods of allocating quotas. thesegrandfatherprovisions. Adhering to the principle of nondiscrimination is important, even though it may seem at first sight Eliminating Barriers to Entry that there is no real cost to granting preferential treatment in services. Because the protective instru- Apart from the few instances just noted, China ment is often a restrictive regulation that does not has chosen not to limit the number of service generate revenue (or rents), there is no cost to providers. The only barrier to new entry is the granting preferential access because there is no rev- requirement to enter as a joint venture, which is enue (or rents) to lose. When the protective instru- discussed in the next section. But two questions ment is a quota, the implications of preferences arise: Is it good policy to allow unrestricted entry in depend on who appropriates the rents. Where rents all service sectors? Will China's commitments are appropriated by exporters or dissipated, prefer- deprive the government of an important policy ential liberalization is again necessarily welfare- instrument? One reason for entry restrictions enhancing for the importing country. might be the existence of significant economies of Preferences in services may, however, impose scale. For example, networks incur substantial fixed other, more subtle costs. The cost of trade diversion costs, and competitive entry could lead to ineffi- may well be the establishment of poorer-quality cient network duplication.7 providers, not just in terms of poorer-quality For several reasons, however, entry restrictions services but also in terms of smaller social benefits are increasingly difficult to defend in principle in in the form of knowledge and technological the face of technological change and in the face of spillovers. Furthermore, the greater importance of mounting evidence that competition works. First, sunk costs in some service sectors, ranging from entry restrictions change the nature of interaction basic telecommunications to financial services, sug- between incumbents and may well make collusion gests that preferential liberalization may have more more likely. Second, such restrictions dampen the durable consequences than in the case of goods. For impact of competition on productive efficiency. example, allowing the second-best provider to Third, the regulator is usually not better placed establish a business may mean that a country is than the competitive process to determine the opti- stuck with such a provider even when it subse- mal number of firms in the market, especially given quently liberalizes on an MFN basis.6 the difficulty of obtaining information about the These considerations suggest that China would cost structure of firms and other sources of regula- do well to resist the reported pressure to grant pref- tory failure. Fourth, technological advances have erential access to the service suppliers of certain significantly lowered network costs, even in a sector WTO members. Such pressure resulted in, for exam- like fixed-line telecommunications, and vertical ple, the grandfather provisions in China's commit- separation (e.g., through network unbundling) has ments on insurance services, described earlier in this widened the scope for competitive entry (Smith chapter. These provisions reflect an emphasis on 1995). Inefficiencies introduced by the duplication The Services Dimension of China's Accession to the WTO 131 of networks may be small compared with the oper- can learn by collaborating. As with all such argu- ational inefficiencies that can result from a lack of ments, it is difficult to judge whether the current competitive pressure.8 On this basis, it would seem costs are likely to be offset by the eventual benefits. that there is little reason to worry about excessive Third is the adjustment cost argument: an immedi- entry; rather, the priority should be eliminating the ate transfer of control could lead to drastic cuts in barriers to entry where they remain (e.g., in med- surplus labor, which gradual reductions in owner- ical services and retailing). ship help prevent. A central issue is whether it is possible to address these adjustment costs through direct support to the affected factors rather than by Eliminating Restrictions on Foreign Ownership staggered liberalization. Finally, the most impor- The most important restriction on foreign presence tant reason is probably a purely political reluctance in China is the requirement to enter as a joint ven- to allow foreign control of an essential service. ture, often with limits on the extent of foreign own- Again, these political concerns should diminish if ership. In most areas these limits on ownership are multiple competing foreign firms, not just one for- being gradually phased out, but in some areas such eign monopolist, are providing the service. as telecommunications and life insurance they will remain even after all the liberalization commit- Eliminating Restrictions Gradually Across Regions ments have been phased out. What is the rationale for such restrictions, and what are their implica- A remarkable feature of China's dramatic expan- tions? Furthermore, what are the consequences of sion in international trade over the past two the manner in which they are being phased out? decades has been the concentration of export- Joint ventures may, of course, be the preferred oriented industries in coastal regions. Three coastal choice of the foreign investor if, for example, local provinces (Guangdong, Jiangsu, and Fujian) and firms have specific assets that can be accessed only Shanghai have been the main recipients of through collaboration. However, binding owner- outward-oriented foreign investment, with the ship restrictions may adversely affect firm perform- remaining portion going either to other coastal ance, because the incentive to undertake costly provinces or to regions adjoining coastal areas. transfers of technology and improvements in man- Thus, although China's economic reforms have agement is related to the expected gains, which in been successful in raising living standards for a turn are related to an owner's share of the profits considerable share of the population, a large num- of. Moreover, changes in the permissible share of ber of Chinese people in inland provinces have ownership in the vicinity of 50 percent (e.g., from benefited substantially less. 49 percent to 51 percent) may have a particularly A factor responsible for coastal agglomeration large impact on performance as a firm obtains has been the inefficiencies in China's internal serv- fuller control of its operations and has greater free- ice systems, most obviously transport and telecom- dom to make the changes it deems necessary. munications, but also financial, business, and other At this stage the evidence available is not suffi- services. Transport infrastructure disparities cient to evaluate the magnitude of these effects, but between the coastal and inland provinces narrowed the plausibility of their existence raises the question considerably after policies were adopted in 1990 of why the Chinese government is willing to bear aimed at promoting more regionally balanced eco- such costs. Four reasons are possible. First, limi- nomic development. However, there is evidence to tations on ownership may seek to balance the suggest that it is not the availability of transport efficiency enhancing and the rent-appropriating infrastructure per se that has precluded inland aspects of foreign investment. However, rent appro- provinces from actively participating in foreign priation could to some extent be prevented by ex trade. Rather, the inadequacies associated with ante auctions of equity or ex post taxation of prof- transport services have been the more serious bind- its.9 A more basic question is why rent-generating ing constraints on better integrating China's hinter- restrictions on competition continue to exist. Sec- land economy.10 ond is the "infant entrepreneur" argument: foreign- The geographic limitations on liberalization ers are induced to form EJVs so that local investors commitments in certain sectors could well 132 China and the WTO accentuate these interregional inequalities, even Industry (MII) has been the regulator of telecom- though the limitations do not apply to transport munications and other information technology- services and most are to be phased out over time. related products and services, including data The existing enclaves of development are likely to communications, wireless communications, elec- see faster improvements in service quality from tronics, computers, the Internet, and software. early liberalization, and these improvements will In August 2001 the high-profile State Council cause even more economic activity to gravitate to Information Office (SCIO) was formed with a man- these areas. Eventually, though, liberalization of date to develop and implement information policy services in the hinterland may lead to greater diffu- in China (see chapter 10). Similar steps may need sion of activity. Yet insofar as sunk costs are impor- to be taken in a variety of other network services, tant, the earlier agglomeration is unlikely to be including transport (terminals and infrastructure) completely reversible and inequalities may persist. and energy services (distribution networks). The importance of these considerations needs to be The creation of a regulator is only a first step. established through further empirical research, but Persuading the dominant interest groups to con- at least intuitively would seem to strengthen the cede control is fraught with difficulty. For example, case for eliminating the geographic restrictions in India a conflict between the Department of simultaneously rather than sequentially. Telecommunications (DOT) and the Telecommu- nications Regulatory Authority of India (TRAI), as it was initially constituted, hampered progress What Are the Priorities for toward an efficient telecom infrastructure.12 Absent Complementary Domestic Reform? a truly independent regulator empowered to rebal- If China is to make the most of the liberalization it ance tariffs,enforce fair interconnection agreements, is now committed to undertake, improving the reg- and ensure rapid, equitable issuance of radio spec- ulatory framework is critical. Regulation in services, trum, the benefits of a sector opened to allow pri- as in goods, arises essentially from the need to rem- vate participation and foreign investment could be edy market failure--attributable to the problems of significantly limited. natural monopoly and inadequate consumer In certain market segments it may not be possi- information--and to ensure equitable access. ble to create conditions for effective competition in the supply of certain telecommunications, trans- port, and financial services, even if all barriers to Efficient Regulation: Making Competition Work entry are eliminated. Two related factors are likely to The first regulatory priority arises in the so-called be involved. First, unlike for goods, national markets locational services. These markets are concentrated are often segmented from the international market because of the large fixed costs required to create by the infeasibility of cross-border delivery. Second, specialized distribution networks: roads and rails changing technologies may have reduced the opti- for land transport, cables and satellites for commu- mal scale of operation as well as the sunk costs in nications, and pipes for sewage and energy distri- these sectors, but not enough for small markets to bution (UNCTAD and World Bank 1994). Unless sustain competitive market structures. Some form the appropriate regulatory mechanisms are put in of final price regulation may therefore be needed. place, the incumbent can frustrate competition by denying rivals access to essential facilities such as Regulation to Remedy Inadequate Consumer distribution networks and terminals. Information China has accepted the regulatory principles specified in the Telecommunications Reference In China's increasingly open markets, priority must Paper. Thus it has committed to instituting an be given to strengthening the quality of prudential independent regulator for basic telecommunica- regulation in intermediation and knowledge-based tions services to ensure that the incumbent supplier services, where consumers have difficulty securing does not undermine market access by charging pro- full information about the quality of service they hibitive rates for interconnection to its established are buying (UNCTAD and World Bank 1994). networks.11 Since 1998 the Ministry of Information For example, consumers cannot easily assess the The Services Dimension of China's Accession to the WTO 133 competence of professionals such as doctors and ization were not determined by the market, but lawyers, the safety of transport services, or the were set administratively and then kept artificially soundness of banks and insurance companies. low for certain categories of end users or types of When such information is costly to obtain and dis- service products. Liberalization threatens these seminate and consumers have similar preferences arrangements. The elimination of restrictions on about the relevant attributes of the service supplier, entry implies an end to cross-subsidization, because the regulation of entry and operations in a sector it is no longer possible for firms to make extranor- could increase social welfare. However, the estab- mal profits in certain market segments.And privati- lishment of institutions competent to regulate well zation could mean the end of government support. is a serious challenge, as revealed by the difficulties Reform programs can accommodate universal in the financial sectors elsewhere--not only in var- service obligations by imposing this requirement ious developing countries but also in the United on new entrants in a nondiscriminatory way. Thus States, Sweden, and Finland in the 1980s and 1990s. in several countries such obligations have been part A separate problem is that domestic regulations of the license conditions for new entrants into fixed to deal with market failure may themselves become network telephony and transport. Often, however, impediments to competition and trade as a result subsidies have proved more successful than direct of differences across jurisdictions in technical stan- regulation in ensuring universal access (Estache, dards, prudential regulations, and qualification Foster, and Wodon 2001). The Chilean government requirements in professional, financial, and numer- adopted a scheme that permitted it to leverage over ous other services. For example, China requires for- $2 million in public funds into $40 million in private eign doctors to obtain a license from the Ministry investment. As a result, telephones were installed in of Public Health and foreign accountants to pass 1,000 localities at about 10 percent of the cost of the Chinese national certified public accountant direct public provision. Public subsidies can also be examination. In many instances the impact on directed to the consumer rather than the provider trade is an incidental consequence of the pursuit of (Cowhey and Klimenko 1999). The choice of a legitimate objective, but in some instances reg- appropriate instrument will have to be made on a ulation can be a particularly attractive means of sector-by-sector basis. protecting domestic suppliers from foreign compe- tition. Multilateral trade rules on domestic regula- Addressing Adjustment Costs tions may contribute to domestic reform by help- ing to sift the legitimate from the protectionist. To Different modes of supply have different effects on this end, negotiations are under way to develop factor markets. Cross-border trade and consump- GATS rules for domestic regulations. The core of tion abroad resemble the goods trade in their these disciplines may well be the so-called necessity implications. The impact of the movement of fac- test, which seeks to establish whether a particular tors depends critically on whether they are substi- regulation is more burdensome than necessary to tutes or complements for domestic factor services. achieve a legitimate objective. Given the structure of factor prices in China, liber- alization would typically lead to an inflow of capital and skilled workers. Such inflows would tend to be Regulation to Ensure Universal Service to the advantage of the unskilled poor--increasing Opening up essential services to foreign or domestic employment opportunities and wages.13 Interest- competition could have an adverse effect on the ingly, it has been shown that even when foreigners poor--a point that is often cited as a reason for the compete with local skilled workers in a service sec- persistence of public monopolies. A more efficient tor, the productivity boost to the sector from allow- solution is to have regulations with a social purpose. ing foreigners access could lead to an increase in the Where China is a relatively inefficient producer demand for domestic skilled workers--that is, the of a service, liberalization and the resulting foreign scale effect could outweigh the substitution effect competition are likely to lead to a decline in domes- (Markusen, Rutherford, and Tarr 2000). Given tic prices and improvements in quality. However, these predictions, why are workers in China some- there is a twist. Frequently, the prices before liberal- times skeptical about the benefits of liberalization? 134 China and the WTO One concern is the possible reduction in employ- (2001), in a study of telecommunications reform in ment in the formerly public monopolies that have Asia,Africa, and Latin America, found that although frequently employed surplus labor. each element of reform has a positive impact on per- Evidence suggests that such pessimism may formance,the effect of each is magnified when others not always be justified. For example, some develop- are also implemented. Moreover, privatizing the ing countries have managed to maintain or even incumbent after introducing competition (the increase employment in their liberalized telecom- Chinese route) is likely to lead to a higher level of munications sectors. Because China still has a low mainline penetration than the opposite sequence. teledensity outside the big cities (about 10 lines per Although China has chosen the preferred 100 people), much of telecom investment is being sequence in telecommunications, it must still make directed toward building wire line and mobile net- important choices in other areas, such as postal and works that are labor-intensive and that are probably air transport services. Introducing privatization helping to maintain or raise employment levels. The before introducing competition may create a privi- introduction of competition should help.For exam- leged incumbent that has a first-mover advantage ple, Petrazzini and Lovelock (1996) found in a study in the market. Among other things, the incumbent of 26 Latin American and Asian economies that tele- may be able to make certain strategic decisions or com markets with competition were the only ones indulge in lobbying to affect the eventual form of that consistently increased employment levels,while competition.14 For example, in South Africa foreign two-thirds of the countries with monopolies saw and domestic shareholders in the privatized considerable declines in their telecom work force. telecommunications monopoly managed to per- Despite these optimistic projections, it is likely that suade the government to allow only one new reform programs will require complementary poli- entrant rather than the planned two. Regulating the cies to mitigate any social and economic costs of terms of access to essential facilities for new adjustments in factor markets. entrants may also be more difficult in an arm's length relationship with a private provider whose costs are difficult to observe than with a public Sequencing Regulatory Reform and Trade and provider whose cost information may be easier Investment Liberalization to access. Entry may be easier on symmetric Regulatory improvements take time. Changes in the terms with an inefficient public incumbent than patterns of competition and ownership can be on asymmetric terms with an efficient private implemented instantaneously in principle, but incumbent. China, like many other countries, has chosen to introduce changes gradually. One question that Financial Services In financial services, interna- arises in this context is: How does the impact of dif- tionalization raises several concerns: the threat to ferent elements of reform depend on the extent to the survival of local banks and financial companies; which other reforms have been implemented--that the loss of monetary autonomy; and the increased is, how does the interaction of the different elements volatility of capital flows. Many of these concerns of policy reform affect performance? The other, do not relate just to the internationalization of more subtle question is:Does the sequence of reforms financial services, but also to the processes of finan- have transitory and permanent effects on perform- cial deregulation and capital account liberalization. ance? Some preliminary observations on these ques- However, the extent of the benefits and costs of tions, which are relevant to the design of transition internationalization depend to a great extent on strategies, are presented in the sections that follow. how it is phased in with these other two types of financial reform, and, in particular, the strengthen- Telecommunications Figure 8.2 depicts the ing of prudential regulation and supervision. sequence of telecommunications liberalization Many countries that have had successful experi- in nine Asian countries. China is among the few that ences opening up to foreign financial firms (Brazil, have allowed some degree of competition (in long- Chile, Hungary, Ireland, Poland, Portugal, and distance services) before allowing a change of owner- Spain, among others) also engaged in a process of ship in the incumbent supplier and creating an inde- domestic deregulation and consequently reaped pendent regulator. Fink, Mattoo, and Rathindran substantial gains (World Bank 2001).The experience The Services Dimension of China's Accession to the WTO 135 FIGURE 8.2 Sequence of Telecommunications Reform in Nine Asian Countries, 1989­99 LD China 1 2 Privatization Mobile Fixed competition Regulation Local India 8 14 19 19 20 19% 23% Indonesia 1 3 4 6 7 10% 20% 29% ILD LD Local Korea, 1 2 5 Republic of 25% ILD Local, LD Malaysia 2 4 7 8 12% Pakistan 2 3 100% ILD Local, LD Philippines 2 5 11% 17% Local Singapore 1 2 34% Local, LD Sri Lanka 1 2 3 4 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 Notes: The percentages indicate the share of private equity ownership in the incumbent operator. Local, LD, and ILD refer to the local, long distance, and international fixed-line service segments, respectively. The number(s) in the mobile row corresponds to the number of cellular operators in the country. "Regulation" only captures the existence of a separate regulatory agency. Source: World Bank/ITU telecommunications policy database. of the countries acceding to the European Union regulatory framework for the provision of financial suggests that internationalization and domestic services. It also added to the credibility of rules. deregulation can be mutually reinforcing. Most of these considerations are relevant to Increased foreign entry bolstered the financial China. As discussed by Bhattasali in chapter 11, sector framework by creating a constituency for on-performing loans account for a large fraction of improved regulation and supervision, better disclo- the assets of the big four banks, which account for sure rules, and improvements in the legal and more than two-thirds of the assets of the banking 136 China and the WTO sector. The big four are already thought to be insol- Conclusions vent (Fitch and Moody's). One of the main reasons China's GATS commitments represent the most for the many bad loans is that interest rates are still radical services reform program negotiated in the controlled by the People's Bank of China (PBC). WTO. China has promised to eliminate over the They are unusually low to make it easier for state- next few years most restrictions on foreign entry owned enterprises to borrow funds at well below and ownership, as well as most forms of discrimi- the market-clearing equilibrium rate of interest.15 nation against foreign firms. Trading partners are This situation has been responsible for a huge naturally interested in the market access dimension transfer of wealth from individual savers to state- of these commitments and are keen to see them owned enterprises. Two-thirds of credit resources fully implemented.17 However, realizing the gains went to state-owned enterprises that generate only from, and perhaps even the sustainability of, liber- one-third of industrial output. The official target of alization will require the implementation of com- 2003 for interest rate liberalization has already been plementary regulatory reform and the appropriate pushed back to 2005 unofficially for fear that the sequencing of reforms. ensuing competition for deposits will drive busi- Although China's commitments, for the most ness away from the four large banks, which would part, promote good policy, two elements raise some then collapse, leading to a serious banking crisis concern. First, restrictions on foreign ownership-- (Business China, May 7, 2001). temporary in most sectors, but more durable in Although the two reform processes (interna- telecommunications and life insurance--may tionalization and domestic financial deregulation) dampen the incentives for foreign investors to are mutually reinforcing, they are not sufficient in improve firm performance. The rationale for these themselves. More than in other sectors, the gains restrictions merits greater scrutiny. Second, initial and costs of financial reform depend on the regula- restrictions on the geographic scope of services tory and supervisory framework (Barth, Caprio, liberalization could encourage the further agglom- and Levine 2001). Experience shows that it is vital eration of economic activity in certain regions-- to strengthen the supporting institutional frame- indeed, to an extent that is unlikely to be reversed work in parallel with domestic deregulation and completely by subsequent country-wide liberaliza- internationalization. In the absence of such tion. It may therefore be worth examining whether strengthening, foreign entry may entail risks. For- these restrictions could be phased out more quickly. eign bank entry can destabilize local banks by tak- This chapter has identified three main priorities ing away the lowest-risk business--including large for complementary reform. First, improved pru- exporting firms--leaving local banks to venture dential regulation and measures to deal with the further out on the risk frontier. large burden of nonperforming loans by state The benefits of a supportive institutional frame- banks are necessary to realize the benefits of liber- work are even more obvious when it comes to cap- alization in financial services. Second, in basic ital account liberalization.16 Experiences in recent telecommunications and other network-based years, most recently in Asia, have shown that services meaningful liberalization will be difficult achieving the potential gains, and avoiding the to achieve without strengthened pro-competitive risks, of capital account liberalization depend to a regulation. Finally, in virtually every important great extent on whether domestic institutions and service sector, an effective universal service policy prudential authorities have developed sufficiently must be put in place because liberalization alone to ensure that foreign finance will be channeled in will not deliver improved access for the poor. productive directions (Eichengreen 2001). The Services Dimension of China's Accession to the WTO 137 Appendix: Structure of China's Market Access Commitments TABLE 8.A.1 Mode 1 Mode 2 Mode 3 Mode 4 Sector F P N F P N F P N F P N Business services Legal services 0 1 0 0 1 0 0 1 0 0 1 0 Accounting/auditing/ 1 0 0 1 0 0 0 1 0 0 1 0 bookkeeping Architectural services 0 1 0 0 1 0 0 1 0 0 1 0 Medical and dental services 1 0 0 1 0 0 0 1 0 0 1 0 Data processing services 0 1 0 0 1 0 0 1 0 0 1 0 RandD services (natural 0 1 0 0 1 0 0 1 0 0 1 0 sciences) Advertising services 1 0 0 1 0 0 0 1 0 0 1 0 Management consulting 0 1 0 0 1 0 0 1 0 0 1 0 services Communication services Courier services 0 1 0 0 1 0 0 1 0 0 1 0 Voice telephone services 0 1 0 0 1 0 0 1 0 0 1 0 Private leased circuit services 0 1 0 0 1 0 0 1 0 0 1 0 Electronic mail 0 1 0 1 0 0 0 1 0 0 1 0 Online info and database 0 1 0 1 0 0 0 1 0 0 1 0 retrieval Audiovisual services 0 1 0 0 1 0 0 1 0 0 1 0 Construction, engineering Construction work (building) 0 0 1 0 1 0 0 1 0 0 1 0 Construction work (civil engin.) 0 0 1 0 1 0 0 1 0 0 1 0 Distribution Wholesale trade 0 1 0 0 1 0 0 1 0 0 1 0 Retailing services 0 1 0 0 1 0 0 1 0 0 1 0 Educational services Secondary education 0 0 1 0 1 0 0 1 0 0 1 0 Adult education 0 0 1 0 1 0 0 1 0 0 1 0 Environmental services Sewage services 0 1 0 0 1 0 0 1 0 0 1 0 Refuse disposal 0 1 0 0 1 0 0 1 0 0 1 0 Financial services Non­life insurance 0 1 0 0 1 0 0 1 0 0 1 0 Acceptance of deposits 0 1 0 0 1 0 0 1 0 0 1 0 Lending of all types 0 1 0 0 1 0 0 1 0 0 1 0 Trading in securities 0 1 0 1 0 0 0 1 0 0 1 0 Health-related, social services Hospital services Social services Tourism services Hotels and restaurants 1 0 0 1 0 0 0 1 0 0 1 0 Travel agencies 1 0 0 1 0 0 0 1 0 0 1 0 Recreational services Entertainment services News agency services Transport services Maritime (freight) 0 1 0 0 1 0 0 1 0 0 1 0 Rail (passenger) Rail (freight) 1 0 0 1 0 0 0 1 0 0 1 0 Road (passenger) Road (freight) 1 0 0 1 0 0 0 1 0 0 1 0 Total number 7 20 4 10 21 0 0 31 0 0 31 0 (31 subsectors) Percent 23 64 13 32 68 0 0 100 0 0 100 0 F full commitment; P partial commitment; N no commitment. Sources: WTO services database. 138 China and the WTO Notes 11. Several countries have found it difficult to create open, competitive telecommunications sectors because of a weak regu- 1. See, for example, chapter 9 by Luo and Findlay, chapter 10 latory environment. Poland opened up its telecommunications by Pangestu and Mrongowius, and chapter 11 by Bhattasali. sector to private competition as early as 1990. There was a rush 2. The only explicit sectoral exclusion from the GATS is cer- to invest, and about 200 licenses were awarded in the first six tain "hard" rights in the aviation sector. years of the newly liberalized regime. However, the dominant 3. Note that a GATS scheduling rule requires that discrimi- state operator, operating in a weak regulatory system, limited natory quantitative restrictions be listed as market access restric- access to its network and benefited from unequal terms for rev- tions even though they are also inconsistent with national enue sharing. By 1996 only 12 of the 200 licenses were still being treatment. used by the few competitive operators to survive. 4. The commitment on mode 1 indicates a cross reference to 12. The Indian government announced a new telecommuni- mode 3, the meaning of which is not clear. cations policy on March 26, 1999, that addressed several of these 5. It was reported in the Financial Times of December 6, key outstanding issues. 2001, that a compromise had been reached on this issue. Appar- 13. Because the poor are likely to be unskilled, the question ently, AIG would be given permission to open two more arises as to which service sectors are likely to employ them? Data 100 percent­owned branches, but would thereafter have to abide on the skill composition of the work force in service sectors are by the same 50 percent rule as all other foreign insurers in available only at a rather aggregate level. Still, a certain pattern the Chinese market. The deal was seen as consistent with WTO can be inferred. Construction, distribution, and personal rules because the two extra licenses were to be awarded to AIG services tend to be unskilled labor­intensive, whereas commu- before China entered the WTO on December 11, 2001. The nications, financial, and business services tend to be skilled licenses were to be granted for the cities of Suzhou, near labor­intensive. Shanghai, and Beijing, the capital. 14. A public sector provider could also behave in this way, 6. Where this happens will depend on the importance of but presumably if the government's objective is liberalization, it sunk costs relative to differences in costs and quality. is somewhat easier to draw along a public provider. 7. One such possibility is in the "nonsustainability" of a nat- 15. In 1995 savings deposits received a paltry 3.15 percent in ural monopoly. This situation could arise, for example, under interest, while loans for working capital were made at only some natural monopoly cost conditions, when there exist no 11 percent--they should have been made at about 21 percent prices that will not attract entry, even though supply by a single after factoring in the rate of inflation. The current one-year rate firm is efficient. Armstrong, Cowan, and Vickers (1994, p. 106) on savings deposits is about 2.25 percent. conclude: "Notwithstanding the logical possibility of this hap- 16. Because China has made only limited commitments on pening, we are doubtful whether it provides a good case for cross-border trade in financial services, its GATS commitments entry restrictions in the utility industries, which are not for the do not require it to allow a high degree of capital mobility, most part remotely contestable and where there is little evidence except insofar as capital inflows are required to establish com- that cost conditions give rise to non-sustainability." mercial presence. 8. Interesting evidence in this context is available from the 17. In fact, in the WTO's Services Council the United States Indian telecommunications sector. Das (2000) estimates a has already raised the issue of whether China's current rules for frontier multiproduct cost function for the incumbent fixed- express delivery services and branching by non­life insurance line operator, covering 25 years from 1969 to 1994. The study companies conform to its GATS commitments (Inside US Trade, finds the existence of very high economies of both scale and March 29, 2002). scope in the technology used--the parameter estimates even suggest that telecommunications in India is a natural monop- oly. However, the incumbent operator displays great ineffi- ciency, leading to a 26 percent increase in the operator's cost of References production. Based on these findings, Das concludes that India's market liberalization program, started in the mid-1990s, is jus- The word processed describes informally reproduced works that tified, but he argues that there may be a need to regulate entry may not be commonly available through libraries. in order to reduce the unnecessary duplication of common Armstrong, M., S. Cowan, and J. Vickers. 1994. Regulatory costs. Moreover, with continued improvements in technology, Reform: Economic Analysis and British Experience. Cambridge, the fixed costs of entrants are likely to fall, reducing losses Mass.: MIT Press. of scale economies and thus increasing the costs of entry Barth, J. R., G. Caprio Jr., and R. Levine. 2001."Bank Regulations restrictions. and Supervision: What Works Best?" Paper presented at 9. The fear of creating a disincentive for investors might be a Thirteenth Annual World Bank Conference on Development reason to refrain from taxation. Economics, April. 10. For example, although there has been a significant Cowhey, P., and M. M. Klimenko. 1999. "The WTO Agreement increase in the volume of container traffic in China since 1990, and Telecommunication Policy Reforms." Draft report for the increase is largely confined to coastal regions and associated the World Bank. Graduate School of International Relations with the ocean-going leg of travel. Container traffic in inland and Pacific Studies, University of California at San Diego, areas is much less, with no significant change in the percentage March. of seaborne containers traveling beyond port cities and coastal Das, N. 2000. "Technology, Efficiency and Sustainability of provinces. Truck rates for moving a container 500 kilometers Competition in the Indian Telecommunications Sector." inland are estimated to be about three times more, and the trip Information Economics and Policy 12: 133­54. duration five times longer, than they would be in Europe or Eichengreen, B. 2001. "Capital Account Liberalization: What Do United States. The intermodal transport system was found to be the Cross-Country Studies Tell Us?" World Bank Economic poorly integrated, with no streamlined procedures to support Review 15 (3): 341­65. the continual movement of containers between the coast and Estache, A., Vivian Foster, and Quentin Wodon. 2001. "Account- inlands. ing for Poverty in Infrastructure Reform: Learning from The Services Dimension of China's Accession to the WTO 139 Latin America's Experience." WBI Development Studies No. International Institute for Communication Telecommunica- 23950. World Bank Institute, Washington, D.C. tions Forum, April 22­23. Sydney, Australia. Fink, Carsten, Aaditya Mattoo, and Randeep Rathindran. 2001. Smith, P. 1995. "End of the Line for the Local Loop Monopoly." "Liberalizing Basic Telecommunications: The Asian Experi- Public Policy for the Private Sector, Note No. 63. World ence." Policy Research Working Paper 2718. World Bank, Bank, Washington, D.C. Washington, D.C., November. United Nations Conference on Trade and Development Ianchovichina, E., W. Martin, and C. Wood. 2001. "Economic (UNCTAD) and World Bank. 1994. Liberalizing Interna- Effects of the Vietnam­US Bilateral Trade Agreement." tional Transactions in Services: A Handbook. New York and World Bank, Washington, D.C. Processed. Geneva: United Nations. Markusen, James, Thomas F. Rutherford, and David Tarr. 2000. World Bank. 2001. "Finance for Growth: Policy Choices in a "Foreign Direct Investment in Services and the Domestic Volatile World." World Bank Policy Research Report. World Market for Expertise." Policy Research Working Paper 2413. Bank and Oxford University Press. World Bank, Washington, D.C., August. WTO (World Trade Organization). 1999. "Structure of Com- Petrazzini, B. A., and P. Lovelock. 1996. "Telecommunications in mitments for Modes 1, 2 and 3." Background Note by the the Region: Comparative Case Studies." Paper presented at Secretariat, Council for Trade in Services, S/C/W/99, March. 9 Logistics in China: Implications of Accession to the WTO Wenping Luo and Christopher Findlay Study of the geography of poverty has revealed how evolve, and recently it became a "hot" topic in changes in transport costs, even when they make up China, with logistics services eliciting strong a small share of the value of a product, can lead to interest from both businesses and policymakers. large changes in value added and in growth rates Moreover, a significant part of seminar and man- (see Radelet and Sachs 1998 and Redding and agement training time in China is now devoted to Venables 2000). The rate at which transport costs developing what might be called the "logistics way accumulate over distance is usually taken as given of thinking." There is an appreciation that systems, in this research, and the questions asked are about processes, and perspectives--that is, the software of the links between location and rural producer the business--are just as important as the physical incomes. Policy prescriptions then usually refer to assets, if not more so, in the growth of enterprises processes under direct government control such as that are internationally competitive in logistics. port operations or customs procedures. However, We argue in this chapter that WTO accession has in this chapter we stress that policy reform, espe- significant implications for the logistics sector. cially through commitments to the World Trade Although there is no easy match between the vari- Organization (WTO), can have significant effects ous categories of the General Agreement on Trade not just on transport prices, but also on the other in Services (GATS) and the modern scope of logis- costs of getting goods to market in the form that tics, it is possible to make a concordance. We com- consumers prefer. These effects are important both plete that work in order to comment on the in international trade and in trade within an econ- implications of China's accession commitments. omy. Reductions in this set of costs can add to value The requirement to make these associations illus- added, especially in remote, poor areas. trates an important feature of scheduling commit- We illustrate these points in this study of the ments under GATS in emerging business areas. logistics sector in China and of the impact of The next section of this chapter reviews the def- China's accession to the WTO on the performance inition of logistics services. It is followed by an of that sector. The concept of logistics continues to overview of their evolution in China.1 The authors thank the editor of this volume and seminar participants for comments. Errors remain their own. 141 142 China and the WTO We then outline the WTO commitments made having a single transportation manager who by China related to this sector, and reach a set of could coordinate inbound and outbound trans- conclusions on their direct effects and their indirect portation. Then companies became aware of the implications. The direct effects, we emphasize, opportunities to view the whole process, from include substantial savings in real resources and not raw materials to work-in-process inventory to just transfer differences that represent relatively finished goods, as a continuum that, managed small gains in efficiency. from a systems perspective, could lead to more The indirect effects include reference to policy- efficient operation. making processes and administrative structures. · Stage 3: supply chain management (from 1980s The analysis here highlights the value of comple- to 1990s). Companies expanded their perspective mentary reforms in domestic institutions and regu- on the logistics processes to include all the firms latory agencies and their processes in order to involved, making use of partnerships/alliances maximize the gains from liberalization. between manufacturing companies and their The chapter then continues with a discussion of suppliers/vendors, customers (channels of dis- the likely orders of magnitude of the effects of a tribution), and other logistics-related parties more open logistics sector on total logistics costs, such as transportation and public warehousing and discusses the significance of these changes for companies. poor areas in China. We pay special attention to the role of the gains in increasing the value added in pro- The concentration on fewer carriers operating duction processes in poor areas. Some of the main under long-term contracts and on outsourcing the points are revisited in the final section of the chapter. whole package of services stimulated the develop- ment of third-party logistics (3PL) firms. As noted, several subsectors are involved in logistics--the Logistics Defined transport sector (including different modes of Logistics is the process of planning, implementing, transportation), the warehousing sector, and and controlling the efficient flow and storage of related providers of value-added services. The 3PL goods, services, and related information from the business is complex, because it operates across all point of origin to the point of consumption to meet the subsectors involved. customers' requirements. The provision of logis- Luo and Findlay (2001) reviewed developments tics services requires inputs from various service in each subsector in China, including the basic facts providers, including the providers of transport and of capacity, volume, and growth; evolution and warehousing as well as other value-adding activities. reform of the subsectors; subsector infrastructure The concept of logistics has undergone many and organization; and administrative, regulatory, significant changes. Coyle, Bardi, and Langley and policy issues. Some of the results relevant to (1996) divided the development of logistics into this chapter are outlined in the next section. three stages: The Logistics Sector in China · Stage 1: physical distribution or outbound logis- tics system (during 1960s and 1970s). Businesses Growth of the logistics sector in China has been attempted to manage systematically a set of dramatic. Freight turnover in China increased from interrelated activities, including transportation, 76 billion ton-kilometers in 1952 to 4,381 billion distribution, warehousing, finished goods, ton-kilometers in 2000. The volume increased by inventory levels, packaging, and materials han- 3.8 times between 1980 and 2000. dling, to ensure the efficient delivery of finished goods to customers. Evolution of Transport Services · Stage 2: integrated logistics management (dur- ing the 1970s and 1980s). Companies began to In the political and economic environment of the recognize the additional opportunities for sav- early 1950s, China gave priority to heavy industries. ings by combining the inbound side (materials A precondition for their development was a supply management) with the outbound side (physical of cheap raw materials. This development in turn distribution). Initially, savings was achieved by demanded a supply of cheap transport, and rail and Logistics in China: Implications of Accession to the WTO 143 FIGURE 9.1 Choice of Transport Modes: China, 1970­2000 Freight traffic by mode (metric tons) Percent 100 80 60 40 20 0 1970 1980 1990 2000 Rail Road Water Air Source: Luo and Findlay 2001. water were two modes of transportation that could in the manufacturing sector. But in the transporta- provide transport services for large volumes at a tion sector the change has been relatively small; relatively low price. large, state-owned enterprises still dominate the Figure 9.1 shows the composition of freight sector. In recent years, the use of joint ventures has carried by mode (measured in weight) from 1970 increased, especially in the container berth, con- to 2000. After 1978 the transport system in China tainer freight station, and motor carrier activities. was characterized by the rapid development of road However, these firms still account for only a small and air transportation. The market shares of water portion of the total number of enterprises. Road and rail transport decreased sharply, reflecting the transport and shipping were deregulated in 1986, shift from the traditional heavy industries to light and rail and air transportation were also decentral- industries that demanded speedy, flexible trans- ized to some extent. The ownership structure has portation services and that could bear a higher changed to different degrees in different modes of transport service price. Figure 9.1 shows the rising transportation. The road and inland shipping com- share of road transport and the fall in the shares of panies are dominated by collective and private rail and water in the total tonnage. Although it is enterprises, and coastal and international shipping, not evident, data for air transport were included in rail, and air transportation enterprises are still the construction of the chart. Those numbers are mostly state-owned. very small compared with those for other modes, The transport sector is still a weak link in China's and therefore are not evident in the columns of the economy: chart, but air freight volumes are growing fastest among all modes. As for trip lengths, they vary · The number of either rail or road trunk lines significantly by mode. The length of road trips is linking regions is insufficient. The major railway relatively short on average. Rail and water show lines are overloaded. The trunk lines of the high- longer trip lengths and therefore account for much way system linking provinces are not well net- higher shares of the transport task measured in worked and cannot meet the demand. These terms of ton-kilometers. But even in these terms, impediments also contribute to differences in the road share continues to grow, mainly at the growth rates between regions. expense of rail. · The coordination between different modes requires planning and development. Within each mode of transport, there are imbalances. In Constraints on Development the rail sector, the mixing together of passenger China's program of economic reform has gradually trains with cargo trains on the busy lines leads changed the ownership of its enterprises, especially to slower speeds and low efficiency. In the air 144 China and the WTO sector, there is an imbalance between the trunk Beijing, Tianjing, Shanghai, Guandong, and and feeder lines and in the number of big and Shandong have also paid attention to the develop- small aircraft. In the road sector, the construc- ment of their local logistics sectors. tion of the national-level trunk road system has Most of logistics activities are still operated in- been accelerated, but in the rural areas the road house, and demand for 3PL services is low among network is far less developed. traditional Chinese firms. State-owned companies · Service quality is generally low, as is the techni- in China have inherited an operating model based cal level of the equipment. For example, in the on the provision of services in-house, including rail sector double lines and electrified lines rep- logistics. According to investigations by the China resent a relatively small share of the total length Warehousing and Logistics Council, a third party of lines. Less than 20 percent of the freight undertakes only 18 percent of the inbound logistics. trucks can carry containers. As for outbound logistics, 59.8 percent is operated together with a 3PL firm, and only 16.1 percent is The warehousing and storage system in the operated totally by a 3PL firm. Most Chinese firms planned economy before reform was character- keep their own logistics assets and personnel (Luo ized by department or sector administration and and Findlay 2001). operations--that is, sectors built and owned their Although many 3PL firms have appeared in the own warehousing and storage facilities and served marketplace, most of them offer a low standard of their own demands only. Many sectors formed their service and efficiency. For one thing, most of the own closed warehousing systems.2 Meanwhile, each firms provide only simple delivery and warehous- local government at the provincial, city, and county ing services: value-added services such information levels established their own warehousing system to services, inventory management, logistics cost con- serve their local economy.Overall,therefore,govern- trol, or logistics system design are not offered. For ment administration and enterprise functions are another, the scale of the firms is usually small, and not separated, which constrains enterprise perform- most of them come from among traditional firms ance; the administration is fragmented in both verti- with weak network and organizational capabilities. cal and horizontal dimensions, which inhibits the Moreover, modern logistics facilities such as integration of services; and in some activities the transshipment facilities for different modes of quality of the physical infrastructure is insufficient. transport, public warehousing and transport facili- ties, and logistics centers are not available. At low levels are the technology used for logistics Demand for 3PL Services equipment and the extent to which facilities are The demand for specialized logistics or 3PL serv- standardized and information technology is ices has begun to grow faster in recent years, applied. Finally, logistics facilities are distributed especially that originating from the multinational unevenly--for example, transport facilities are less companies, the leading Chinese state-owned com- developed in the middle and western parts of the panies, and some of the private companies, as well country. as that arising from the e-commerce sector. In the An important constraint on the capacity of ser- recent years, quite a few 3PL companies have vice providers to offer integrated services lies in the appeared, including (a) international companies administration system--in particular, in the verti- such as Maersk, UPS, and TNT; (b) traditional cal and horizontal separations in the administrative Chinese transport, warehousing, and forwarding systems that have carried over from the planning companies such as COSCO and Sinotrans; and system, the self-protection that remains, the mixing (c) emerging companies such as the Bao Gong of the enterprise with administration functions, Logistics Group. and the lack of transparency in policy measures. All Relevant departments in the central government these issues, and the extent to which they impede have paid greater attention from different perspec- the scope to gain from reform in the logistics sector, tives to the logistics sector and its development, and are examined in more detail after a review of they have studied policies that they expect will the implications of China's WTO accession for the promote it. Local governments such as Shenzhen, sector. Logistics in China: Implications of Accession to the WTO 145 Accession to the WTO and Implications Our assessments of the effects of the likely policy for the Logistics Sector changes associated with WTO accession are divided into two types: direct and indirect. Direct effects are To provide integrated logistics services, a logistics those on the costs and quality of services provided. sector must cover the following service elements: Indirect effects are those on the regulatory or international and domestic transport (all modes); administrative environment. warehousing and storage; international and domes- tic freight forwarding; other value-added services such as logistics information (tracing and track- Direct Effects ing), packaging and labeling, and quality control/ inspection; and third-party logistics services. The direct effects of WTO accession on the logistics Various GATS categories must be examined in sector are presented in the appendix to this chapter order to incorporate the scope of logistics services and are summarized in tables 9.1­9.3. The tables as just defined. Although these categories are listed show the starting position (table 9.1) and the in the appendix to this chapter, this list is not com- impacts of the commitments one year after acces- prehensive, and in some cases the impact of logis- sion (table 9.2) and in the period three years after tics activities must be inferred from packages of accession (table 9.3). more specific commitments--for example, in rela- In general, China's accession to the WTO will tion to 3PL services. have a positive affect on performance in the logistics Another issue related to the definition of the services sector as a whole. Of the elements listed in sector is coverage of the GATS category called "dis- the tables, only the air transport and water transport tribution." Distribution includes some activities sectors are isolated from a significant change in pol- that might normally be associated with logistics, icy. The commitments in many areas extend existing such as warehousing and inventory management.3 policy (e.g., customs clearance, container depot It appears, therefore, that some parts of a logistics services, maritime agency services, road transport, sector, as might be defined in ideal terms, are rail transport, warehousing and forwarding, courier already included in the category of distribution. services). In others, the commitments bind current However, distribution refers to these activities policy (international shipping). In some cases, geo- when undertaken in the context of wholesaling and graphic limits apply--in internal waterway trans- retailing. Thus their coverage is much narrower port, foreign firms can operate only on routes to than that which is relevant here. ports open to foreign shipping. TABLE 9.1 Starting Position of China's Logistics Sector Area Starting Position Air transport Relatively closed in both domestic and international routes Water transport Inland shipping open to domestic companies; coastal and international shipping relatively closed Road transport No restriction for domestic entry; foreign entry only by joint ventures Rail transport State monopoly Forwarding International forwarding: licensing system applies; foreign participation only via a joint venture Domestic forwarding: relatively open for domestic companies; not open to foreign companies Storage and warehousing No restriction for domestic entry; foreign entry only by joint ventures Integrated services (3PL) Determined by the above Other (packaging) No restriction for domestic entry; foreign entry only by joint ventures Courier service No restriction for domestic entry; foreign entry only by joint ventures Source: The authors. 146 China and the WTO TABLE 9.2 Position of China's Logistics Sector One Year after WTO Accession Evaluation of Direct Impacts Area High Medium Low Air transport Water transport Road transport Rail transport Forwarding Storage and warehousing Integrated services (3PL) Courier service Other (packaging) Source: The authors. TABLE 9.3 Position of China's Logistics Sector Three Years after WTO Accession Evaluation of Direct Impacts Area High Medium Low Air transport Water transport Road transport (3) Rail transport (6) Forwarding (4) Storage and warehousing (3) Integrated services (3PL) Courier service Other (packaging) Note: The numbers in parentheses show the number of years after accession when wholly owned foreign establishments will be permitted. Source: The authors. Discrimination against foreign suppliers will foreign suppliers of second-best forms of participa- continue (mainly in terms of rules on the forms of tion in the market (joint ventures instead of wholly establishment) for varying periods, which delays owned operations). the benefits of reform. A schedule might be As discussed earlier, the demand for integrated designed to capture some of the rents created in the services (3PL) is increasing, and so, more and more, transition to full openness. It also might be logistics service firms are trying to provide or con- designed to slow down the transition and therefore trol the whole chain of the services. Therefore, a (it is hoped) reduce the costs of adjustment. The 3PL firm needs to have access to all the relevant consequence, however, may only be one of delaying licenses, which could include those for interna- those costs and adding to the risks of backsliding. tional freight forwarding, different modes of trans- Another goal may be to limit for a longer time for- portation (air, international shipping, domestic eign ownership in sensitive sectors. However, costs shipping, road, rail), storage and warehousing, would be incurred, including the imposition on container station and depot services, and courier Logistics in China: Implications of Accession to the WTO 147 services. The opening up of the different sectors in · Self-protection is evident at the departmental logistics, as discussed earlier, would have a signifi- and district levels. Policies and regulations are cant impact on the ability of logistics firms to pro- often designed according to the self-interest of vide the whole chain of services. the department, sector, or district involved. A local logistics firm will often seek protection from competition through the application of pol- Indirect Effects icy by departmental or district agencies, which This section assesses the indirect effects on the hampers the development of cross-department logistics sector of the reforms associated with or cross-district logistics networks (see box 9.1 China's accession to the WTO. These effects are for an example of local protectionism). summarized in table 9.4. Reform of the administrative system will have to Local Protectionism Vertical and horizontal be accelerated to meet WTO rules. The more rapid divisions within administrative systems constrain reform will break down the scope to apply protec- development of the logistics sector. For example, tion at the departmental and district levels, and the administrative system that applies to trans- it will facilitate competition in the market at portation is divided according to the mode of a national level. Bosworth (2002) points out, transport, and the same systems operate at both the however, that the obligations under the GATS cov- local and higher levels of government. The result is ering subnational governments are weaker than overlapping responsibility and power between the those applying to national governments. The departmental and district administrations. These requirement is that WTO members take only administrative structures are no longer suitable for "reasonable measures" to ensure that subnational the development of logistics activities for several governments meet their obligations. However, reasons: China's commitments under its Protocol of Acces- sion specifically require that it maintain a uniform · Coordination of the different modes of trans- system of administration. Furthermore, the central port services is difficult because the separation government is required to establish a mechanism of different modes has resulted in differences in whereby those concerned about problems of organization, service standards, and technical regional protection can bring their concerns to its and equipment standards. This lack of consis- attention. These specific provisions seem likely to tency leads to the difficulties in developing mul- make the disciplines on China stronger than those timodal services. under general WTO rules. · Coordination of the construction of infrastruc- ture is poor, which leads both to duplication of Separation of Government Administration from investments and to neglect of the development Enterprises Some government departments are of transshipment centers. still directly or indirectly involved in enterprise TABLE 9.4 Summary of Indirect Effects of WTO Accession on China's Logistics Sector Area of Existing Problems Evaluation of Indirect Impact Local protectionism GATS obligations also apply at subnational government level. Mixing of government administration State-owned enterprises are under pressure to transform and enterprise management to a market-oriented operation, and government adminis- Inadequate infrastructure and tration can be separated from enterprise management. facilities/equipment Sector likely to attract more foreign investment. Lack of research and professional skills Foreign-invested enterprises will bring know-how, innovation, technology, and management. Source: The authors. 148 China and the WTO BOX 9.1 Local Policies and Their Implications for Logistics Operations The Jia Yu Freight Group Company is a new pri- transport companies, but we have to fol- vate transport and logistics services company. Its low their special practice. For example, in more than 20 branches cover most of the coun- Guangzhou we have to rent their designated try to provide road express service in a fast- office room with a higher rental in a specified growing business. The company manager, place. Otherwise, we are not allowed to enter Mr. Zhai, complains about the business environ- the market. In Guangzhou, we have to ment, especially the local protection: deposit a larger amount of money and administration fee. But whenever we meet We face many obstacles in operating our difficulties, there is no assistance from them. business and providing our services to cus- tomers. Transport needs both registration in In general, local governments do not have iden- the Administration Bureau of Industry and tical rules and regulations. The road sector has Commerce of the local government and a been deregulated, but the local government license from the Road Administration Bureau always wants to protect its local companies by [RAB]. Each local RAB has a different prac- various means, which limits the development tice. Some cities like Changzhou in Jiangsu of cross-regional logistics services. "The strange Province and Wenzhou in Zhejian Province thing is the more developed the region, the do not issue licenses to companies from other more restrictions there are on the outside firms. regions. Our trucks are not allowed to enter For example, in Chongqing we haven't met these cities. If we carry cargoes from their much difficulty in establishing our presence and cities, we will be fined. Some cities are not so operating our business," Mr. Zhai added. extreme in their treatment of the outside Source: Author interviews. activities. As already noted, in the railway and air relevant know-how includes not only the operation sectors state ownership is still dominant. Port of particular segments of the logistics operation, administration is mixed with port enterprise oper- but also the manner in which the elements fit ations. These situations not only affect government together, as well as the management of the relation- functioning, but also weaken the competitiveness ships between all the suppliers and the customers. of the enterprises. Opening other service sectors, such as education, will also raise the quality of options available for Logistics Infrastructure and Facilities/Equipment education and training in this sector. The scale of the infrastructure and the quality of facilities and equipment available to logistics ser- Cost Savings in the Third-Party vices suppliers are relatively poor. The assessment Logistics Sector by the Hong Kong Trade Development Council (2002) is that the airport infrastructure is a "major Little empirical work has been conducted on the impediment to aviation industry growth." Further- efficiency of the logistics sector in China or on the more, there is an imbalance in development impacts of reform. Some ideas about methodology between the eastern and western parts of the coun- and some indicators of the scale of impacts are try. Road and rail transport, storage and warehous- available, however, from work done in other ing, container station and depot services, and markets--for example, on the effects of the integra- airports are some of the areas that could attract tion of transport markets in Europe (European more foreign investment after accession. Commission 1997).4 Those studying the European logistics services Human Resources Research on logistics, espe- markets identified eight input-oriented measures cially among business firms, is rare, and the profes- of logistics performance: (1) customer orientation; sional education system has so far included little (2) integrated long-range planning; (3) supplier logistics training. Foreign investment and establish- partnerships; (4) cross-functional operations; ments in the logistics services sector will bring with (5) continuous improvement process; (6) emplo- them technology and management know-how. The yee empowerment; (7) integrated information Logistics in China: Implications of Accession to the WTO 149 technology (IT) systems; and (8) measurement, According to Gibson (2001), electronics is associ- comparison, and action. ated with a significantly higher share of logistics Different stages of development of the logistics costs in total revenue than textiles. Gibson also sector can be identified through the use of these notes that in some sectors, such as fresh food, logis- indicators. Higher levels of performance, as meas- tics costs can be 50­60 percent of total revenue. ured by these indicators, lead to better performance The table also illustrates the variation in the sig- in terms of the costs of each element of the service nificance of logistics costs in Asia and those in and in terms of other indicators such as on-time Europe--the latter are lower by 24 percent. Gibson delivery, order completeness, invoice accuracy, and (2001) suggests that in China logistics could account damage-free delivery.All these features can therefore for 30­40 percent of the total cost of manufactured be summarized in the total cost of logistics services, goods. Another source suggests that "supply-chain- which refers to the sum of inventory, administrative, related costs can be 30% to 40% of wholesale prices warehousing, and transportation costs. Not only in China, compared to 5% to 20% in the U.S."5 The does market opening lower the prices of the direct Hong Kong Trade Development Council (2002) inputs (e.g., transport), but it also raises services refers to an assessment that logistics account for quality and thereby lowers other costs (e.g., ware- 40 percent of general production costs and that housing or inventory, or allowances for losses). "logistics . . . takes 90% of the whole production Transport is usually the single major component cycle time." of logistics costs. For example, in 1992 in Europe A key factor in the performance of the logistics logistics costs were on average 10.1 percent of rev- sector is the damage rate. One source reports that enue, down from 14.3 percent in 1987. The 1992 for fast-moving consumer goods, the damage rate share comprised 3 points for transport, 2.3 points is 5 percent in China compared with "well below" for warehousing, 1.9 points for administration, and 1 percent in developed economies.6 High rates of 2.9 points for carrying inventory. The logistics cost product loss lead to requirements for higher inven- share of revenue was expected to fall to 9.1 percent tory levels. The costs of inventory and product loss in 1997. In the firms regarded as leaders in the oper- can be significant, which reduces the relative ations of logistics, the logistics share of revenue was importance of the transport component of the total estimated to be as low as 6.5 percent (1992), 36 per- cost of distribution. Switching to a higher-quality cent lower than the overall average. The largest logistics service, even one that charges higher fees, decreases over this period in Europe were observed for example, a foreign provider can then lower in the component associated with inventory costs, overall logistics costs if the costs associated with followed by those linked with administration, ware- product loss are reduced sufficiently. housing, and transportation. Transport, even in Boillot and Michelon (2000) provide a break- those leading companies, remains the largest single down of the total logistics costs in China. They component of total logistics costs. report that losses and damage account for 48 per- The significance of logistics costs varies between cent of the total, compared with 21 percent for sectors of the economy as illustrated in table 9.5. transport activities. They also quote industry TABLE 9.5 Logistics Costs as a Share of Total Revenue, Europe and Asia, 2001 (percent) Sector Europe Asia Textile 8.7 11.5 Automotive 8.9 13.5 Chemicals 10.2 13.8 Fast-moving consumer goods 10.4 14.0 Electronics 12.6 14.0 Average 10.4 13.7 Source: Gibson 2001. 150 China and the WTO estimates that total logistics costs could fall by output. Changes in logistics costs even in the range 37 percent if the current arrangements for manag- suggested here can make a significant difference in ing logistics were replaced by a subcontracting the value added available for distribution to the fac- approach. The bulk of this reduction appears, in tors of production. If the payment to capital is their assessment, to be attributable to the reduction fixed, then the amount of valued added available to in losses and damage rather than a reduction in labor changes by an even greater proportion.8 This profit margins. Management fees may increase to example illustrates the significance of logistics effi- reflect the higher costs of the higher levels of man- ciency for inland areas of China. agement inputs in time and experience. A more efficient logistics sector will benefit peo- In summary, a more open market for logistics ple living in poor areas in several ways. For exam- services that transfers the technology for the ple, as implicit in the example just given, their redesign and management of integrated services is terms of trade will improve; the costs of items they likely to have significant effects on total logistics buy from the rest of China will be less, and the costs in China. The composition of those costs prices they receive for the items they export to the will change--for example, management fees will rest of China will be higher. The impact of reform, increase, but this change will be offset by reduc- which reduces logistics costs, also generates exten- tions in other items, not necessarily in the transport sive real gains: the impacts on real incomes are component but more likely in the costs associated greater than those of the removal of a tax, for exam- with damage, losses, and inventory. ple. The impact of tax removal includes transfer To summarize this discussion, if total logistics effects, but when logistics costs are reduced, costs are 30­40 percent of wholesale prices in resources are saved. In other words, the rectangle China, as suggested earlier, and if those costs fall by effects are gains, not just transfers (see Deardorff 35 percent, then logistics costs as a share of whole- 2001). sale prices would fall by 10­14 percentage points, to Reductions in transport costs can also lead to 20­26 percent of wholesale prices. But even this significant increases in trade orientation. Some of assessment may be conservative, because logistics the orders of magnitude involved in these effects costs as a share of the wholesale price in China is are illustrated by the work on Africa by Limao and still above the average for Asia in table 9.5. It also Venables (2001) who estimate the response of trade lies above that observed in developed market to reductions in transport costs. They find an economies, although there are limitations on the elasticity of 2.5, which could be applied to the extent to which comparisons of these ratios can be reduction of 35 percent in logistics costs, according made across countries.7 to the assessment reported earlier. The opening up of markets stemming from improvements in transport and logistics services Implications for Less-developed also can have effects on competition in the local Areas in China markets. And competition itself has additional As noted in the previous section, reforms associ- effects. One is the impact on the rents available. ated with China's accession to the WTO could The suppliers who previously had monopoly result in a fall in logistics costs equal to 10­14 per- power are constrained by the options now avail- cent of wholesale prices in China. In an example able to their customers to seek supplies from out- developed by Radelet and Sachs (1998), suppose side the region. Thus not only does regulatory that a region of China faces a perfectly elastic reform lower costs, but also its competitive effects demand for its output and a perfectly elastic supply force the passing on of those cost reductions to of purchased intermediate inputs that come from consumers in business and in households.9 Firms outside the region. Suppose that the share of pur- that previously earned rents from their protected chased inputs in gross value is 30 percent. If then position are clearly worse off, but overall the logistics costs are 35 percent of the value of output region will realize a welfare gain from the intro- and inputs, the value added is 24.5 percent of the duction of competition. In terms of the welfare of value of output. If logistics costs fall to 23 percent those people living in the poor area, this effect is of the value of output, the value added rises by even larger when those who captured the rents 64 percent, to reach 40 percent of the value of were not local firms. Logistics in China: Implications of Accession to the WTO 151 Competition, and a greater foreign (both from China can learn much from the considerable out of the country and out of the region) presence experience in the rest of the world with community in all markets within a poor area, also could have a service obligation policies, and the scope for China dynamic effect on, for example, the transfer of to innovate in policy development in this field is technology and productivity growth. This effect is great. likely to be greater in those services markets where an establishment is required to deliver the activity, and where there is therefore more interaction Conclusions between the new supplier and local firms. The China experience illustrates that the design of After markets become more competitive, some commitments in the GATS is complicated in new localities may continue to lack services of a quality business areas. For example, the GATS categories typical of those in more developed areas. Govern- do not match exactly the business structures used ments often respond to this situation by subsidiz- to provide logistics services. However, it is possible ing the provision of services. Subsidies might be to make a concordance: China's commitments, arranged within the sector (supported by a regula- which are significant and which in most parts go tory structure that facilitates these transfers beyond a binding of current policy, provide a tem- between consumers), or through explicit payments plate that other developing economies may con- from the budget. The design of these policies sider for this sector. includes important efficiency considerations, both China's experience also illustrates how WTO on the funding side and in the manner in which the commitments can open up the component activi- subsidies are applied. ties of the logistics sector, including areas that in Gruen (2001) has questioned and proposed the past might have been regarded as nontrade- options for the management of service obligations. ables. The commitments thereby also facilitate the The popular approach to a service obligation policy is growth of third-party logistics providers. The to use cross-subsidies between consumer groups to design and management of the provision of ser- provide specific services (e.g., a rail link) to all resi- vices in this field is important; not just the extent dents in a particular area. Gruen suggests instead and quality of physical capacity matters. The pres- strategies in which services are provided to well- ence of foreign providers of these integrating ser- defined target households and where the recipients vices will be facilitated by the WTO commitments. have options for the manner in which they receive A feature of the set of commitments reviewed that support ("cashing out," for example, or at least here was the application of timetables for their substitutability between modes of supply--in this implementation. Some motivations for that case,of transport services).Analysis of costs and ben- approach can be identified, but it comes at some efits of alternatives is important--for example, those cost, including the higher risks of backsliding and of buildingaroadtoanexistingcommunityorimple- the use of inefficient forms of business organ- menting an option for relocating that community.10 ization by firms that enter in anticipation of An important consideration in the design of further market opening--for example, where joint policies to support provision of services to poor ventures are permitted but full foreign ownership areas is their consistency with obligations under is not. the WTO on subsidies. Models for the design of One of the challenges, however, is to promote these policies are available, however--for example, complementary administrative and regulatory in the WTO's April 1996 reference paper on basic reform. The China case highlights the risks associ- telecommunications, which notes: ated with administrative structures, even at the cen- Any Member has the right to define the kind tral level, that add to the complexity of establishing of universal service obligation it wishes to main- the most efficient business structures (e.g., for tain. Such obligations will not be regarded as intermodal transfers of freight). Perhaps an even anti-competitive per se,provided they are admin- more important issue is to translate national com- istered in a transparent, non-discriminatory and mitments to the WTO into local government pol- competitively neutral manner and are not more icy, especially in regions where there has been a burdensome than necessary for the kind of uni- tradition of using regulatory instruments to protect versal service defined by the Member.11 local businesses. Implementation of the principles 152 China and the WTO of transparency and nondiscrimination in the 1. Aircraft Repair and Maintenance Services application of business regulation are important COMMITMENT elements of the extent to which markets are actually Foreign service suppliers are permitted to estab- opened to foreign competition. lish joint venture aircraft repair and maintenance A more open logistics sector, especially through enterprises in China. The Chinese side should hold the growth of 3PL providers, can lead to significant a controlling share or be in a dominant position in cost reductions. One estimate, based on assess- the joint ventures. Licenses for the establishment of ments of the state of logistics services in China by joint ventures are subject to an economic needs test. industry commentators, is that the more wide- spread presence and use of efficient and interna- IMPLICATIONS tionally competitive 3PL providers could lower The commitment effectively binds current pol- costs by an amount equivalent to at least 10 percent icy. Two joint venture aircraft repair and mainte- of the wholesale price of manufactured goods on nance service firms are already located in China: average. The impact on these relative terms may be AMECO (founded in July 1989) with a 40 percent even higher for fresh food produce. These amounts foreign share and GAMECO (founded in August are not simply transfers, but real gains associated 1989) with 50 percent share. Binding policy may with the saving of resources. For example, an reduce uncertainty from the perspective of poten- important source of gains is the lower wastage rates tial entrants. Further entry would add to competi- in transport and warehousing. tion and may facilitate the introduction of new Changes of these magnitudes in the quality of technology and management expertise in aircraft transport services could have important effects for repair and maintenance, so that a higher volume of poor areas. Their lack of access to markets in the qualified services are made available within China. rest of the country and in the rest of the world is an important constraint on their development. A 2. Computer Reservation System (CRS) Services more open and competitive logistics sector will remove some of the impediments to their partici- COMMITMENT pation in trade within the economy and with the For cross-border supply: rest of the world. The effects of lower logistics costs on value added and on the total value of wages paid · Foreign computer reservation systems, when hav- and available for distribution to workers are signif- ing agreements with Chinese aviation enterprises icant. In an example reported here, a 12 percentage and the Chinese computer reservation system, point reduction in the logistics costs in the whole- may provide services to Chinese aviation enter- sale value of goods leads to an increase in value prises and Chinese aviation agents by connecting added of more than 60 percent. with the Chinese computer reservation system. However, some impediments remain to the sup- · Foreign computer reservation systems may pro- ply of services in remote areas.Policies to support the vide services to representative offices and sales provision of services where these impediments apply offices established in the destination cities in are not inconsistent with the GATS, but they should China by foreign aviation enterprises that have be evaluated against alternative strategies for pro- the right to engage in business under the bilat- moting the income growth of those communities. eral aviation agreements. · Direct access to and use of foreign computer Appendix: China's Commitments reservation systems by Chinese aviation enter- on Logistics Services­Related prises and agents of foreign aviation enterprises Sectors and Their Implications are subject to the approval of the general admin- istration of the Civil Aviation Administration of Air Transport China (CAAC). Air transport remains quarantined from GATS cov- erage except for a limited number of complementary IMPLICATIONS services.12 Even in these areas, China's commitments The commitment refers only to foreign connec- are relatively limited. tion with the Chinese system. The commitment Logistics in China: Implications of Accession to the WTO 153 confirms policy as it is now. Chinese carriers have 4. Maritime Agency Services code-sharing arrangements with carriers in more COMMITMENT than 10 foreign countries, but direct foreign partic- Foreign participation is permitted in the form of ipation in the provision of CRS services is not joint ventures, but the foreign equity share may not allowed. Commercial presence by foreign providers exceed 49 percent. is still not permitted. IMPLICATIONS Maritime agency services are not presently open, Water Transport (International and so this commitment is significant. The level of for- Domestic Shipping) eign ownership remains constrained, however. 1. Maritime Transport Services: International Transport (Freight and Passengers, (CPC 7211 5. Internal Waterways Transport: Freight Transport and 7212 Less Cabotage Transport Services) (CPC 7222) COMMITMENT COMMITMENT Registered companies may be added for the pur- For cross-border supply: international shipping pose of operating a fleet under the national flag of is possible on routes to ports that are open to for- China. Foreign service suppliers are permitted to eign vessels. establish joint venture shipping companies, but the IMPLICATIONS foreign investment shall not exceed 49 percent of This policy will apply only in ports open to for- the total registered capital of the joint venture. The eign shipping. In 1999 there were 130 such ports. chairman of the board of directors and the general manager of the joint venture shall be appointed by the Chinese side. Road Transport IMPLICATIONS Road Transport Services: Freight Transportation by The commitment is the same as current policy. Road in Trucks or Cars (CPC 7123), Storage and Warehousing Service (CPC 742) 2. Auxiliary Services: Maritime Services (CPC 741), COMMITMENT Customs Clearance Services for Maritime Foreign participation is permitted in the form of Transport joint ventures, but the foreign equity share may not COMMITMENT exceed 49 percent. Within one year of China's Foreign participation is possible only in the accession, foreign majority ownership is permitted. form of joint ventures, but foreign majority owner- Within three years of China's accession, wholly ship is permitted. foreign-owned subsidiaries are permitted. IMPLICATIONS IMPLICATIONS Joint ventures are already allowed, but currently In the road transport sector joint ventures are the maximum foreign share is 49 percent. The allowed at present, but foreign partners are allowed commitment relaxes that constraint. only a minor share (less than 49 percent). This sector is now very competitive with low margins 3. Container Station and Depot Services because it was subject to domestic deregulation in 1986. Road transport is an essential link of the COMMITMENT logistics chain and, of all the transport sectors, Foreign participation is possible only in the plays the most important role. Moreover, of the form of joint ventures, but foreign majority owner- four modes of transport, the road sector is the ship is permitted. fastest growing. Foreign logistics firms might have IMPLICATIONS an interest in participating in the road transport Joint ventures are already allowed, but currently sector by either providing new capacity (perhaps the maximum foreign share is 49 percent. The in special road transport services) or buying an commitment relaxes that constraint. existing fleet. Some of the effects of deregulation 154 China and the WTO include greater use of subcontracting and larger- and within three years of accession wholly foreign- scale and lower-cost operations. owned subsidiaries are permitted. IMPLICATIONS Rail Transport Storage and warehousing are an important link Rail Transport Services: Freight Transportation by in logistics services. Joint ventures are allowed at Rail (CPC 7112) present, but foreign partners are allowed only a minor share. The margins in this sector are low. COMMITMENT There also appears to be excess capacity, but the Foreign participation is permitted in the form of service standard is low, and the mix of services joint ventures, but the foreign equity share may not available does not match demand. The opening up exceed 49 percent. Within three years of China's of the market could help foreign-owned logistics accession, foreign majority ownership will be per- suppliers establish their own presence in this sector mitted and within six years of accession wholly by either building new warehouses or buying the foreign-owned subsidiaries will be permitted. existing warehouses that are appropriate. IMPLICATIONS Railway transport continues to be monopolized Forwarding by the government. China's WTO commitment in this sector reveals its intention to open this trans- Freight Forwarding Agency Services (CPC 748 and port up to the outside world. The process is, how- 749, Excluding Freight Inspection) ever, slower than that in the road sector. Only six years after China's accession to the WTO will COMMITMENT wholly foreign-owned subsidiaries be permitted. Upon China's accession, foreign forwarding agen- Some parts of the rail sector are profitable, and oth- cies that have at least three consecutive years of expe- ers are not (e.g., in rural areas). Foreign firms will rience are permitted to set up freight forwarding take an interest in the profitable lines or regions. agency joint ventures in China, with foreign invest- Evidence from other markets indicates that pro- ment not to exceed 50 percent. Within one year of ductivity in rail is related to government policy. Key China's accession, foreign majority ownership is factors include degree of managerial autonomy and permitted. Within four years of accession, wholly the extent of subsidies made available. Higher levels foreign-owned subsidiaries will be permitted. The of private (and foreign) ownership would be minimum registered capital for a joint venture is to expected to add to productivity in this sector. One be no less than US$1 million. With four years of important issue is how the unprofitable lines can be accession, national treatment will be accorded in this maintained if that is required to meet service obliga- respect. The term of the joint ventures shall not tions. The effects of reform in this area could be sig- exceed 20 years.After one year of operating in China, nificant. According to one source, "Most logistics the joint venture can set up branches, but another companies and distribution managers continue to US$120,000 should be added to the original regis- argue that there is never a good reason to use rail for tered capital of the joint venture for the setup of each transport." Factors in this view include damage branch. Within two years of China's accession to the rates, delays, slow speed, lack of information and WTO, this rule on an additional registered capital booking requirements, and lack of route flexibility.13 requirement will be implemented on the national treatment basis. A foreign freight forwarding agency may set up a second joint venture after its first joint Warehousing and Storage venture has been in operation for five years. Within Storage and Warehousing Services (CPC 742) two years of China's accession to WTO, this require- ment will be reduced to two years. COMMITMENT Foreign participation is permitted in the form IMPLICATIONS of joint ventures, but the foreign equity share may Joint ventures are allowed at present, but the not exceed 49 percent. Within one year of China's foreign partner is allowed only a minor share. accession, foreign majority ownership is permitted, Within four years of China's accession, wholly Logistics in China: Implications of Accession to the WTO 155 foreign-owned subsidiaries will be permitted. trade goods (wai mao), railway, road, and water transportation Foreign-owned firms could enter the market easily (jiaotong), military, and basic food (liang shi). 3. Distribution (trade) is different from logistics. But the after the restriction is lifted. Many firms have related subordinated services are increasingly likely to be already presented themselves in the market in the included in the logistics sector, because the sector has taken on form of a joint venture. Although, according to the more and more of the value-added services demanded by ship- pers. As defined, distribution trade services are composed of regulations, only a minor share is allowed, the for- four main subsectors: commission agents' services, wholesaling, eign partner actually may control the business. retailing, and franchising. The principal services rendered in Some firms enter the market by using the "shell" of each subsector can be characterized as reselling merchandise, accompanied by a variety of related subordinated services, a domestic company. Complete opening of the including inventory management; assembly, sorting, and grad- market will reduce for these firms the costs of pro- ing of bulk lots; breaking bulk lots and redistributing into viding forwarding services and also enhance their smaller lots; delivery services; refrigeration, storage, warehous- ability to provide "whole of chain" services. ing, and garage services; sales promotion, marketing, and adver- tising; and installation and after-sales services, including maintenance and repair and training services (covered by CPC Related Areas [Central Product Classification] 61, 62, 63, and 8929). 4. The experience of regulatory reform in the United States Courier Services (CPC 75121, Except Those Currently can be used as another comparator. A longer time-series of data Specifically Reserved to Chinese Postal Authorities by on the European experience may also now be available. Law) 5. Andrew Tanzer quoting McKinsey & Co. in "Chinese Walls," Forbes Global, November 12, 2001. Available at COMMITMENT www.forbes.com/global/2001/1112/091.html. 6. Economic Intelligence Unit, China Hand (database), Upon China's accession, foreign service suppli- December 1999, Chap. 12, Sec. 1. ers are permitted to establish a joint venture, with 7. The ratio of logistics costs to wholesale prices in different foreign investment not exceeding 49 percent. countries will depend not only on the characteristics of the Within one year of accession, foreign majority product, but also on the factor intensities of the different pro- duction processes of both the logistics services and the manufac- ownership is permitted. Within three years of tured products and the relative factor prices in the economy accession, foreign service suppliers are permitted to concerned. establish wholly foreign-owned subsidiaries. 8. For example, if the cost of payments for capital is 20 per- cent of the value of output, the funds available for the wage bill IMPLICATIONS rise by a factor of 3.5 as logistics costs fall from 35 to 23 percent of wholesale prices. Courier services are an important part of logis- 9.According to anecdotal evidence,this effect has appeared in tics services. A joint venture is allowed at present, markets for light industrial products made in the past by town- but only with a minor share for foreigners. The ship and village enterprises. Brand-name products from suppli- opening up of this sector could bring up more ers based in eastern China are now much more widespread. 10. Gruen also discusses the value of broadening the funding competition in this market, and there is already evi- base for service obligation policies. dence of the interest of foreign firms in this sector. 11. Available at www.wto.org/english/tratop_e/serv_e/ telecom_e/tel23_e.htm. Packaging Services (CPC 876) 12. See Hong Kong Trade Development Council (2002) for another review of China's WTO commitments in transportation COMMITMENT and logistics. Foreign service suppliers will be allowed to 13. Economic Intelligence Unit, China Hand (database), establish a joint venture in China. Within one year December 1999, chap. 12, sec. 3. of China's accession, foreign majority ownership is References permitted. Within three years of China's accession, foreign service suppliers are permitted to establish Boillot, Jean-Joseph, and Nicholas Michelon. 2000. "The New wholly foreign-owned subsidiaries. Economic Geography of Greater China." China Perspectives 30 (July­August): 18­30. Bosworth, M. 2002. "Most-Favoured-Nation Treatment and Notes National Treatment in the GATS." In Sherry Stephenson and Christopher Findlay, with Soonhwa Yi, eds., Services Trade 1. These sections are summaries of the material in Luo and Liberalisation and Facilitation. Canberra: Asia Pacific Press, Findlay (2001). For a review of developments in the logistics sec- ANU. tors in three East Asian economies, including China, see Heaver Coyle, J. J., E. J. Bardi, and J. C. Langley. 1996. The Management of (2001). Business Logistics. 6th ed. St. Paul, Minn.: West Publishing Co. 2. These systems operated in these sectors: raw material and Deardorff, A. 2001. "International Provision of Trade Services, components (wu zi), manufactured consumer products (shang Trade, and Fragmentation."Review of International Economics ye), rural area suppliers and distributors (gong xiao), foreign 9 (2): 233­48. 156 China and the WTO European Commission. 1997. Impact on Services: Transport Luo W., and C. Findlay. 2001. "Analysis of Logistics Service Sec- Networks. The Single Market Review, subseries II, vol. 11. tor in China: Accession to the WTO and Its Implications." Luxembourg: Office for Official Publications of the Paper prepared for Authors' Workshop on WTO Accession European Communities. and Poverty Reduction, Beijing, October 26­27. Gibson, Ken. 2001. "Analysing the Trends: Predicting the Future Radelet, S., and J. Sachs. 1998. "Shipping Costs, Manufactured of Logistics in Asia." Presentation to the conference "Toward Exports and Economic Growth."Paper presented at American the Final Frontier: Logistics and the Efficient Supply Chain." Economic Association Meetings, Harvard University. September 27. Available at www.dnmstrategies.com/ Processed. Available at www2.cid.harvard.edu/hiidpapers/ presentation/tnt/sep_2001/KenGibsonV1.0-0945.pdf. shipcost.pdf. Gruen, N. 2001."Beyond the Safety Net--A View from Outside." Redding, S., and T. Venables. 2000. "Economic Geography and Presentation to the APEC Telecommunications Working International Inequality." CEPR Discussion Paper Series Group, Canberra, March. No. 2568. Centre for Economic Policy Research, London, Heaver, T. 2001."Logistics in East Asia." Development Economic September. Research Group, World Bank, Washington, D.C. WTO (World Trade Organization). 1996. Telecommunica- Hong Kong Trade Development Council. 2002. "China's tions services: Reference Paper, Negotiating Group on WTO Accession: Enhancing Supply Chain Efficiency-- basic telecommunications. WTO, Geneva. Available at Transportation and Logistics." Hong Kong. http://www.wto.org/english/tratop_e/serv_e/telecom_e/ Limao, N., and T. Venables. 2001. "Infrastructure, Geographical tel23_e.htm. Disadvantage and Transport Costs." World Bank Economic Review 15: 315­43. 10 Telecommunications Services in China: Facing the Challenges of WTO Accession Mari Pangestu and Debbie Mrongowius Despite some significant reforms, China's telecom- the five-year phase-in period under its WTO acces- munications sector since the country's accession to sion agreement. The second section sets out to the World Trade Organization (WTO) remains one analyze the effects of current reforms on greater effi- of the most restricted and regulated of those in the ciency, reduced costs, improved quality, and greater major developing countries in the region. During access in the sector. The third and fourth sections the WTO negotiations, liberalization of this sector evaluate the post­WTO accession environment was a critical issue both because of its growth faced by China and analyze the possible direct and potential and because it was considered one of indirect implications. The final section draws gen- China's "key national industries."1 By the end of the eral conclusions and policy implications about the decade, China is projected to become the largest telecommunications sector in China. market for telecommunications services in the world, creating a strong incentive for foreign play- ers to enter the market.2 The Landscape of the Telecommunications Industry Reforms were sought before and in association before WTO Accession4 with China's WTO accession, as well as in response to technological advances and the gradual com- In anticipation of the liberalization commitments it moditization of services.3 However, competition would have to make upon its accession to the WTO, remains constrained, and obstacles include lack of China began in 1994 to undertake reforms to clear and concise legislation, regulatory implemen- increase competition in its telecommunications sec- tation, and regulatory independence. The key chal- tor. It took tentative first steps in 1994 to change the lenge for China is to ensure that regulatory reforms monopoly structure, but only in the last five years, and new institutions will create greater competition beginning in 1998,has it undertaken major reforms. that will maximize the benefits for the Chinese economy as well as meet equity objectives. Tentative First Steps: Constrained Competition This chapter has several goals. The first major section seeks to examine the pace and progress of Up to 1994 China Telecom was the sole basic change in China's telecommunications services sec- telecommunications services provider in China. tor and to assess China's position at the beginning of New market entry was prohibited, and the Ministry 157 158 China and the WTO of Posts and Telecommunications (MPT) not only State Development Planning Commission (SDPC). regulated the sector but also operated telecommuni- In response to the inefficiency in the telecommuni- cations services. In 1994 the China United Telecom- cations system and its inability to provide connec- munications Corporation (Unicom) was created to tion, rural communities began to finance and improve service quality by introducing managed establish their own systems to provide basic service competition (see chapter appendix). However, pric- in their areas. These town and village enterprises ing policies, such as service fees, remained under the (TVEs) then merged into larger organizations pro- MPT, and individual operators were not able to viding wider coverage and eventually set up admin- compete through the price mechanism. Unicom istrative relationships with the local office of faced an uphill battle in its early years.5 the MPT. In addition, various networks had also The biggest problem faced by Unicom was get- evolved within separate industries and institutions ting access to the backbone infrastructure con- in order to compensate for the supply shortage in trolled by China Telecom. When Unicom was first the public network maintained by the MPT. These granted fixed-line services in some provinces, the networks included those for the academic and local China Telecom operators refused to allocate research communities, the military, the railways, phone numbers to Unicom, which held up inter- and the oil industry. connection for over a year. Another problem was To separate the regulatory and operational func- securing access to capital, which Unicom did tion of the MPT, as well as to respond to changes in largely by using "gray area" measures, the most well technology and administrative reforms, the Chinese known of which was the Chinese-Chinese-Foreign government created the Ministry of Information (Zhong-Zhong-Wai or CCF) joint venture enter- Industry (MII) in March 1998.6 This"super agency" prise structure, to circumvent the ban on foreign was created by merging the MPT; the Ministry direct investment (FDI) in the sector. The govern- of Electronics Industry (MEI); and parts of the ment subsequently initiated an investigation into State Administration of Radio, Film and Television the legality of CCF ventures in light of the 1993 (SARFT), China Aerospace Industry Corporation, directive forbidding foreign investment in the sec- and China Aviation Corporation. The MII worked tor. The investigation resulted in restrictions on closely with the State Economic and Trade Com- CCF arrangements: a mandatory limit on agree- mission (SETC) in setting policies for the informa- ments of 15 years, a five-year limit on profit shar- tion industry about the introduction of foreign ing, and transfer of 90 percent of all network assets investment and technologies, as well as in establish- to the Chinese party. ing industry strategic development plans. However, Another company, Jitong, was created in 1994 as the independence of the MPT and China Telecom part of the "Golden Bridge Network"--a govern- was not convincingly demonstrated, in part because ment initiative to develop a series of information many of the senior staff at the MII came from the networks. Jitong's original objective was to provide MPT. data services via satellite connections, and it was intended to compete with China Telecom. Most of Breaking Up the China Telecom Monopoly Jitong's services have now been shifted to an Internet protocol (IP) network infrastructure. These services The MII's fear that the monopoly it enjoyed would include Internet access and gateway services; value- be sacrificed as a "pawn" for China's accession to added services such as Web site hosting, server man- the WTO, and its conviction that national interests agement, and e-commerce solutions; and VoIP would be compromised if China's telecom industry (Voice over Internet Protocol). In early 1999 Jitong did not survive the resulting competition, galva- was the first Chinese operator to offer VoIP, and it nized the ministry into devising a survival plan, became at once a dominant player in the market. beginning in 1998. This strategy involved creating a The reforms undertaken up to 1998 were signif- favorable policy environment for China Unicom icant first steps, but the competition remained con- and breaking up the China Telecom monopoly. strained, the service quality poor, and the prices In 1999 the MII launched major restructuring uncompetitive by global standards. Prices were still efforts by splitting China Telecom into three state- set by the MPT, the State Pricing Board, and the owned companies and reinvigorating Unicom. The Telecommunications Services in China: Facing the Challenges of WTO Accession 159 three companies emerging from China Telecom allowed to charge 10 percent lower rates than China were: China Telecom for fixed-line business; China Mobile, the right to operate a code division multiple Mobile Communications Corporation (CMCC or access (CDMA) mobile phone system7 and to list its ChinaMobile)formobilephonebusiness;andChina shares on the Hong Kong (China) stock exchange. Satellite (ChinaSat) for satellite communications. In Unicom also has a modern management structure, addition, a new company, China Netcom Corpora- because it did not inherit an existing state-owned tion (CNC) was established to build a broadband enterprise structure. Internet protocol network. The division of enter- China Mobile has listed shares from its most prises by subsector reflected MII's plan to speed up profitable subsidiaries through China Mobile Group the development of individual service sectors. (Hong Kong) Co. Ltd. The latter company listed In 2000 a license was also given to Railcom,which 24.4 percent of its shares on the Hong Kong and New was part of the Ministry of Railways, to provide all York stock exchanges. Its stated development strat- basic telecommunications services except mobile. egy is to shift from traditional circuit-switched net- Thus, together with Jitong, at the end of 2001 there works to broadband packet-switched IP networks were seven major operators in the telecommunica- and to diversify from pure wireless voice service to tions sector. More than 3,000 different enterprises mobile image and data services. Its competitive edge are engaged in Internet-related and other value- over Unicom in the mobile sector is its financial added businesses, which enjoy greater operational resources, geographic coverage, and brand name. freedom and are market-oriented. Current opera- China Telecom owned most of the backbone tors are only granted licenses that specify the type infrastructure and leased it to the others.Only China of telecom service they may offer (table 10.1). Telecom and China Mobile, which originally was Only Unicom has a license to undertake all serv- part of China Telecom, operated in all provinces. As ices except for satellite. Although mobile telephony for mobile phone market, Unicom has been given a is the most rapidly growing sector in the industry, boost since its first uphill struggle due to some of the currently there are only two mobile phone licensees advantages mentioned above, and it has reached a and no clear indications about whether an addi- market share of about one-third. Even though other tional license will be issued despite much specula- operators are covering some areas, Unicom's cover- tion that China Telecom will be awarded the coveted age appears to be quite extensive, with expansion license. Other advantages of Unicom include being plans in place. TABLE 10.1 Network Operators and Licensing, China (end of 2001) License International Company Fixed-Line Cellular Paging VoIP Dataa Gateway Satellite China Telecom Unicom China Mobile b China Netcom Jitongc d Railcome ChinaSat aData includes fax and Internet transmission. bOnly mobile data transmission. cJitong serves only corporate clients and has since merged with China Netcom and part of China Telecom. dOnly VSAT. eAlthough Railcom is licensed to provide services other than fixed-line, it has yet to develop the capability to do so. Source: Compiled by authors. 160 China and the WTO TABLE 10.2 Network Operators in China: Description of Network and Coverage Networks Coverage Company Own Leasea Current Planned China Telecom All provinces Add capacity for 18.5 million new telephone subscribers and 7.1 million data communication users. Unicom 180 cities; mobile: CDMA coverage for 200 cities 322 cities China Mobile All provinces; GPRS GPRS for all provinces service in 25 cities and 16 provinces China Netcom 18 major cities Add 15,000 km of fiber-optic cable Jitong 120 cities; 30 cities 180 cities by 2002 covered by city-area networks; 1,000 VSAT stations Railcom 28 cities, 14 provinces 100 cities, all provinces ChinaSat All provincesb n.a. n.a. Not available. Note: GPRS general packet radio service. aAlthough all operators have built their own backbone networks, China Telecom still maintains a monopoly over end local loops throughout the country. Currently, interconnection fees must be negotiated individually at every China Telecom province administration. bThis includes 120 duplex remote stations providing access to remote regions such as Guangxi, Guizhou, Inner Mongolia, Qinghai, Tibet, and Yunnan. Source: Compiled by the authors. As for universal service obligations (USOs), all Council) and the MII and other government bodies operators are subject to these obligations, but there in charge of information security and enforcement. are no clear implementing regulations. Because The SILG was created in the mid-1990s in response China Telecom is the primary owner of the func- to the growth of Internet service and computer tional local end loops, some believe it eventually will crimes. The group wields much control and influ- be the major deliverer of universal service. Already ence over the development of the sector, and required to provide telecommunications services to approves and modifies the framework for industry government organizations and emergency services, regulations and the future direction for the indus- it has implicitly been the one given the task of try, and measures to implement policy. The State providing universal service. Council Informatization Office (SCIO) was formed under the SILG in August 2001 as the executive arm with a mandate to explain and carry out govern- Regulatory Framework ment policy. SCIO is a high-profile organization, Since 1998 the Ministry of Information Industry and it plays a more direct role than SILG in develop- has been the regulator of telecommunications and ing and implementing information policy in China. other information technology­related products Because the Internet involves information flows and services, including data communications, wire- to the public, which raises issues of censorship and less communications, electronics, computers, the security, other relevant authorities are the State Internet, and software. The MII reports to the State Secrets Bureau, Ministry of Public Security Council and is a member of the State Information (network security), and the State Administration Leading Group (SILG), which is the government for Industry and Commerce, which is responsible agency that links the central government (State for registering Internet service providers (ISPs) and Telecommunications Services in China: Facing the Challenges of WTO Accession 161 Internet content providers (ICPs). The MII, State discussions. Public hearings on the proposed prices Development Planning Commission, and Ministry must then be held before the prices are resubmitted of Finance (MOF), with approval from the State to the State Council for final approval and imple- Council, determine prices and rates. mentation. In practice, basic telecommunications Prior to the reforms that began in 2001, various service fees have come down since 1999. issues arose about the regulatory framework. One In 1999 the MII slashed service fees, in part to was the independence of MII from the incumbent address consumer complaints, but also to increase telecom operators, since the latter still comprised of consumer access to advanced technology and to officials from the Ministry of Post and Telecommu- rationalize the fee structure in anticipation of nications (MPT) structure and still influential in the China's WTO entry. The move was also part of a new MII structure. Other issues are that although strategy to provide universal service. The drastic cuts the MII supervises the telecom network, SARFT pre- accelerated telephone popularization in rural areas, sides over cable markets, and other departments reg- doubled the number of Chinese accessing the Inter- ulate the Internet. Given the number of authorities net, and enabled Chinese Internet service providers in charge of the overall sector, it is often not clear to expand their business and telecom facilities,which who is in charge of policies and guidelines for the inturnbenefitedChineseInternetcontentproviders. sector. Another general problem--not just for regu- Reducing long-distance rates was intended to cut lation of the Internet but more generally--is the gap down on "information smuggling"--that is, chan- between national and local policies. Shanghai, neling calls to China through Hong Kong (China), for example, has taken the lead in encouraging where rates were lower. e-commerce,and local authorities have forged ahead In early 2001, the MII, the SDPC, and the MOF without approval from the central government. jointly announced massive cuts in telecommunica- To address the problems between national and tions and Internet usage fees. Those cuts were fol- regional policies, provincial telecommunications lowed by a second round of cuts in July 2001. Local authority (PTA) offices were set up in 2001 to con- phone charges per unit dropped from being charged duct the regulatory function within each province. at three minute to one minute intervals, and the charge was set at 0.10 yuan ($0.018) per minute. Monthly service fees for fixed-line telephones also Performance of the Sector: Impact dropped from 24 yuan ($2.90) to 18 yuan ($2.17), of Reforms Prior to WTO Accession and fixed-line phone installation fees were elimi- The effects of reforms and the introduction of nated completely. Domestic long-distance charges competition, albeit controlled by the government, have been standardized at 0.70 yuan ($0.08) per are expected to include lower prices, better quality minute and additional charges on long-distance of service, high growth and telecom penetration calls will no longer apply. Long-distance charges for rates, and major changes in the structure of the calls to Hong Kong, Macao, and Taiwan were stan- industry. dardized at 2 yuan ($0.24) per minute,while the rate for international calls was set at about 8 yuan ($0.97) per minute. Initial installation fees have Decline in Prices been eliminated as well, and the price to lease lines Prices for basic telecom services have come down from China Telecom was lowered by 72.8 percent. considerably in the last few years, but more because On top of the mandated price cuts,several opera- of mandated price reductions than pure competi- tors continued to announce further rate cuts of up to tion. After all, the MII retains a considerable degree 40 percent for domestic long-distance calls, while of control over the sector through price controls as doubling the rates for local calls to maintain their well as through licensing. The MII justifies its con- operational profits. At the local level, individual trol of prices by pointing to its desire to maintain operators have underpriced themselves slightly stability in the market and avoid disruptive price when they have felt local conditions required it. wars.The procedure for telecom rate changes begins Most competition in pricing takes place through when an operator submits a proposal, which is fol- packaging mechanisms, which are not subject to the lowed by State Council department and ministry constraints of the established price ranges. 162 China and the WTO An example of price competition and its effect regularly reveal the financial conditions of the on reducing prices is evident in those services, such listed subentities, but a significant portion of their as IP telephony,where the government does not have businesses remains undisclosed. Corporatization price controls. There, intense competition has led has not severed the state connection, and it is not to reductions in the price of long-distance calls. A possible to discern exactly how much support they price war on IP telephony was introduced in 2001 still receive from their shareholders and how many by China Netcom. The IP phone card, a prepaid are actually sustained purely by generated revenues. device for long-distance calling, became very popu- Furthermore, many board members and top-level lar because it can save callers up to 70 percent on management continue to be closely connected with their long-distance call charges. CNC then further the state authorities. The support that these rela- reduced their charges by another 50 percent, pric- tionships could continue to generate, in any terms, ing IP domestic long-distance calls at 0.3 yuan is utterly unquantifiable. China Telecom, China ($0.036) per minute and overseas calls at 2.4 yuan Mobile, Unicom, and China Netcom claim that ($0.28) per minute. When the other operators-- they are no longer subsidized by the state and are China Telecom, Unicom, Jitong, and China able to run a stand-alone operation profitably. Mobile--followed suit and reduced their IP pre- Licensing and the forced divestiture of China paid cards, the government had to lower long- Telecom's mobile and paging components have distance rates in the face of such competition. reduced cross subsidization between products or Most of these rates appear to be artificially low interproduct subsidization, although this possibil- and to require significant rebalancing before the ity has not been eliminated entirely. sector can become fully competitive. IP card rates may be the only ones that closely reflect the true costs of these services. Growth and Competition in the Different Subsectors Profitability of the Industry Fixed-Line The number of fixed-line subscribers The resulting decline in prices for long-distance has increased substantially in recent years, espe- calls had a serious effect on China Telecom's over- cially in the post-1998 reform era. In 1996 there seas listing plans, because China Telecom holds a were 54.9 million fixed-line subscribers--a number franchise on fixed-line phone service and makes that had doubled to 108.8 million by 1999. By the most of its money from long-distance calls. China end of 2002 this number had almost doubled again, Mobile (listed in Hong Kong) and Unicom should with the number of users reaching 180.4 million. be able to benefit significantly from the changes By the end of 2003 the number of fixed-line users because both companies lease lines from China reached 210 million (figure 10.1) or slightly lower Telecom for their voice and Internet businesses. By than the number of mobile users. This increase rep- contrast, China Mobile, Unicom, China Netcom, resents a growth rate of 27 percent p.a. or per year and Jitong could be adversely affected by the over the 1996­2003 period. The penetration rates changes in Internet protocol telephony. With China of the population for fixed-line telephony have Telecom's rates falling, the discount margin for gone up from 10 percent in 1998 to 14 percent in using IP telephony has been narrowed significantly, 2003, still leaving considerable scope for growth. removing much of the attraction for using the serv- Even though Unicom and Railcom had licenses ice in the first place. In 2001 China Mobile's average to provide fixed-line services, China Telecom revenues per user fell by 35 percent, leading to a fall remained the dominant provider with 99 percent of in its share prices and those of Unicom's. The the market. additional users of mobile are not subscribers; they are using prepaid cards for which the cost of Mobile Phones Mobile telephony began to administration and use is much lower. increase significantly with the introduction of the It remains unclear how profitable each operator GSM (global system for mobile communications) is under current conditions. Those operators who networks in 1995. Domestic and international have partially listed on the stock exchange must roaming capability was introduced in 1996. The Telecommunications Services in China: Facing the Challenges of WTO Accession 163 FIGURE 10.1 Mobile and Fixed-Line Use in China, 1996­2003 (millions) Millions 300 250 200 150 100 50 0 1996 1997 1998 1999 2000 2001 2002 2003 (est.) Mobile Fixed-line Source: Ministry of Information Industry. number of mobile subscribers in China grew from scribers. In an effort to eliminate this type of 7.3 to 44.3 million over the 1996 to 1998 period, "arbitrage" opportunity between operators, the MII and surpassed the number of users in the United has begun to standardize. States in 2001, reaching 144.8 million (compared The issuance of additional mobile licenses is the with 120 million in the United States) (fig- subject of great debate. It remains unclear whether ure 10.1).8 Growth has continued, and the number the government will issue new licenses, despite indi- of subscribers reached 268.7 million by the end of cations that such plans are under way to increase 2003, and for the first time was slightly higher than competition in the sector and despite the need to the number of fixed-line users at 263.3 million. have new licenses issued to operators during the This increase represents a dramatic growth rate of transition to third-generation (3G) technology. 86 percent over the 2001­2003 period. As a result, However, it is unclear which technology standards the penetration rates have also increased. An and which 3G technology should be used. A time important contributing factor to the increase in the division synchronous code division multiple access number of mobile phone users has been the intro- (TDSCDMA) was recently created within China, duction of prepaid cards, which do not require sub- with standards parallel to those of the wideband scriptions or monthly user fees. code division multiple access (WCDMA) standard Of the only two licensed mobile phone opera- favored by European and Japanese companies. tors in China, both are state-owned companies. TDSCDMA was also approved by the International Although the market share of UNICOM has Telecommunication Union's telecom division as increased, China Mobile's share of the market is one of the three international 3G standards. still dominant at about two-thirds. In this mobile telephony market, the two participants in the Internet The number of Internet users9 in China duopoly have been reluctant to work together. For has also increased rapidly, rising 20-fold--from example, lack of number portability eliminates about 1.2 million in 1997 to reach 22.5 million in much of the incentive for consumers to switch 2000 (figure 10.2). By the end of 2002 the number service operators. Newly attained market share had more than doubled again, to 59.1 million. rarely comes at the expense of the incumbent; About two-thirds are dial-up users, and the instead, most of the uptake comes from new sub- remainder are connected through leased-line 164 China and the WTO FIGURE 10.2 Internet Users in China, 1996­2002 Millions Percent per year 70 180 160 60 140 50 120 40 100 30 80 60 20 40 10 20 0 0 1996 1997 1998 1999 2000 2001 2002 Millions Growth Source: Ministry of Information Industry 2003. connections or use both dial-up and leased lines. to convergence--is the evolution of Voice over Despite the high growth rates, China is a long way Internet Protocol (VoIP) services. Despite signifi- from being considered a "wired" nation. cant resistance from existing operators and the MII, The chief bottlenecks for further expansion of the possibility of providing reduced long-distance Internet use include high costs and slow speeds in costs, up to 70 percent less than conventional long- most areas. The maximum transmission rate of the distance rates over the fixed line provided by China telephone lines used by most people to access the Telecom, led to intense competition for the provi- Internet is only 56 kilobits per second, making sion of this service. In recognition that it would be downloading and uploading very time-consuming. difficult to control this development, MII legalized Although an integrated services digital network VoIP, and incumbents such as China Telecom have (ISDN) was introduced in 1998, it remains costly seen their revenues drop by an estimated $130 mil- and is still of insufficient bandwidth to allow for lion because of the increase in VoIP usage. This more sophisticated online usage and programs. increase in competition could come about very Shanghai has been able to benefit by using the quickly because IP networks can be built and more advanced asynchronous digital subscriber deployed more rapidly than other networks. The line (ADSL) services since the beginning of 2000.10 MII was forced to create a framework to allow the ADSL is not the only broadband technology avail- incumbents to provide these services as well. Cur- able. More mature technologies include hybrid rently, five of the six telecommunications carriers fiber/coax (HFC), digital subscriber line access are licensed to provide VoIP services. Foreign net- (DSLx), local multipoint distribution service work providers and local paging companies have (LMDS), and fiber-optic cable. HFC has to date come together in alliances to seek a share of the been applied in only a few selected pilot project lucrative market.11 locations. Much of the difficulty in getting this tech- nology installed in a wider area arises from issues of Competition and Improvements convergence. The HFC broadband access technol- in Quality of Service ogy is the cable modem, which allows transmission of data to computers and television sets at the same Five regulations governing the quality of services time. Regulatory reforms are required for this tech- have been promulgated by the MII in the last few nology to be approved on a nationwide basis. years related to periodic reports about the quality Perhaps one of the more interesting develop- of service, frequency of publication, and service ments in the industry--and a step on the way quality maintenance, including how to manage Telecommunications Services in China: Facing the Challenges of WTO Accession 165 TABLE 10.3 Impacts of Reforms on China's Telecommunications Industry, 1998­2003 1998 1999 2001 2003 Backbone optical cable line 180.8 194.1 340 (yuan per thousand km) Fixed asset investment (billions of yuan) 168.1 151.8 Telecom services business volume 226.49 313.24 350 (10 million of yuan) (est.) Price (yuan) June 1 Fixed (LAN) line installation (analog) 1,010 725 Free Free Domestic leased lines (digital, per month) For telecommunications services 9,000 3,000 1,500 For Internet services 4,500 2,000 1,000 International leased lines (per month) Hong Kong, Macao, and Taiwan (China) 27,400 2,800 Other Asia 28,900 14,000 Other 30,800 15,000 Domestic long-distance fee (per minute) 0.5­1.00 0.7 (depending (flat fee) on range) Internet connection fee (per minute) 0.02 Mobile network access fee 800 500 Free Users (millions) Fixed-line 87.4 108.8 180.4 263.3 Urban 62.6 74.6 171.3 Rural 24.8 34.1 92.0 Mobile phonesa 23.9 43.2 144.8 268.7 Fixed-line penetration (per 100 people) 10.5 13 18.9 21.2 Computers with Internet access 2 8.9 22.5 Competition Number of players 2 5 7 6b Mobile 2 2 2 2 Fixed-line 2 2 3 3 Data 1 2 2 2 VoIP 0 1 5 5 Market share of China Mobile (percent) 95 89 75 68 Market share of Unicom mobile (percent) 5 11 25 32 aVarious estimates say mobile phone use will hit 300 million by 2005. bAfter the breakup of China Telecom into north and south, with north and Jitong being absorbed by ChinaNetCom. Source: Various issues of China Daily and MII. complaints from customers. It is difficult to meas- Notices on complaint handling will be published ure whether the limited increase in competition to every three months. Service quality reports submit- date has led to improvements in quality. However, ted regularly by the telecom agencies to the MII will interviews do reveal that there has been a change in be published every six months. The results of sur- attitude at China Telecom, which is servicing its veys and evaluations of customer satisfaction will be customers better. published annually. All other information related Since 2000, the Telecom Consumer Complaints to service quality will be published irregularly and Handling Center under MII has made public according to circumstance. Recently, telecom com- all complaints made against telecom companies. panies were warned that if they do not provide good 166 China and the WTO service, they will be given a warning by the MII or WTO Commitments will face losing their license altogether. However, the China's WTO commitments in the telecommunica- effectiveness of this mechanism on the quality of tions sector open up the telecom market to foreign service remains to be seen. participation, and bring the domestic regulatory and business environment in line with interna- China's WTO Commitments tional standards (box 10.1). The commitments and Changes in the Regulatory have two main components. The first is removal Framework of limitations on market access by allowing right A series of reforms and regulatory changes pre- of establishment and removal of limitations on cededChina'saccessiontotheWTOonDecember11, national treatment. Foreign investment is allowed 2001, and reforms continue in line with meeting to enter, but initially at a lower ownership level and China's WTO commitments, which are to be com- with geographic restrictions. Over a period--two pleted over the next five years. However, delays in years for value-added services, five years for mobile completion of the regulatory framework prompted telephony; and six years for domestic and interna- some of China's fellow WTO Members to express tional services--the geographic restrictions are concerns during China's transitional review (WTO removed. However, foreign ownership is capped at 2003). 50 percent for value-added services and 49 percent BOX 10.1 Summary of China's WTO Commitments The main commitments are to open up for for- on national treatment. On right of establish- eign entry, which is currently prohibited. The ment, upon accession joint ventures will be opening up is undertaken in phases for geo- allowed in Beijing, Guangzhou, and Shanghai, graphic area and percentage of foreign owner- with foreign ownership of up to 25 percent. ship, and agreement to give national treatment Within one year, areas will be expanded to to foreign firms. China also has to abide with the include services in Chengdu, Chongqing, WTO's April 1996 reference paper on basic Dalian, Fuzhou, Hangzhou, Nanjing, Ningbo, telecommunications. Qingdao, Shenyang, Shenzhen, Xiamen, Xian, Value-added services (electronic mail, voice Taiyuan, and Wuhan, with foreign investment mail, online information and database retrieval, allowed up to 35 percent. Within three years of electronic data interchange, enhanced facsimile accession, foreign investment will be no more services, code and protocol conversion, online infor- than 49 percent, and within five years the geo- mation and data); paging services: no limitations graphic restriction will be removed. on cross-border supply except for right of estab- Domestic and international services (voice, lishment, no limitations on consumption abroad, packet-switched, circuit-switched, facsimile, domes- and no limitations on national treatment. On tic private leased circuit services, international right of establishment, upon accession joint closed user group voice and data services): no limi- ventures with up to 30 percent foreign ownership tation on cross-border supply except under right will be allowed to provide services in Beijing, of establishment, no limitations on consumption Guangzhou, and Shanghai. Within one year, this abroad, and no limitation on national treatment. arrangement will be expanded to Chengdu, On right of establishment, upon accession joint Chongqing, Dalian, Fuzhou, Hangzhou, Nanjing, ventures will be allowed in Beijing Guangzhou, Ningbo, Qingdao, Shenyang, Shenzhen, Xiamen, and Shanghai, with foreign ownership of up to Xian, Taiyuan, and Wuhan. The amount of foreign 25 percent. Within five years, areas will be investment is not to exceed 49 percent. Within expanded to include services in Chengdu, two years, there will be no geographic restriction, Chongqing, Dalian, Fuzhou, Hangzhou, Nan- and foreign ownership can be up to 50 percent. jing, Ningbo, Qingdao, Shenyang, Shenzhen, Mobile voice and data services (analog/ Xiamen, Xian, Taiyuan, and Wuhan, with foreign digital/cellular services, personal communications investment allowed up to 35 percent. Within six services): no limitation on cross-border supply years of accession, foreign investment will be no except under right of establishment, no limita- more than 49 percent, and the geographic tions on consumption abroad, and no limitation restriction will be removed. Telecommunications Services in China: Facing the Challenges of WTO Accession 167 for mobile telephony and domestic and interna- major elements of the regulations are described in tional services, indicating that a foreign majority is the following sections. not desired. The second component requires China to adhere to the WTO's April 1996 reference paper Defining Coverage The working definition used on basic telecommunications, which sets out prin- in the WTO schedule is that basic telecom service ciples on the regulatory framework for basic tele- encompasses local, long-distance, and international communications services. services for public and non-public use; may be pro- vided on a facilities basis or by resale; and may be provided through any means of technology (e.g., September 2001 Regulations cable, wireless, satellite). In anticipation of this Prior to China's formal accession to the WTO in working definition, the 2001 regulations defined December 2001, some observers had been antici- telecom services as services provided by means of pating the introduction of a telecommunications carrying, sending, or receiving sound, data, images, law, because work on one had been under way since or any other information through a hard-wired or 1986, with earlier attempts dating back to the early wireless system. This definition is intended to cover 1980s. However, the law was not announced prior broadcast networks, the Internet, and related to China's accession and provisional telecommuni- services in order to provide the legal basis for the cations regulations were issued in September 2001 "convergence"of information technologies (Horsley in line with China's WTO commitments in the tele- 2000). A summary of definition of coverage is pro- com sector.12 These regulations and some follow-up vided in box 10.2. regulations are a very important first step toward developing a comprehensive and pro-competitive Foreign Investment Foreign investment issues regulatory framework, but many issues of interpre- are not explicitly addressed in the regulations, tation, clarity, and implementation remain, and although a schedule for phasing in foreign owner- some areas have not been adequately addressed. ship was set out in China's protocol of accession to Moreover, as regulations, they do not have the full the WTO. In addition, Regulations for the Adminis- effectiveness of a true telecommunications law. The tration of Foreign Invested Telecom Enterprises BOX 10.2 Coverage and Definition of China's Telecommunications Sector Basic telecommunications services are fixed-line data processing and transaction handling, local and long-distance telephony; mobile tele- value-added fax, Internet access service, Internet communications including analog and digital information service, and videoconferencing. trunk systems, analogue cellular services, second An additional five categories were introduced (2G) and third (3G) generation mobile telecom- in 2001: fixed-line-based, value-added telecom munication services; satellite communications; services, including telephone information, pag- Internet, and information transmission; sale or ing, message recording, and restoring and video lease of bandwidth, wavelength, fiber optics, conferencing services; mobile phone value- cable, or any other network elements; network added telecom services; satellite value-added access and out-sourcing; international telecom- telecom services; Internet value-added telecom munications infrastructure and international services, including Internet access, data center, telephony business; wireless paging; and resale information, analog private network, videocon- of basic telecom services. ferencing imaging, and other Internet-based Value-added services refers to services in services; and other data transmission and which the supplier adds value to the information network telecom services, including computer by enhancing its content or providing storage information, electronic data exchange, messag- and retrieval. These include nine categories: ing, e-mail, fax restoring and forwarding, and e-mail, voice mail, online information database analog private network services. and retrieval, electronic data exchange, online Source: Telecom Service Classification Catalogue (annex to PRC Telecommunication Regulations, October 1, 2000) and Notice Regarding Adjustment of the Telecom Service Classification Catalogue, June 21, 2001. 168 China and the WTO (FITE Regulations) came into effect on January 1, Universal service is to be provided in a transpar- 2002. An additional MII directive was issued in early ent, nondiscriminatory, and competitively neutral January 2002. The Administrative Measures for way, and operators should carry out any relevant Telecommunications Business Operating Permits national provisions for universal service. The regu- defined changes in registered capital requirements, lations refer to some form of subsidy or cost recov- permit applications, and approvals processes, as well ery without being specific and linking it to the as operating and annual auditing requirements. competitive safeguards. However the regulations did not make clear how universal service will be Responding to the Requirements of the Refer- implemented. ence Paper The September 2001 regulations also Resource allocation of scarce resources such as incorporated many of the specific requirements on frequencies are to be done in an objective, transpar- creating pro-competitive regulations and adminis- ent, timely, and nondiscriminatory way. The regu- tration that are contained in the WTO Agreement lations in fact go beyond the reference paper by on Basic Telecommunications Services. The main referring to the establishment of an auction-based elements of pro-competitive regulations and proce- system of allocating telecom resources including dures contained in the regulations are as follows. frequencies and network numbers. Leading opera- Competitive safeguards are incorporated to pro- tors are required to ensure that such users can uti- hibit unreasonable cross-subsidization of other lize their network resources. However, it is not clear businesses, actions that limit subscribers from what form the implementing regulations will take. using services of other operators, and the provision Restructuring of the regulation and administra- of below-cost services. In addition, an operator is tion of the telecom sector in line with the reference not allowed to restrict subscribers to using only paper requirement to have an independent regula- services or equipment specified by the operator, tor has begun, but it is still unclear what the final refuse or delay service without justification, outcome will be and how long the process will take. increase rates or items for which it charges fees The reference paper calls for the regulator to be without the approval of the regulator or customer, separate and not accountable to operators and sup- provide misleading information, or refuse to carry pliers, and be impartial and nondiscriminatory in out its commitment to subscribers. dealing with all market participants. As for licensing, all telecom service providers In the September 2001 regulations the MII need a license and have to meet the transparency remained the institution designated "to oversee requirements and the criteria of eligibility for a and control the telecommunications industry of license. A time limit has been set for making deci- China." Surveillance and control of the sector were sions on licensing applications and for making avail- to abide by the following principles:"the separation able the reasons for denial. In issuing the license, the of administrative departments and enterprises; the issuing agency must consider its impact on national breakdown of monopolies and the encouragement security, network safety, the sustainability of tele- of competition, transparency, fairness and just com resources, environmental protection, and practice." The official position is that the opera- competition. The pending issues here are the imple- tional and regulatory functions were separated menting regulations and details on how the MII and through the creation of the MII in 1998 (WTO the provincial regulators will evaluate these factors 2003: para. 78). How objective and impartial MII and handle issues such as auctions. can be in dealing with China Telecom and China The provisions related to domestic interconnec- Mobile remains a source of concern, given close tion are in line with the reference paper in requiring associations between the incumbents and MII. major operators not to refuse other operators' In September 2001 the government set up a new requests for interconnection and the use of non- commission to decide on major policies for the discriminatory and transparent interconnection telecommunications sector and prepare the new procedures, including how disputes should be law. The MII will be in charge of implementing resolved. The operators must also "unbundle" the those rules and administering standards (Dow network elements they sell so that the supplier only Jones Newswire, September 20, 2001). The setting pays for the network components or facilities up of the commission is intended to lead to the required. required objective changes in policies and laws and Telecommunications Services in China: Facing the Challenges of WTO Accession 169 will pave the way for the independent regulatory China's State Administration of Taxation has set up body. However, since at present MII is still charged a strategy and research group to discuss the issue of with implementing the rules and administrative e-commerce taxation and conduct research on the standards, one has to wait for the implementing logistics of an e-commerce tax policy. The group's regulations and mechanisms to be in place to assess findings will be the basis for new e-commerce regu- how objective the outcomes will be. lations to meet the requirements of China's Inter- Other provisions in the telecom regulations net economic development, differentiate and stan- address the provision and quality of telecom serv- dardize business activities on the Internet, enhance ices, network construction, and security. Require- the government's role in the sector, protect the law- ments of greater transparency imply that there could ful rights and interests of enterprises and con- be a greater role for the nongovernment sector-- sumers, crack down on illegal business activities, operators, users, and other relevant parties--in maintain the socioeconomic order, and establish an providing input to rates, as well as implementation e-commerce operations registration system.14 of regulations, setting of standards, and so on. Recent Changes in Institutional Structure and Regulations Regarding Internet and E-Commerce13 Preparation of a New Telecommunications Law Given the rapid development of this segment of the The MII minister was appointed director of industry and of technology, regulations are still the State Council Informatization Office in June catching up with current conditions. Currently, 2003--the first time one person was placed in China has over 300 Internet service providers and charge of both MII and SCIO. Some say it is a step over 600 Internet content providers. Of the ISPs, closer to merging the two agencies to oversee the only 53 have MII approval to provide nationwide country's telecom, electronics, and information services, and the remainder are approved by the technology industry and become the new inde- local provincial telecommunications authorities, or pendent regulatory agency. However, at this stage it PTAs, to provide restricted services. The latest regu- is not clear how and when the regulatory agency lations require ISPs to obtain permits to do busi- will be set up. ness and require ICPs to register and record within Meanwhile, the telecommunications law was the MII/PTA system currently in place. ISPs must still being drafted in late 2003 (WTO 2003: para. no longer require foreign investors to find a 71); policy reforms and rapid changes in technol- Chinese partner in order to gain approval. ogy had forced revisions of the drafts before they Regulations issued in 2001 on Internet informa- were even issued. There have been various updates tion services, electronic bulletin board services, and and clarifications since the September 2001 regula- online news publishing specify registration proce- tions and they are going in the right direction. This dures for providers. Even though the rules exist, includes greater transparency such as specification local authorities have reiterated the rules, and in of deadlines for approval of licenses in basic tele- some cases extended the deadlines, for the neces- com (180 days) and value added services (60 days) sary approvals. In conjunction with the Ministry of and requiring the authorities to provide an expla- Public Security, the Ministry of Culture, and the nation for decisions. In August 2003 the State State Administration for Industry and Commerce Council announced its "opinion" on the broad (SAIC), the MII also issued in 2001 new measures framework for telecommunication market admin- for the "Administration of Places that Provide istration as part of the results of the commission set Internet Access Services"--that is, they govern up to look into telcom regulations.15 This included Internet cafes. Those cafes that have already been strengthening and facilitation interconnection pol- approved by the relevant authorities are now icy with fees to be based on costs incurred by the required to apply for reexamination and approval. operator, and sanctions and penalties on violation For example, Internet cafes may not set up within a of interconnection. This issue is important in light 200-meter radius of primary and middle schools of the disagreements on interconnection pricing and are forbidden to set up in a residential building. between mobile and fixed line networks. The scope Whereas for E-Commerce the issue of whether of the telecommunication law confirms that the to tax e-commerce has not been resolved, and authorities have convergence in mind, that so, that 170 China and the WTO FIGURE 10.3 Restructuring of China's Telecommunications Industry, 1994­2003 China Telecom CT (all fixed-line, south) MII China Telecom China Mobile CMCC (GSM, long-dist. IP) China Satellite CSAT (satellite comm.) Guoxin Paging Unicom Unicom (all, CDMA) China Netcom CNC (Internet, fixed-line, north) Jitong CRC (fixed-line, long-dist., China Railcom Internet) 1994 1998 2000 2002 Today Source: Compiled by authors. the framework law is intended to include opera- Jitong to form the China Netcom Communications tions of telephony, internet, and cable TV indicat- Group Corporation. The 21 southern operations ing that all of the subsectors will come under one retained the China Telecom name and operate as a framework law. However, even though the organi- separate entity. Figure 10.3 shows the restructuring zation and administration of the new regulatory efforts since 1994 up to the last change. authority was announced in broad terms, it was not Corporatization efforts to improve the perform- clarified how the responsibilities between MII and ance of the sector and raise capital continue. In SARFT will be reallocated. Other issues addressed November 2002 China Telecom listed the four with greater transparency and market based princi- profitable networks in Guandong, Jiangsu, ples in mind, were implementing regulations for Shanghai, and Zhejiang in the Hong Kong and market entry, distribution and administration of New York capital markets, raising $1.5 billion. scare resources (e.g. frequencies), service fees, qual- Meanwhile, China Netcom is still undertaking ity standards, and information security. internal reorganization in merging the three sepa- rate entities (China Telecom north, Jitong, and the original China Netcom), and, as a result, its plans to Restructuring: Government-Driven Mergers list shares on the capital markets have been delayed, and Consolidation with a public offering now planned for 2004. Other than regulatory changes, the authorities have also restructured the sector through government The Potential Direct Impact of rather than market driven mergers and consolida- Reforms and Outstanding Issues tions. In 2002 China Telecom was split along regional lines. Operations covering the northern 10 What is the expected impact of the current set of provinces were merged with China Netcom and reforms and WTO accession commitments on the Telecommunications Services in China: Facing the Challenges of WTO Accession 171 telecommunications sector? First, the sector will be decline in prices and the fall in monopoly rents challenged to be competitive and efficient before with the commoditization16 of fixed-line use. The the likely entry of foreign investment stemming only area in which prices could decline further is from China's WTO commitments, and it will face international rates, which will face increased pres- increased pressure to meet customer satisfaction, as sures to fall given the advent of VoIP. Perhaps a well cope with the restructuring that is under way more important concern is that most foreign tele- domestically. Second, the industry has to respond com operators are currently not in a position to to rapid changes in technology. Meeting these chal- undertake huge investments in China. lenges will require further restructuring of the sec- Thus it is unlikely that new entrants will enter tor, new investments from both the government the fixed-line subsector, except possibly in segments and private sector, and the transfer and develop- requiring low capital requirements. Increased ment of the appropriate technology and infrastruc- competition in fixed-line is unlikely to come from ture (hard and soft). The evolution of the sector liberalization or privatization, because the majority should be managed carefully, because there is the ownership is expected to remain with the state. danger of over investment and over capacity, which Rather, the priority will be on creating the regula- will be influenced by the regulatory environment tory framework, including the institution, to create facing the sector. The sections below address the the pro-competitive operating environment and potential competition and impact on the different complete the corporatization process so that China subsectors, and issues related to developing the Telecom and China Netcom will operate more effi- appropriate regulatory environment. ciently. The government-driven process of mergers and consolidations is likely to continue to be the modus operandi. The Potential for New Entrants As for mobile phone services, the rapid progress and Managing Competition of technology in this subsector, combined with It is unlikely that China would have opened up its declining costs and an infrastructure cost lower telecommunications sector to foreign entry with- than that for fixed-line telephony, make it attractive out the pressure it faced in bargaining for WTO to investors. Yet toppling the major barrier to entry, accession (Mueller and Lovelock 2000; Zhang obtaining a license, remains elusive. Licenses to 2001). Now, even though the schedule allows new operate mobile services, as with any other basic entry and permits foreign investors to come in over service, are issued at the discretion of the central the next few years under the WTO commitments, it authorities, and although the procedures for is unlikely that a flood of new foreign entrants will greater transparency in the issuance of licenses materialize. Because of the large size of the invest- were laid out in the September 2001 regulations, ment and the higher risks of an uncertain regula- there is no indication when new licenses will be tory environment, foreign interest will be limited opened. Additional licensing to operate mobile unless there is a credible commitment from the services has so far been linked to technology such as regulators to implement the planned regulations. 3G and standards such as those represented by In any event, foreign majority ownership will not UNICOM's CDMA, although a decision on the be allowed in key areas, and foreign ownership is technology standard has yet to be made. capped at 49­50 percent, with entry still based on The mobile telephone market is still a duopoly obtaining a license to operate. with very little competition in this subsector. Interviews with foreign telecom operators and Backdoor attempts at entering the market by, for industry people conducted from mid-2001 to mid- example, mobiles that operate from the fixed-line 2002 indicate that capping foreign ownership was network within limited areas have been halted not a major disincentive, but the lack of specificity indefinitely.17 Moreover, because the applications on the regulations for foreign investors was. An for value-added services that use the latest technol- important barrier to entry to fixed-line operators is ogy do not yet exist in China, technology alone will the dominance of the incumbent, China Telecom; not provide enough of a catalyst for users to switch. the lack of certainty surrounding interconnection Another problem for competition is that as yet to its backbone infrastructure; and the drastic there is no number portability--that is,users cannot 172 China and the WTO FIGURE 10.4 Share of Revenue by Operator, 2001 and 2002 Percent 60 50 40 30 20 10 0 CT CNC Unicom CMCC Railcom CSAT 2001 2002 Source: Ministry of Information Industry. switch services without switching their numbers. Internet data. The first broadband services system Therefore, the technology alone does not provide using television cable was recently introduced in enough of an incentive to switch services. The Daqing City in Heilongjiang Province. The opening attractiveness of mobile telephony is also reduced of the system on CATV represents a major break- because roaming charges are based on the caller pay through in the country's broadband progress. The party principle. benefits to the industry could be substantial, if the The main challenge facing the mobile industry in government can come to a final decision on how China--as everywhere--is the commoditization of fast it wants convergence to take place. call fees as the average revenue per user continues to decline. To find new revenue streams, mobile opera- Development of Internet and the New Economy tors need to seek out new partnerships with value- added service providers. One model is the coopera- Although certain barriers such as low credit card tion between Unicom and ICPs (e.g., Montro). penetration and an underdeveloped distribution Another value-added service that is booming is the infrastructure will exist for some time in China, short messaging service (SMS). Foreign and new Internet usage and everything associated with it, entrants are likely to take part in the value-added including e-commerce, are expected to continue to and specialized services segments of this subsector. grow rapidly. Although foreign companies will be Another issue is what will happen to the existing able to enter this market to provide value-added internal networks that developed in response to the services, it seems likely that partnerships between inefficiencies and high cost of using China Tele- the two will become the order of the day.The foreign com? Many ministries such as the State Power companies will provide the technology,brand-name Commission and the Ministry of Railways built reputation, and skills, and the Chinese companies their own internal networks. One of the important will provide the local connections, language abili- government-supported networks is the one devel- ties, and legal foundations for the new venture's oped by the agency that administers the state radio, existence. Again, however, as the recent changes in SARFT. The network already has 90 million users regulations have demonstrated, the lack of regula- and is planning to connect all existing networks. tory certainty, including on cross-border taxation, Finally, the government also plans to upgrade security, and content, have to be resolved before the the CATV network so that it can also transmit subsector can really take off. Telecommunications Services in China: Facing the Challenges of WTO Accession 173 Overall, then, China does not yet have the infra- required joint ventures between existing telecom structure needed to take advantage of the new econ- operators and new partners to seek new telecom omy, and the chief bottlenecks to such development licenses and to adhere strictly to the business cate- are less in the technology than in finance and logis- gory and geographic area of operation specified by tics. Much needs to be done in terms of developing the telecom license already in hand. these sectors of China's old economy before it can leapfrog its way into the new global economy. This Toward a Pro-Competitive Regulatory Frame- constraint also points to the need for a comprehen- work and Institutions Given the nature of and sive approach in developing the services sector. Cur- developments in the telecommunications sector, it rently, the only real possible use of new economy is unlikely that increased competition will come systems in China rests with the foreign multination- from liberalization or privatization. In any case, als that have already implemented various supply studies have shown that privatization will flow chain management systems, customer relations from the introduction of competition, but that the management systems,and other e-solutions systems reverse is not necessarily true (Petrazinni 1996; throughout their global operations. In many cases, Fink, Mattooo, and Rathindran 2001). In China, China continues to be the one country in which their inefficiency has stemmed not from government operations cannot be fully implemented. ownership per se, but from a lack of competition and ineffective government regulation (Stiglitz 1998). Given that the major operators in basic tele- Managing Policy Reforms and Increased com are state-owned and that the situation is Competition unlikely to change even in the short term despite The discussion in this chapter thus far clearly the liberalization of the sector through WTO acces- points to the importance of creating an appropriate sion as already outlined, the key challenge for regulatory environment to attain the emergence of China's telecom sector is to ensure that the few and healthy competition and an efficient telecommuni- predominantly state-owned players will be compet- cations sector. Some of the key regulatory issues ing in a pro-competitive environment. related to policies on convergence, government- The September 2001 regulations go some way mandated restructuring, consolidation of the toward responding to some of the obligations for state-owned enterprises, and the creation of pro-competitive regulations and institutions as set pro-competitive regulations are described in this out in the WTO agreement on basic telecommuni- section. cations, but there are still areas of uncertainty, issues of implementation, and lack of a timetable Cable and Convergence Issues China's cable tel- for implementation. Given the current political evision network now reaches about 90 million sub- institutions and structure in China, it is unrealistic scribers throughout the country, and it is expected to expect full regulatory independence in the near to eventually be the largest in the world. Major term. Not only does the Communist Party have recent regulatory reforms have enabled fixed-line overall responsibility for policy, but the MII still has operators such as China Telecom, Unicom, CNCC, a strong and complex relationship with the major and Railcom to offer cable television over their operators. This situation gives rise to specific and broadband networks. Cable TV companies were major challenges to the creation of a truly inde- also looking forward to being able to deliver Inter- pendent regulator (Zhang 2001: 47). If regulatory net and voice services over their networks. How- reform continues to proceed slowly, the impact ever, until there is an implementation schedule on of WTO accession on liberalization of China's convergence, explicitly indicating what is or is not telecommunications market could end up being permissible for cable operators, the cable industry limited. remains reluctant to see convergence because it The problems and the issues surrounding the fears broadband would significantly erode its cur- creation of an independent regulatory agency are rent subscriber base. not just confined to China, but also are being The MII has remained generally unsupportive replayed in most developing countries attempting of convergence. Regulations introduced in 2002 to open up their telecom sector. Therefore, careful 174 China and the WTO thought should be given to the sequence of steps infrastructure for telecommunications connectivity taken in developing such an institution. The first still does not exist, and mobile stations are limited steps should be adopting transparency and training to serving only the large cities. human resources adequately, as well as setting The objective of the universal service obligation administrative guidelines as to just what constitutes is to ensure that the public has access to affordable an "independent" agency. These measures could be telecommunications services, including people liv- followed by a step-by-step phase-in of a more inde- ing in high-cost service areas such as rural and pendent regulatory agency. remote regions as well as low-income groups. The authorities have not yet assigned the delivery of universal service to any of the current operators, Provision of Universal Service to Poor although the expectation is that China Telecom will and Isolated Areas18 remain the major provider for the less developed As in other countries, before reforms China Tele- areas and the poor regions. com was the only operator in China and had an The WTO reference paper on basic telecommu- obligation to provide universal service. The model nications services does not provide guidelines on adopted for China Telecom was the traditional universal service, although it does make clear that it cross-subsidization model in which subsidization should be implemented in an anticompetitive fash- flowed from the eastern and more developed ion, that it should be administered in a transparent, coastal areas to the western provinces of China and non-discriminatory, and competitively neutral from wireless and long-distance services to local manner, and that it should be no more burdensome services. The need to provide all households and than necessary for the kind of universal service small firms with telecommunications was used as a defined by the WTO Member. reason to defend monopolistic telecom companies. The experience of and lessons learned by other Furthermore,it was believed that profit-maximizing countries in implementing universal service pro- private firms would not go to poor and remote grams have yielded some caveats and models that areas. However, experience with protected state China could consider (World Bank 2000, Box 10.3). monopolies has shown that the outcome is likely to First, it is important that the definition of the uni- be one of inefficiency and underinvestment, with versality target is clear and that it is linked to realis- little or no service actually provided to the poor tic implementation measures, including how it areas. Indeed, many developing countries have paid should be funded. The universality target differs capital costs of $4,000 per line, or three to four times from country to country, depending on local eco- the achievable cost (Stiglitz 1998). Evidence also nomic and sector conditions, the distribution of suggests that liberalization in the form of greater the population, geographic considerations, and competition or privatization will generally increase security issues. Universal service refers to individual network availability (Petrazinni 1996, p. 37). or private access, whereas universal access refers to With the restructuring of the China's telecom community or public (shared) access (e.g., a phone sector, the authorities are shifting from the tradi- booth in every village), and in developing countries tional cross-subsidization model to the provision of the latter is more important. universal service by popularizing access through There are several common models for achieving significant price reductions. By means of such universal access: applying market-based reforms actions, it is hoped that telecom services will gradu- through privatization, competition, and cost-based ally become more accessible, despite the lack of an pricing; requiring operators,including new entrants, explicit universality mandate. The growth rates of to assume mandatory service obligations; imple- fixed-line and mobile phone users have indeed menting cross-subsidies between and within services been dramatic, leading to penetration rates of provided by incumbent operators; using access about 20 percent and 21 percent of the population, deficit charges paid by telecommunications opera- respectively, based on 2003 estimates. This average tors to subsidize the access deficit of incumbents;and is, however, misleading, because a much lower rate creation of a fund to provide for universal service of teledensity--only about 6 percent--prevails in (World Bank 2000). Other models and combina- rural areas and western provinces (MII statistics). tions of these approaches are available,but the choice Furthermore, in large areas of the country the basic depends on country-specific issues and conditions. Telecommunications Services in China: Facing the Challenges of WTO Accession 175 BOX 10.3 Experiences of Chile, Peru, South Africa, and Hong Kong (China) with Universal Access Chile neously on three projects and permitting bid- ders to bid on any combination of three proj- After market opening and privatization of its ects. In 2000 Peru modified its target to fund telecommunications sector, Chile introduced access to the Internet and other advanced serv- policy intended especially to provide universal ices, and to allow for funding the operation and access to people living in low-income and rural maintenance of designated services, not just areas. The supporting funds, which came from installation. the central government budget, took the form of targeted subsidies to private operators for the installation of public telephones in the unserved South Africa areas. The process included competitive bidding and the funds were not linked to the rollout obli- In South Africa, rollout obligations are imposed gations of operators to provide services in cer- on operators in underserved areas and com- tain areas Over the period 1993­99, a total of bined with a universal service fund. The fund is 183 projects to serve almost 6,000 localities generated by annual contributions from all were approved. Competition between bidders telecommunications licensees, and is used to reduced the actual subsidies paid, although provide direct subsidies to a targeted priority there were some delays in installation. The tar- population to subsidize the higher costs of gets are now to finance community Telecenters telecommunications and to subsidize the cost of that have access to the Internet and other new network rollout to underserved areas by opera- information and communications technology. tors. The network rollout, service quality targets, and pricing are closely monitored by the respon- sible agency. The funds have been used to assign Peru priority to the setting up of telecenters in part- nership with communities and donor agencies. Peru imposed rollout obligations as part of priva- The telecenters typically contain telephones, fax tization and established universal access and photocopy machines, personal computers, telecommunications funds, whose revenues and access to the Internet. were taken from the 1 percent of the gross rev- enues of the telecommunications sector. The funds were used to meet the target of extending Hong Kong (China) services to 5,000 unserved localities by the year 2003. Universal access was defined as services In Hong Kong, the cost-based universal service provided by public operators and available to a regime is funded through charges on external majority of users, including voice telephony, (i.e., international) traffic. The designated serv- low-speed fax and data, and free emergency ice provider has the obligation to provide PSTN calls. Criteria for the selection of localities were (public switched telephone network) access set as rural towns (300­400,000 population), service in Hong Kong and receives a fair contri- district capitals, and towns in high social interest bution from other licensees toward the net cost areas. Several innovations were introduced in of serving customers and providing public tele- the bidding process such as, for economies of phones. In 1999 the funds accounted for about scale, allowing a single operator to bid simulta- 1 percent of total sector revenues. Source: Kerf, Schiffler, and Torres 2001; Wellenius 1997. Cross-subsidization has traditionally been used rural regions, which currently lag far behind their when a state has one dominant provider. The disad- eastern coastal counterparts. The fund is intended vantages, however, are by now well known, such as to separate government administration from enter- inefficiency and not reaching the target group.China prise management in the sector, shifting responsibil- has used this model in the past, and so now must ity for infrastructure, operations, and fund-raising rethink how to deliver universal access. to the telecom companies themselves. Because the The MII plans to set up a universal service fund costs of these activities are high and have low that will be used for developing the western and returns, the MII is proposing tax incentives and 176 China and the WTO investment, financing, and human resource policies monopoly with large fixed costs to that of a set of for the industry, in tandem with the continued more complex and competitive activities. New opening of pricing issues toward full market force technologies have lowered unit costs dramatically. management.However,the plan is still under discus- Since the mid-1990s, there has been a tendency sion. In February 2003 the government announced toward a convergence of telecommunications, that it would reinstate the universal service fund computing, and broadcasting. New services, such as with would require contribution from all domestic the Internet, mobile, and wireless telecom, use operators to subsidize the construction of telecom different platforms and transmit through various infrastructure in China's rural areas. However the media, and they provide alternatives to traditional details on the size of contributions, how they will be wire line telephony at a much lower cost. These disbursed and managed remains unclear.19 services have developed beyond regulators' control The approach of using universal funds has the and can bypass the public switched network and advantage of being most effective in providing tar- control of the incumbent. These changes have also geted subsidies to expand or support uneconomic meant moves away from traditional forms of own- services, such as the basic telecommunications ership and the market structure of state monopoly services needed in the underdeveloped western to one of joint ventures, strategic alliances, and new provinces and rural areas. Indeed, it is potentially and foreign entrants (Petrazzini 1996, p. 24; Intven the most efficient and transparent approach, but 2000). In addition the justification of maintaining much depends on how it is structured and admin- state monopoly in telecom and undertaking cross istered. The scheme can be administratively com- subsidization has lost ground. plex, so that the transaction cost are higher than China's reforms in the last few years have been the subsidy, and there is the potential for bad quite dramatic compared with the situation in governance if accountability and monitoring other countries, and currently the telecom services mechanisms are not built in and implemented. sector in China is moving in this direction also. Furthermore, it is often difficult to predict the Increased competition has occurred, although it associated costs and revenues. has been "managed" by the government through Experience has produced important lessons for introducing new state owned players, breaking up anyone using the universal funds approach to pro- the state monopoly and restructuring as well as vide universal access. First, the target must be consolidation efforts, and through government clearly specified (e.g., high-cost regions such as the driven price reductions. Increased competition has western provinces of China, rural areas without in turn led to greater efficiency, price reductions any access, a low-income population). Second, the beyond the government mandates, competition financing, be it from direct government funding, through better service and product differentiation, contributions from operators, or proceeds from and greater penetration rates. Even though many privatization, must be transparent. Third, the body challenges lie ahead, considering the dramatic that administers and disburses the funds must be changes that the sector has witnessed since 1998 independent from the operators. It also should be and the unstoppable pace of global technological market-neutral in that it does not favor the incum- change, it seems reasonable to expect that the sector bent vis-à-vis new entrants, and it must seek com- will continue to liberalize. petitive bidding. Fourth, the subsidies given (i.e., in Whilst a number of foreign players have entered the form of a fiscal or investment incentive) should and will continue to enter value added services, as cover only the uneconomic portion of the cost of already discussed above, even after basic telecom providing the services, and the operators, whether sector is opened up under China's commitments, private or government, should finance the rest of competition will remain between domestic (which the cost from their revenue. will remain largely state owned) companies. There- fore to reach the desired level of fair competition and efficiency, and attract new investments, much Conclusions depends,however,on the ability of the authorities to In recent years the view of the telecommunications come up with a solid legislative framework that fos- sector has changed--from that of a natural ters a stable business environment by ensuring and Telecommunications Services in China: Facing the Challenges of WTO Accession 177 enforcing a level industry playing field and private take some time and the issue of how "independ- property rights,and how the state owned companies ent" the regulator will be from the domestic will be "corporatized" to be managed in a commer- operators, if the regulator emerges out of MII, cially viable manner. In 1999 the MII issued a state- will continue. However, it is important that ment that a telecom law would be finished prior to China addresses these issues by beginning with accession, but in December 2003 this law was still the introduction of greater transparency. For being drafted. Given that even the United States, instance, by establishing a clear and transparent considered to be the most wired nation, continues scheme within which licenses can be allocated. to have considerable difficulty with creating and The deadlines provided for consideration of updating regulations on various subsectors of the license approval are welcome, but more details telecom industry, it is perhaps not surprising that are needed with regard to the criteria for eligibil- the process of drafting this legislation has been so ity of licenses. time-consuming. A number of issues related to the new telecommunications law such as interconnec- Other than the regulatory framework and tion policy, the creation of an independent regula- implementation thereof to ensure greater trans- tory authority and convergence appear to be moving parency and a level playing field, another major in the right direction. However, a number of other challenges facing the sector of providing universal issues still need to be clarified: service. China is already moving in the right direc- tion by recognizing that separate policies are · Regulate and enforce a level playing field for pri- needed to address this social objective and not vate and foreign investment. Although hedging being burdened on the objective of developing an against risk is an inherent part of good business efficient and competitive telcom sector. China practice,the current environment creates a major should take note of the experience of other coun- disincentive for investment of the foreign and tries, and choose what will be appropriate for its private funds needed to develop the industry. conditions. Given China's size and vast differences · Consider implementation of less interventionist between provinces, more than one model is likely and government driven price mechanism, such as to work depending on the existing conditions and introducing price caps, that incorporate quality geographical location in question. However, the and other self-regulating factors. Current prices principles and lessons from other countries should reflect an effort to broaden the use of telecom- be borne in mind--especially with respect to it munications services, and significant sector and being market based, transparent and with the right geographic rebalancing will be required. But they incentive structure--when designing the policy. will affect cross-subsidization and will not reflect true costs. Appendix: Summary of WTO Reference Paper on Basic · Decide, like all other countries, which 3G tech- Communications (Annex to Fourth nology standard to adopt. The debate over Protocol of General Agreement on which 3G technology to use contributes to the Trade in Services [GATS], Agreement environment of uncertainty and has the poten- on Basic Telecommunications, tial to hold back investments. Effective January 1, 1998) · Adopt a clear and concise, but flexible, frame- work for the regulation of value-added services, China's WTO accession commitments in telecom- including e- and m-commerce platforms. munications oblige it to also adhere to the princi- · Devise anticollusion mechanisms given the like- ples of the WTO reference paper. The paper pro- lihood of mergers between the current major vides for a pro-competitive environment of policy market players. Such mechanisms become par- and regulation. ticularly important once the MII is relieved of its Competitive Safeguards control over the industry and pricing issues are left to market forces. Prevention of Anticompetitive Practices in Tele- · Establish an independent regulatory agency. communications Appropriate measures shall be Given past experience, it is likely that this will maintained for the purpose of preventing suppliers 178 China and the WTO who, alone or together, are a major supplier from Allocation and Use of Scarce Resources engaging in or continuing anticompetitive practices. Allocation of scarce resources such as frequencies, numbers, and rights of way will be carried out Safeguards Anticompetitive practices include, in in an objective, timely, transparent, and non- particular, engaging in anticompetitive cross- discriminatory way. The current state of allocated subsidization, using information obtained from frequency bands will be made publicly available. competitors with anticompetitive results, and not making available to other services suppliers on a timely basis technical information about essential Notes facilities and commercially relevant information 1. For the United States, European Union, and Japan, open- that are necessary for them to provide services. ing up telecommunications services was a major point of con- tention during their bilateral negotiations with China. China's rejection of foreign direct investment (FDI) in telecommunica- Interconnection tions services was unusual when compared with its stance in other economic sectors or with that taken by other countries at a Interconnection is to be provided under nondis- similar level of development. 2. About 40 foreign operators lost up to US$1.4 billion when criminatory terms, conditions, and rates, and in a the government cracked down on investments that circum- timely fashion, and, upon request, at points in vented the FDI ban in the sector. Despite this, the U.S. company addition to the network termination points offered AT&T recently entered into a joint venture to form Shanghai Symphony Telecommunications, and all other major foreign to the majority of uses. The procedures applicable operators remain in China, awaiting further entry opportunities. for interconnection to a major supplier will be 3. A decline in average revenue per unit without increasing made publicly available, and a major supplier will usage. 4. See Gao, Pin, and Kalle (2000); Perkins (2001), Price make publicly available either its interconnection Waterhouse Coopers (2001), World Bank (1992, 2000) and agreements or a reference interconnection offer. An Zhang (2000) for greater detail and background. independent domestic body will resolve intercon- 5. The lack of cooperation is not surprising, because Unicom nection disputes. was created by the rival Ministry of Electronics and an official in the State Council--not by MPT. Unicom's original objective was to share 10 percent of the fixed-line market and 30 percent of the mobile phone market with China Telecom by 2000 (Asian Com- Universal Service munications, September 1998). 6. Creation of the agency was part of the institutional reor- Any Member has a right to define the kind of ganization undertaken at the Ninth National People's Congress. universal service obligation it wishes to maintain. 7. In 1998 the government forced the People's Liberation Such obligations would not be considered anti- Army (PLA) to divest itself of all of its business interests as part of a government crackdown on corruption and forced a han- competitive as long as they are administered in a dover of Great Wall Telecom Corporation, which had the right transparent, nondiscriminatory, and competitively to operate a CDMA system, to Unicom. neutral manner. 8. As reported in China Daily, 5 August 2001. 9. The Chinese Internet Network Information Center (CNNIC) semiannual survey defines Internet users in China as Licensing Criteria Chinese citizens who use the Internet at least one hour a week. 10. ADSL employs the existing double twisted copper tele- The following should be made publicly available: phone lines to provide broadband network access, enabling a downstream rate of 8 megabits per second and an upstream rate all licensing criteria and the period of time required of 760 kilobits per second. However, transmission quality can be to reach a decision on a license application, the maintained only within a distance of 5 kilometers, thus requir- terms and conditions of individual licenses, and the ing one base site every 5 kilometers, and with heavy initial investment by telecom operators in network construction. Since reasons for denial of a license. ADSL was first introduced, charges have dropped from about 450­500 yuan to about 130 yuan per month. 11. Companies such as AT&T and Edge2net Inc. have already Independent Regulators entered the market. The latter has teamed up with China Motion Telecom to create Shenzhen China Motion Telecom. The regulatory body should be separate from and 12. The regulations have since been updated with a revised not accountable to any supplier of basic telecom- catalogue of value-added services in June 2001. munications services. The decisions of and proce- 13. See Baker and McKenzie (2000); Liu He, Qin Hai, and Lu Yanrong (2001) for more details. dures used by regulators shall be impartial to all 14. New regulations were announced in early 2003 that cov- market participants. ered the legal definition of E-Commerce as trading, production, Telecommunications Services in China: Facing the Challenges of WTO Accession 179 and services provided over the electronic network (e.g. informa- Liu He, Qin Hai, and Lu Yanrong. 2001."E-Commerce in China" tion search, ordering, and payment) and authentication of (First Draft). e-signatures. They did not as yet discuss the issue of taxation. Mueller, Milton, and Peter Lovelock. 2000. "The WTO and 15. State Council, Opinions Regarding Further Strengthen- China's Ban on Foreign Investment in the Telecommunica- ing the Supervision and Administration of Telecom Market, tions Sector: A Game-theoretic Analysis." Telecommunica- August 2003. tions Policy 24: 731­59. 16. A decline in average revenue per unit without an Perkins, Tony, and Steven Shaw. 2000. "China and the WTO: adequate increase in minutes of usage. What Will Really Change?" McKinsey Quarterly no. 2. 17. After essentially being cut out of the mobile market with Petrazinni, Ben A. 1996. "Competition in Telecoms-- its divestiture of CMCC, China Telecom tried to enter the Implications for Universal Service and Employment." mobile business by introducing a mobile phone that functioned Public Policy for the Private Sector. World Bank Group, essentially as an extension of its fixed-line networks. Therefore, Washington, D.C., October. the mobiles only functioned within a limited geographic area Price Waterhouse Coopers. 2001. "Briefing Overview of China around the line. Nevertheless, 2.5 million lines in about 50 cities Telecommunications Sector," February. were already installed when the MII, in response to the lobbying Qiang, C.Z.W., and Xu, L.C. 2000. "Reforming China Telecom," by two licensed mobile operators, ordered China Telecom to World Bank DECRG. cease such operations. The attraction for users was that handsets Smith, Peter. 1997. "What the Transformation of Telecom cost about half the price of normal handsets. Markets Means for Regulation." Public Policy for the Private 18. See Petrazinni (1996); Smith (1997a, 1997b); Stiglitz Sector. World Bank Group, Washington, D.C., July. (1999); and Wallenius (1997) for more information. Smith, Warrick. 1997. "Utility Regulators--The Independence 19. It was recently announced under the Rural Communica- Debate." Public Policy for the Private Sector. World Bank tion and Universal Services Initiative that, by 2005, 40,000 vil- Group, Washington, D.C. October. lages will be reached and each of the six major telecom operators Stiglitz, Joseph. 1998. "Creating Competition in Telecommuni- will be asked to cover certain regions: China Telecom (Shanxi); cations." Paper presented at World Bank Conference China Netcom (Inner Mongolia autonomous region); China on Managing the Telecommunications Sector Post- Mobile (Sichuan); Unicom (Guanxi autonomous region); China Privatization, Washington, D.C., April. Satcom (Sichuan); and China Railcom (Henan). Stiglitz, Joseph. 1999. "Promoting Competition and Regulatory Policy: With Examples from Network Industries." Washington, D.C., World Bank, Beijing, July. References Wellenius, Bjorn. 1997. "Extending Telecommunications Ser- vice to Rural Areas--The Chilean Experience." Public Baker and McKenzie. 2000. Asia Information Associates, Policy for the Private Sector. World Bank, Washington, E-Commerce in China. D.C., February. McKinsey Quarterly. China Statistics Yearbooks 1998­2000. World Bank. 1992."People's Republic of China Telecommunica- Beijing, China Statistics Press. tions Sector Study: Survey, Assessment and Strategy Recom- Fink, Carsten, Aaditya Mattoo, and Randeep Rathindran. 2001. mendations." Report No. 9413-CHA, February. "Liberalizing Basic Telecommunications: the Asian Experi- World Bank. 2000. "China: Services Sector Development and ence."Paper presented at the Conference on Trade,Investment Competitiveness." Washington, D.C., December. and Competition Policy in the Global Economy; the case of World Bank. 2001. "Seizing the 21st Century: Using Knowledge the International Telecom Regime. Germany, January. for China's Development." Washington, D.C., June. Gao, Pin, and Lyytinen, Kalle. 2000. "Transformation of China's WTO (World Trade Organization). 2003. "Report of the Meet- Telecommunication Sector: A Macro-perspective." Telecom- ing held on 5 December 2003: Note by the Secretariat." munications Policy 24: 719­30. Council for Trade in Services. WTO, Geneva. Intven, Hank, ed., "Telecommunications Regulation Hand- Zhang, Bing. 2001. "Assessing WTO Agreements on China's book." 2000. Infodev, World Bank. Telecommunications Regulatory Reform and Industrial Klein, Michael, and Philip Gray. "Competition in Network Liberalization." TPRC. Industries--Where and How to Introduce It." Public Policy for the Private Sector, World Bank. 11 Accelerating Financial Market Restructuring in China Deepak Bhattasali Close to the date of China's entry into the World a joint venture with one of the largest foreign life Trade Organization (WTO) in December 2001, insurance firms,which led to a further round of new several events appeared to perk up the domestic associations between Chinese financial enterprise financial market. A leading international bank groups and foreign insurance companies. A similar announced that it would charge a small monthly pattern of mergers, acquisitions, and partnerships fee if the amount in an individual bank deposit has emerged in investment banking, funds manage- fell below US$5,000.1 A leading local bank then ment, securities houses, and housing finance. In pointed out that about 20 percent of its savings another development, a leading foreign manufac- deposits were in the amount of 100 yuan or less and turing firm entered into an innovative agreement hinted that this situation was uneconomical. The with a wholly owned foreign bank for loans against China Bank Association then wrote a detailed its receivables and promptly used the funds to pay proposal that reportedly was discussed with the off outstanding credits to two large local banks. To People's Bank of China (PBC)--the central bank-- cap it all, data available during the first quarter of before submission for approval to the State Devel- 2002 suggested that a credit crunch had occurred in opment Planning Commission. The proposal rec- the second half of 2001, as banks tightened lending ommended that banks be allowed to levy fees for in order to improve their portfolios to meet the about 40 different services that traditionally they post­WTO accession challenge. had provided free of charge. By the standards of most financial markets, these A little earlier, fierce competition had broken out events were unremarkable, either singly or together. among Shanghai banks when they slashed buy/ However, to observers of China's financial market it sell margins on foreign currency transactions to was clear that something momentous was happen- 10 basis points (from an average of 40 basis ing; the genie was out of the bottle, competition was points in 2001), but order was quickly restored about to break out! Members of China's moribund by the Shanghai Banking Association. In parallel financial community, especially bankers, saw these developments, a foreign-invested bank and a fully events as the first signs of the imminent destruction owned foreign bank received their first Chinese cus- of the domestic financial system, justifying the fears tomers. Meanwhile, a major local financial con- of protectionists who had argued against many of glomerate opened a new life insurance company in the liberal measures included in China's accession 181 182 China and the WTO agreement (see table 11.2). Against this backdrop it and the successful nationalization of capital flight is interesting to assess the following issues: How did through the creation of an onshore foreign currency the accession agreement threaten the status quo in deposit and loan market. Uncertainties about the the domestic banking market?2 If the policymakers solvency of pension plans, the absence of a mort- were aware of the threat, why did they agree to take gage market (until 1999), and funds for education such radical measures to permit the entry of foreign needs and purchases of other services are also banks? What are some likely scenarios for the evo- believed to have stimulated savings in the form of lution of the banking sector in China during the bank deposits. However, this proposition has not yet period of accession and beyond? been tested rigorously. The next section of this assessment describes the structure and behavior of the Chinese banking Financial Sector Reform industry at the time that the finalWTO-related nego- tiationswereunderway.Itisfollowedbyadescription China's attempts to modernize its financial market of China's WTO accession commitments. The final have been steady, but, beyond an occasional flurry section examines the likely impact of foreign entry of precautionary measures introduced after the on China's banking sector and related policy issues. mini-crises of the 1990s, they have focused mainly on institutional diversification and strengthened administrative oversight. It is possible to distin- Initial Conditions guish four distinct phases of institutional reform in The post-1978 economic reforms in China--when China in the 22 years leading up to WTO accession. agriculture and industry began to grow rapidly The first phase (1979­86) was characterized by in response to improved incentive systems--led the breakup of the mono-bank system. PBC to a large increase in incomes, savings, and bank became the central bank, and the Industrial and deposits. For the most part, deposits were channeled Commercial Bank of China (ICBC) was created to to state-owned enterprises (SOEs) through the state- handle urban commercial banking, thus joining owned banking system, with insufficient regard for other specialized institutions for large-scale con- credit quality. Although investment decisions were struction (China Construction Bank [CCB]), for- controlled by the central government in theory, in eign exchange transactions and international trade practice subnational governments enjoyed remark- (Bank of China [BOC]), and rural lending (Agri- able degrees of freedom in choosing investments cultural Bank of China [ABC]). Foreign banks were and appropriating finance from the local branches initially allowed to open representative offices, but of the state banks. by the end of this phase they also were permitted to The banking system, and later the securities mar- establish commercial branches, albeit with tight ket, played an even more important role in support- geographic, product, and customer restrictions. ing the level of economic activity in the early 1990s, The second phase (1987­91) was marked by a as the government budget became increasingly con- reduction in administratively governed specializa- strained by a drop in revenues. Revenues fell from tion among institutions, the rapid growth of non- about 34 percent of the gross domestic product bank financial intermediaries, the establishment of (GDP) in the early to mid-1980s to below 12 per- the first joint stock universal bank (Bank of Com- cent by the mid-1990s. However, the legacy of poor munications [BOCOM]), the establishment of two loans began to affect the profitability of banks state-owned insurance companies (in addition to acutely by the mid-1990s, although the propensity the People's Insurance Company of China [PICC] of Chinese households to save a large part of their established in 1949), and the beginnings of a capital income (averaging 28 percent of GDP), the relative market by the introduction of secondary market absence of alternative savings vehicles, and the trading in government securities. implicit deposit guarantees at the state banks pro- The third phase (1992­96) saw even greater diver- vided high levels of liquidity to the banking system, sification in the financial market. Stock exchanges helping to mask its problems. The rapid growth of were established in Shanghai and Shenzhen, the household savings in the domestic banking system interbank market was developed,some interest rates has also been possible because of capital controls became flexible, and life and non­life insurance Accelerating Financial Market Restructuring in China 183 licenses were extended to foreign firms, but with (e.g., mortgage loans and automobile finance to tighter product and geographic restrictions than for accompany the rapid growth of newly created hous- domestic providers. ing markets and the anticipated increase in car buy- In the final pre-WTO accession phase of reform ing as their prices dropped after China's accession to (1997­2001), the emphasis shifted to addressing the theWTO).Moreover,the pressure to improve corpo- portfolio problems of the commercial banks and rate governance within financial institutions and governance of the financial market as a whole. In a increase the transparency of financial information sense,the preparations forWTO accession effectively provided by listed companies and banks was notched began during this period. They entailed restructur- up, using a combination of sanctions, personnel ing the PBC, attempting to improve bank supervi- movements,moral suasion,and new regulations. sion and the recognition of bad loans in the portfo- lios of state banks, creating four asset management Preaccession Situation companies(tosegmentaportionof suchloans),clar- ifying the roles of the China Securities Regulatory Despite these measures, China's financial services Commission(CSRC)andtheChinaInsuranceRegu- industry seemed far from ready for increased com- latory Commission (CIRC), and diversifying finan- petition from abroad at the time of WTO accession cialinstrumentstomeettheemergingneedsof savers (table 11.1).3 The size of the Chinese financial TABLE 11.1 China's Financial Services Industry, 2001 Size (Trillions of Yuan) Share (%) Type of Institution Assets Loans Deposits Assets Loans Deposits Deposit money banks 15.0 10.7 13.1 100.0 100.0 100.0 State banks 10.1 7.2 8.7 67.3 67.3 66.4 Other banks 1.9 1.2 1.6 12.7 11.2 12.2 Foreign funded banks 0.4 0.2 0.01 2.7 1.9 0.1 Urban credit cooperatives 0.8 0.6 0.8 5.3 5.6 6.1 Rural credit cooperatives 1.6 1.3 1.7 10.7 12.1 13.0 Finance companies 0.2 0.2 0.3 1.3 1.9 2.2 Other deposit institutions 1.1 1.4 0.2 Size (Trillions of Yuan) Prevalence Assets Premium Income Depth (%) Penetration (Yuan) Insurance companies 0.46 0.21 2.2 169 Trading Volume (Trillions of Yuan) Overnight 7 Days 30 Days 90 Days Total Interbank market 0.10 0.56 0.04 0.005 0.81 Size Market Capitalization Turnover Trading Volume (Trillions of Yuan) (Trillions of Yuan) (Trillion Shares) Securities market 4.4 3.8 31.5 Size (Trillions of Yuan) Share (%) Treasury Financial Corporate Treasury Financial Corporate Bond market 1.95 0.69 0.03 73.2 25.7 1.1 Sources: People's Bank of China 2002­03; Almanac of China's Finance and Banking 2002, Yearbook of China's Insurance 2002. 184 China and the WTO market was remarkable. Total deposits in banks condition of their portfolios or in balance sheets reached the exceptionally high level of 150 percent and income statements can, at best, help only in of GDP in 2001. China had the second-largest secu- understanding broad changes in the direction of rities market in Asia, with 1,189 listed companies, these parameters over time. However, the informa- about 60 million brokerage accounts, and a market tion available on the operating practices of Chinese capitalization approaching 45 percent of GDP. The financial institutions is sufficient to develop a set of four large state banks--which had been trans- characterizations to describe the state of the major formed into universal banks and constituted the industry segments on the eve of WTO accession.7 core of the financial system--accounted for 67 per- Each of China's four state banks is big--they have cent of deposit bank assets and 56 percent of total an average asset size of more than $400 billion, aver- financial assets. age employment of over 415,000 people, and from Nearly 77 percent of the annual financial savings 15,000 to 58,000 branches (see Almanac of China's of households was deposited in the banking system Finance and Banking 2002). These numbers suggest during 1998­2001. Consequently, banks handled that, at least in terms of resources, they are large most of the finance available to enterprises and the enough to compete against most foreign banks,even government, of which the share held by small and after factoring in the potential entry of major inter- medium nonstate enterprises--the most dynamic national players that have grown rapidly in recent segment of the Chinese economy--was tiny. More- years through aggressive mergers and acquisitions. over, the rural finance system was malfunctioning, The core institutional problem, however, is the with the Agricultural Development Bank of China relative lack of profitability and insolvency of the (policy bank), the Agricultural Bank of China, and four state banks, which undermines the many bene- the rural credit cooperatives facing the severest fits that could result from size, incumbency, and repayment problems in the banking sector. During reach.8Furthermore,althoughimproving,thebanks 2001,total bank loans expanded by 1.3 trillion yuan, have limited capacity to make lending decisions while both new equity issues and corporate bond based purely on the commercial appraisal of repay- issues were a negligible fraction of this amount. ment ability and credit risk. Until recently, the PBC- Corporate (i.e., state enterprise) sector leverage was regulated spread between deposit and lending rates very high; equally, a high proportion of loans taken (about 400 basis points, which are high by interna- by state enterprises was nonperforming. Govern- tional standards but not unusual by developing ment estimates placed the nonperforming loans country standards) was sufficient to sustain the (NPLs) of the four large state banks at about 27 per- reported profitability of banks, but it removed cent of their loans outstanding, which excluded the incentives to extend credit to underserved but rela- NPLs transferred in 2000 to four asset management tively dynamic segments of the economy, such as companies (nearly 20 percent of loans outstanding small enterprises. The competitive prospects of the at the four banks at that time). Unofficial estimates state banks are further undermined by their relative of the remaining NPLs in the state banks have run as inability so far to implement efficiency reforms-- high as 40­50 percent of loans outstanding.4 In for example, by cutting staff as much as required addition, it is likely that the other domestic banks or closing unprofitable branches. Since 1997, the have a larger proportion of NPLs.5 However, neither state banks have reportedly closed 29 percent of the application of existing loan classification sys- their offices and cut staff by 15 percent; not only are tems nor the current state of management informa- these measures considered insufficient, the feasible tion systems within the banks gives analysts any rea- limits under current ownership conditions seem son to believe that the various estimates present an to have been reached. Thus, as of the end of 2001, accurate picture of the condition of bank portfolios. the pace of innovation in retail banking--where Current accounting, auditing, and reporting such banks would normally be expected to have standards and practices do not provide sufficient an advantage--was unsatisfactory, despite the information for making judgments about the prof- fact that such services are relatively underdevel- itability of China's financial institutions, the ade- oped in China, where they have massive potential.9 quacy of their capital bases, and the efficiency of Finally, these banks' lack of experience with inter- their operations.6 The information available on the national banking--and their unfamiliarity with Accelerating Financial Market Restructuring in China 185 internationally benchmarked banking practices, in a flight of deposits to the big banks, the nonstate technology, and management--is a handicap.10 banks saw an increase in deposits of 77 percent over If information about the state banks is scanty, the same period. The same is true of loans, with that on the other parts of the banking industry, with these banks accounting for 42 percent of the the exception of the four listed commercial banks, is increase over the period, compared with 21 percent even more deficient. The published data do not for the state banks. Agility in trawling for savings in engender much confidence, and little systematic urban areas, combined with deft exploitation of the research has been undertaken into their operating non-SOE segment of the corporate market, pro- conditions. In general, however, it is believed that duced such outcomes. Both areas of concentration the urban and rural credit cooperatives (as well as represent the fastest-growing parts of the Chinese some city commercial banks) are insolvent and economy. Therefore, it is likely that, unless they mostly illiquid, with periodic reports of failures prove to be equally dexterous, the state banks will among them. In addition to poor management and have a smaller market share in 2006, by which time banking practices, the rural credit cooperatives have all of China's existing WTO commitments on bank- been affected adversely by the stagnation of the ing liberalization are expected to come into force. rural economy since the mid-1990s. Although the On the eve of WTO accession, the 191 foreign ABC generally withdrew from subcounty-level banks in China operated almost as off-shore or banking in 1996, leaving the field clear for the coop- enclave banks.12 The foreign bank penetration rate,if eratives, the franchise value of rural banking did measured by the number of banks or branches, not rise, and may have declined in the face of large- seemed quite high--52 percent and 14 percent of scale rural distress. the total Chinese banking industry, respectively. The other development of note in the period However, they accounted for less than 3 percent of immediately preceding accession to the WTO was assets, 2 percent of loans, and about 0.1 percent of the transfer of a significant portion of directed deposits. Despite these low numbers, several foreign lending to the three policy banks.11 It is not imme- banks were able to become profitable within diately apparent that this transfer reduced the level 12­18 months of entering the Chinese market, of credit risk in the banking industry, because there although market reports also suggest that they had is little evidence that these banks are better placed nonperforming assets equal to 20 percent of their than the state banks to evaluate such risks. More- totalassets(andsomeashighas90percent),andtheir over, directed lending has played an important role business was almost exclusively with enterprises with in the support state banks have given to the fiscal foreigninvestment.Becauseforeignbankswerebeing stimulus program introduced in 1998 and currently limited mostly to foreign currency transactions-- under implementation. Finally, because the two commanding a 24 percent market share in foreign main policy banks (China Development Bank and currency loans and 45 percent share in international China Import and Export Bank, or Eximbank) fund settlements--such banks had co-located with their themselves mainly through the issuance of long- clients, the overwhelming majority of whom were term local currency bonds, the majority of which foreign-domestic joint ventures in the coastal belt of are sold to the state banks, covariant risk exists China.Only a handful of banks--mainly with strong among the major players in the banking system. Hong Kong (China) or other Asian institutional Despite these weaknesses, other segments of the presence--had developed longer-term strategies banking market show potential for growth and aimed at other local market segments. The banks innovation. Although deposit growth at the state operated in an environment that was radically differ- banks continues to be strong--rising nearly 26 per- entfromthatforthedomesticbanks.Theyhadtodeal cent during 1999­2001--joint stock and other with onerous location, product, and investment commercial banks, with a significantly smaller restrictions,andavastlydifferentsupervisoryandtax reach in terms of branch networks, have made rapid regime (foreign banks are taxed at a 33 percent rate, gains. Thus, despite the implicit guarantee that whilelocalbanksaretaxedata50percentrate).Over- deposits at state banks are thought to enjoy, and the all, few linkages exist between foreign and domestic widespread belief among banking analysts that con- banks--which limited technological spillover and cerns about financial sector solvency would result competition in the domestic banking industry. 186 China and the WTO China's WTO Commitments At the time the Chinese government was negoti- in Financial Services ating its specific commitments, a significant body of literature had been building up on the effect of for- China's commitments to liberalizing its financial eign entry into developed and developing country services industry, and indeed its General Agree- financial markets. Issues related to the survival of ment on Trade in Services (GATS) commitments as domestic financial institutions and the possibility of a whole, are possibly the most radical offerings ever more complex macroeconomic risks and manage- negotiated in the history of the WTO and in the ment demands were known and debated. At the Uruguay Round (see table 11.2). Although haz- same time, the evidence of a favorable impact on ardous, it is interesting to speculate briefly about financial market efficiency was equally impressive, the possible motivation for such a fundamental especiallybecauseitwasbasedonverycarefulanalysis attempt to liberalize the financial services industry. TABLE 11.2 China's WTO Commitments in the Financial Services Sector Subsector Commitment Banking Location: Upon accession, foreign currency business is allowed without geographic restriction. Geographic restrictions on the local currency business of foreign banks will be phased out over five years; four cities will be opened upon accession, four additional cities thereafter. Products: Within two years of accession, China will permit foreign banks to provide local currency services to Chinese enterprises; within five years, to all Chinese. Investment: Within five years of accession, all current nonprudential measures regarding the ownership, operation, and establishment of foreign banks, as well as those concerning their branches and restrictions on issuing licenses, will be eliminated (national treatment). Securities Product: Foreign securities companies may engage directly in B share business. Investment: Within three years of accession, foreign investment banks will be permitted to establish joint ventures, with foreign ownership not exceeding 33 percent, to engage (without Chinese intermediary) in underwriting domestic shares (A shares) and underwriting and trading in foreign currency denominated securities (B and H shares, government and corporate debts). Representative offices of foreign securities companies may become special members of Chinese stock exchanges. Fund Investment: Upon accession, the establishment of joint venture fund management management companies will be permitted, with foreign ownership not exceeding 33 percent, to conduct domestic fund management business. Foreign investment shall be increased to 49 percent after three years. Insurance Location: All geographic restrictions will be lifted in three years after the entry. Product: Upon accession, foreign life insurers will be permitted to provide individual (nongroup) life insurance services. Two years after entry, they will be permitted to provide health insurance, group insurance, pension insurance, and annuities to Chinese and foreign customers. Reinsurance is completely open upon accession, with no restrictions. Investment: Upon accession, foreign life insurers will be allowed to hold 50 percent ownership in joint ventures. They may choose their own joint venture partners. For nonlife, China will allow branching or 51 percent foreign ownership upon accession and wholly owned subsidiaries in two years after the entry (i.e., no restriction on the form of enterprise establishment). Licenses will be granted solely on the basis of prudential criteria with no economic needs test or quantitative limits on the number of licenses granted. Source: Ministry of Commerce 2002. Accelerating Financial Market Restructuring in China 187 of country experience and microdata.13 China's and highlights the need to manage the transition decision to join the WTO was a clear attempt to actively. benchmark its economic activities to international standards and to ensure that the future growth of Major Issues and Implications the economy was along the lines of its comparative advantage. Within this broader context, and given This concluding section examines briefly some of certain institutional and political considerations rel- the major banking issues related to WTO accession evant to the financial sector, the evidence on foreign that will have to be addressed in the transition to entry and efficiency seems to have been compelling. market-based finance. It is useful to start with the The basic impulse for financial sector liberaliza- likely effect on general economic management, tion seems to have been based less on the effects it because accession to the WTO is likely to result in a was likely to have on the mobilization of a greater significant remapping of the flow of funds in the volume of financial savings and the promotion of financial market. a higher level of investment--which in standard Interestingly, many of the factors traditionally models appears to lead to a temporary increase in thought to have permitted relatively uncomplicated the aggregate growth rate of the economy--than financial and macroeconomic management, despite on the need for technical innovation that would inadequate economic information, weak accounta- lower the cost of financial services and permanently bility of institutions, and distortions in market sig- boost growth. Moreover, in recent years Chinese nals, are likely to become high-pressure points in policymakers have expressed greater concern about the transition period. They include control over the efficient allocation of resources in the economy. major interest rates, capital flows, and the level of This concern has manifested itself in pricing, com- the exchange rate. The ability to capture and retain petition, and fiscal measures, as well as financial sec- within the financial system a large proportion of tor reform. The rapid liberalization of many real household savings, then to channel it to enterprises sectors--for example, housing and other services-- and government directly or indirectly to support provided the opportunity to apply a similar compe- fiscal policy with quasi-fiscal stimulus through the tition policy model in banking reform, although less financial system, is likely to be affected the most. As so in other financial market segments which, foreign banks and other financial institutions are because of their infancy, attracted stronger protec- allowed to offer a broader range of domestic and tionist sentiment. foreign currency savings instruments, the excess Equally important, given the size of the domi- liquidity of local banks will shrink. If expectations nant players in the banking market, the informa- about the selective credit reach of foreign institu- tion gaps and span of control problems that tions are realized, the liquidity position for meeting accounted for their ineffective management, and the burgeoning needs of the domestic banks' large the difficulties of severing or weakening the links customer base is likely to tighten. Finally, the need between SOEs and banks, fairly strong pro- to correct disparities that have resulted from using competitive measures had to be employed. Foreign an unbalanced growth model to promote coastal entry, suitably phased to provide a manageable provinces represents a special challenge. The effects adjustment process, addressed many of the per- of geography and favorable government policies ceived problems with the financial services indus- toward such provinces have been reinforced by lop- try in China. The expected diversification of finan- sided financial sector development, which discrimi- cial instruments for saving, liquidity management, nates against the growth of the interior and western and financing of economic activity would also provinces of China.14 The geographic liberalization support the transition to market finance. Guid- built into China's WTO accession commitments ance from the recent Sixteenth National Congress will tend to foster more rapid development of the of the Communist Party on active promotion of financial services industry--and the benefits to the private sector is bound to be translated into savers and investors that stem from it--in these demands for a significant redirection of debt and same regions. equity flows. More significant, it sets China inex- A central issue related to financial stability and orably on the path toward market-based finance the future of the banking system is the fate of the 188 China and the WTO state banks. Clearly, rapid development of the non- because their high operating costs and inability to bank financial market will erode the role of all reduce NPLs stem as much from weak management banks--a particular concern with respect to their as from domestic political economy pressures.15 ability to capture household savings and maintain Despite recent attempts to cut the number of lending in the face of declining levels of liquidity. branches and attract new sources of equity (as, for This problem could, of course, be mitigated partly example, in the recent initial public offering by the by removing gradually and according to satisfactory BOC), it is unlikely that small changes in macro- prudential control the current investment restric- economic parameters, banking behavior, or the tions on banks--for example, by permitting pru- level of nonperforming loans will restore the health dent participation in equity markets and investment of their balance sheets. funds. However, the recent portfolio problems Extrapolations from published data on the state of banks will also limit their future risk-bearing bank income statements and balance sheets and a capacity, and possibly their ability to compete with variety of assumptions about macroeconomic foreign banks on the development of new financial parameters are shown in box 11.1. They suggest instruments. State banks are particularly vulnerable that, without fairly radical actions to reduce BOX 11.1 Modeling State Bank Performance: Can the State Banks Survive Liberalization? Within the limitations of the data, it is useful to case), the banks as a group are likely to still assess the prospects of China's main banks after have negative capital by 2007. the post-WTO accession period. The official and · Competitive pressures, either in the form of industry response to banking system liberaliza- rising interest rates to protect deposit bases tion hinges critically on whether the banks are or shrinking loan margins, would worsen the able to earn their way out of their current diffi- base case results. culties while facing foreign entry and associated · The group results are not affected significantly changes. The results of a simple financial projec- by small changes in the assumptions about tions exercise, which draws on publicly available macroeconomic growth, inflation, or the information on the condition of the portfolios of operating costs of banks. However, the effect the four state-owned commercial banks, their on individual banks could differ substantially. accounts, and assumptions about key macro- Also, even large changes in macroeconomic economic parameters, suggest the following variables--for example, higher credit growth broad trends: resulting from 50 percent higher output growth in the economy--will be inadequate · By the end of the accession period in 2007, the for stopping the erosion of the capital base of banks will not, either individually or as a group, the banks. be able to cope with their current and prospec- tive nonperforming loans (a level of 30 percent The conclusion to be drawn from this simple is used for the base case, ranging from 25 to 40 financial modeling exercise is that the state percent for the individual banks). It is likely that banks will come under severe pressure during all but one bank will have negative capital by post­WTO accession liberalization. Fairly sub- 2007, and that, as a group, the four banks will stantial changes in operating costs would be have negative capital. required to improve the condition of the banks. · The operating margins are too low through- Failing this, because they are unlikely to be able out the post-accession period, and not only to simply "grow out of their problems"--that is, do the banks not make profits during the they will be unable to recapitalize themselves transition period, but their capital turns even from provisions and retained earnings--a recap- more negative over time. italization of the state banks from outside · Changes in assumptions about the current sources is likely to be required. A combination of stock of nonperforming loans make a signifi- instruments most probably will be used to cant difference to the results of the base case. restore the health of the state banks' balance However, within a zone of 20 percent around sheets. These results support broadly the find- the base case assumption (about 5 percent- ings of other, similar modeling exercises related age points above and below the 30 percent to China's state banks (see, for example, Li and nonperforming loan level assumed in the base Ma 2001). Accelerating Financial Market Restructuring in China 189 operating costs, the banks will be unable to earn 8. Given their role and their dominance of the financial mar- their way to improved health. In such an event, to ket, the current health and future performance of the state banks cannot help but be a major source of uncertainty about the avoid a loss of confidence in the state banks or a prospects for development of the financial services industry and sharp erosion of their competitive position with China's economy. respect to foreign and smaller domestic banks, 9. Most indicators--number of checking accounts, share of personal loans, mortgages or automobile finance in total credit financial reengineering will have to accompany the or normalized by GDP, automated teller machines per capita-- needed changes in banking practices. Toward this show that Chinese banks lag substantially when compared with end, it is possible that bolder actions to inject new those in neighboring East Asian countries. By contrast, China's retail banking through the postal system has expanded rapidly, capital from private and public sources will have to managing over $46 billion in 110 million accounts at the end of be taken. Private capital would be preferable, espe- 1999 through a network of 66,649 postal branches nationwide. cially if it is bundled with access to new banking The Postal Savings Bank, reconstituted in 2000, has established technologies and financial expertise, but many transactions links with more than 17 countries and boasts a bet- ter electronic network than most state banks. See Leong (2000). market analysts believe that the use of public capi- 10. Almost all of the expertise was located within the BOC, tal is unavoidable. If so, the internal reform of the which accounted for 80 percent of China's financial sector banks should be speeded up and strengthened, so abroad, 40 percent of international trade settlement, and 85 per- cent of foreign exchange clearing. they can move toward self-sufficency. Reforms will 11. They are the Agricultural Development Bank, China reduce the size of the public capital infusion that Development Bank, and China Export and Import Bank, which will be needed which, given many other claims on together accounted for 11 percent of assets, 13 percent of loans, and less than 1 percent of deposits. They have mostly funded government finances, would help smooth post- their operations through long-term local currency bonds, WTO accession and structural adjustment else- mainly purchased by the state banks. where in the Chinese economy. 12. Of these, 92 percent were branches of foreign banks, and the remainder were locally registered banks with full foreign ownership or joint venture arrangements. In addition, 263 rep- resentative offices of foreign banks were situated in China. Notes 13. See, for example, the excellent collection of papers pre- 1. All dollar amounts are current U.S. dollars. sented at the Third Annual International Financial Markets and 2. This chapter focuses on banking, although it refers to Development Conference,entitled"Open Doors: Foreign Partici- other financial market segments where relevant. pation in Financial Systems in Developing Countries."Sponsored 3. Domestic banks compete vigorously in extending credit to by the World Bank, International Monetary Fund, and Brookings the larger corporate customers. One indicator of price competi- Institution, it was held April 19­21, 2001, in New York City. The tion is the tendency to make full, but one-sided, use of the lee- papers by Mathieson and Roldos (2001) and Pomerleano and way allowed by the PBC on loan rates (from 10 percent below to Vojta (2001) contain useful overviews of the issues. Also see 30 percent above the base rate). Because banks have little experi- Claessens, Demigurc-Krunt, and Huizinga (1998). ence with the estimation of risks, when they charge more than 14. See Boyreau-Debray (2002) and Hu and Zhou (2001) for the base rate it is usually for the full amount allowed, which is analyses of regional disparities in financial development. 30 percent, rather than for some amount in-between that would 15. In addition, the risks that could arise from a change in suggest finer differences among their customers. market conditions should be considered. For example, given the 4. See, for example, Lardy (2000). very generous debt-service-to-household-income criteria in the 5. These banks comprise 10 national commercial banks rapidly growing home mortgage market where state bank expo- (accounting for 12 percent of banking assets), 99 city commercial sure is rising, an increase in market interest rates could spark banks, 1,695 urban credit cooperatives (5 percent of banking defaults among households. assets), 39,255 rural credit cooperatives (10 percent of banking assets), and 3 policy banks (10 percent of banking assets). The other financial institutions were 12 financial leasing companies, References 100 trust and investment companies and regional policy banks, 98 securities firms, 13 insurance companies, 191 foreign-funded The word processed describes informally reproduced works that financial institutions, and 71 enterprise group finance companies. may not be commonly available through libraries. 6. China's WTO accession agreement for the financial services Almanac Editorial Board. 2002. Almanac of China's Finance and industry does not seem to have relied significantly on technical Banking. Beijing. calculations of profitability or solvency. Instead, broad product/ Bonin, John P., and Yiping Huang. 2001a. "Dealing with the Bad client/geographic criteria were used to give domestic institutions Loans of the Chinese Banks." Journal of Asian Economics and labor markets time to adjust. In this sense, the financial serv- (summer). ices sector is radically different from other sectors,such as agricul- ------. 2001b. "Foreign Entry into Chinese Banking: Does ture, where the information base was much stronger and permit- WTO Membership Threaten Domestic Banks?" Paper pre- ted disaggregated calculations of impacts,to which a host of finely sented at an international conference on Greater China and sequenced market opening measures were tied. the WTO, March 22­24, Hong Kong. 7. The literature on which to draw is vast. See, for example, Boyreau-Debray, Genevieve. 2002. "Financial Intermediation Lardy (1998), Cheng (1999), Bonin and Huang (2001a, 2001b), and Growth: Chinese Style." World Bank, Washington, D.C., Fang (2001), OECD (2002). March. Processed. 190 China and the WTO Cheng Hang-Sheng. 1999. "Can China Achieve Bank Marketiza- Li Kui-Wai, and Jun Ma. 2001. "China's Accession to the World tion without Bank Privatization? Paper presented at World Trade Organization and Policy Options for Banking Bank, March. Washington, D.C. Reform." Paper presented at Greater China and the WTO Claessens, Stijn, Asli Demigurc-Kunt, and Harry Huizinga. 1998. Conference, March. City University of Hong Kong, Hong "How Does Foreign Entry Affect the Domestic Banking Kong. Market?" Policy Research Working Paper 1918. World Bank, Mathieson, Donald J., and Jorge Roldos. 2001. "The Role of Washington, D.C., May. Foreign Banks in Emerging Markets." Paper presented at Fang Xinghai. 2001. "Reconstructing the Micro-Foundation of Third Annual International Financial Markets and Develop- China's Financial Sector." Paper presented at Conference on ment Conference. April 19­21. Sponsored by the World Financial Sector Reform in China, Harvard University, Bank, International Monetary Fund, and Brookings Institu- Cambridge, Mass., September. tion, New York. Hu Angang, and Li Zhou. 2001. "Market Openness and Good Organisation for Economic Co-operation and Development Governance: The Changing Regional Disparity of Financial (OECD). 2002. China in the World Economy: The Domestic Development in China (1978­1999)." Paper presented at Policy Challenges. OECD: Paris. Conference on Financial Sector Reform in China, Harvard People's Bank of China. People's Bank of China Quarterly University, Cambridge, Mass. Bulletin. 2002­03. Vol. 27. Lardy, Nicholas R. 1998. China's Unfinished Economic Pomerleano, Michael, and George J. Vojta. 2001. "What Do Revolution. Washington, D.C.: Brookings Institution and Foreign Banks Do in Emerging Markets? An Institutional Institute for International Economics. Study." Paper presented at Third Annual International ------. 2000. "When Will China's Financial System Meet Financial Markets and Development Conference, April China's Needs?" Revised version of paper presented at 19­21. Sponsored by the World Bank, International Mone- Conference on Policy Reform in China, February. Stanford tary Fund, and Brookings Institution, New York. University. Yearbook of China's Insurance. 2002. Beijing: Insurance Yearbook Leong, Elaine. 2000. "China's Postal Savings Service Challenges Editorial Board. State Banks." FinanceAsia (September). 12 WTO Accession and the Structure of China's Motor Vehicle Sector Joseph F. Francois and Dean Spinanger An automobile industry has often been a symbol of trade restrictions in Canada). The Chinese govern- economic prestige in the developing world. Brazil, ment has also set prices and limited competition China, India, Indonesia, and Malaysia, among other through a raft of import restrictions, which have countries, have all promoted, and sometimes even included quotas, high tariffs, and differential taxes showcased, the development of a domestic motor that favor local suppliers (see, for example, Li vehicle industry. In China, which has a huge popu- 2003). The limitation of trade has encouraged lation together with a surface area roughly as large inefficient production and has allowed market seg- as that of the United States and almost 15 percent mentation.1 larger than that of Brazil (see table 12.1), almost The integration of Greater China (Mainland every province has its own motor vehicle factory China and Chinese Tapei) into the World Trade and satellite factories. Yet, among the world's major Organization (WTO), and thus into most-favored- economies, China has the largest number of people nation (MFN) principles, has important implica- per vehicle. Even Indonesia, with a 30 percent lower tions for the Chinese economy, not least of all for per capita income than China, has over 50 percent the motor vehicle sector. For example, the accession fewer people per automobile. agreements specify major changes in tariffs, the China's situation is the result of a series of policy elimination of quotas as well as local content measures dating from 1949 (see box 12.1). Further requirements, and changes in the rules governing distorting an efficient structuring of the automo- foreign investment. The perception of the Chinese bile industry have been the internal measures that market by outside investors has already changed, have limited and even prohibited trade through because the application of WTO rules covering the local protectionism (analogous to inter-provincial treatment of foreign firms has reduced uncertainty This paper builds on Francois and Spinanger (forthcoming 2004). Special thanks are extended to Zhang Wenkui for help with data, to L. Alan Winters and Will Martin for detailed comments on an earlier draft, and to five anonymous referees, who offered valuable suggestions. Thanks also go to the participants in a World Bank- sponsored conference in Beijing for their helpful discussion. Finally, the authors would like to thank each other for their individual efforts to improve the quality of the output. Unfortunately, no agreement has been reached on who accepts final responsibility for errors or misinterpretations, and so we blame each other for any shortfalls. 191 192 China and the WTO TABLE 12.1 GNP, Population, and Stocks of Cars: Selected Countries, 2000 PPP GNP Stocks of per Capita Population Cars/Trucks (2002 U.S. (Millions, (Millions, People Surface Area Dollars) 2001) 2001) per Car (1,000 sq. km) India 2,570 1,032.4 6.3/5.9 163.2 3,287 Indonesia 2,990 209.0 3.0/2.4 68.8 1,905 China 4,390 1,271.8 8.5/15.4 149.0 9,598 Colombia 5,870 43.0 1.8/0.8 23.4 1,139 Turkey 6,120 66.2 4.5/1.6 14.6 775 Thailand 6,680 61.2 2.9/4.1 21.4 513 Brazil 7,250 172.4 15.8/4.0 10.9 8,547 Russia 7,820 144.8 21.2/5.1 6.8 17,075 Malaysia 8,280 23.8 4.2/1.0 5.6 330 Mexico 8,540 99.4 12.2/5.6 8.2 1,958 South Africa 9,870 43.2 41.0/2.5 1.1 1,221 Argentina 9,930 37.5 5.4/1.6 7.0 2,780 Rep. of Korea 16,480 47.3 8.9/4.0 5.3 99 Chinese Taipei 17,730 22.4 4.8/0.9 4.6 36 Spain 20,460 41.1 18.2/4.2 2.3 506 Italy 25,320 57.9 33.2/3.8 1.7 301 United Kingdom 25,870 58.8 27.8/3.4 2.1 243 Japan 26,070 127.0 53.5/19.9 2.4 378 France 26,180 59.2 28.7/5.9 2.1 552 Germany 26,220 82.3 44.4/3.6 1.9 357 Canada 28,070 31.1 17.1/0.7 1.8 9,971 United States 35,060 285.3 128.7/88.0 2.2 9,629 Total 9,224 4,017.1 492.2/184.3 8.2 71,200 Low and middle income 4,682 3,274.4 140.6/54.9 23.3 49,263 High income 29,248 742.7 351.6/129.4 2.1 21,937 PPP GNP = GNP adjusted for purchasing power parity exchange rate differences. Sources: World Bank, World Development Indicators, various issues; VDA, various issues. about the general economic climate, thereby induc- The industry itself anticipates significant change. ing notable increases in investment and prompting In recent years, growth in the motor vehicle sector new decisions to enter the market. has been very rapid, with output expanding at an This chapter assesses the effect of these broad annualized rate of 13 percent in the four years changes on the Chinese motor vehicle sector. It ending in 1999,at a 26 percent rate in the three years emphasizes the role of administratively imposed to 2002, and at more than 52 percent in 2003. In inefficiencies within the sector and the role of such view of the new, modern plants that came on line in inefficiencies in structural adjustment. We call 2001 and 2002 and the additional facilities expected these regulated efficiencies. A 1993 World Bank to increase capacity by over 150 percent from 2002 report pointed out that "it cannot be too much to 2005, a large, discrete change in production levels emphasized that in the long run the Chinese auto- is expected. motive sector will have a most efficient industrial At the same time, WTO membership implies organization when its development, dictated by lower prices and steeper foreign competition in international competition and state participation the sector. The response to this shift in the compet- in the sector, is eventually replaced by entrepre- itive landscape will be shaped by continuing prob- neurial ownership" (World Bank 1993, xix). lems with local government protection, lack of WTO Accession and the Structure of China's Motor Vehicle Sector 193 BOX 12.1 Summary of Developments in China's Automotive Sector The following summary is based on the authors' · From 1983 to 1985, the number of automo- survey of the literature: tive companies almost doubled, from 65 to 114 units. 1953­65: Self-Reliance Policy · By 1998, roughly 2,500 production units were in place. · About 60,000 vehicles were produced each · Provincial governments further regionalized year. production. · Production relied on Soviet technologies. · Major international firms began to invest, · Sector had no other international contacts. and then toward 1998 began to invest quite · Provincial governments set up production rapidly. units. · Volkswagen had already established opera- · By 1960, 16 auto producers and 28 assem- tions in China--in 1978. bly companies were in place. · These joint ventures accounted for about 60 percent of production during the period 1966­80: Security-Oriented Policy 1981­98. · Government invested heavily in western regions (Sichuan, Shanxi, and Hubei). 1999­ Opening Up and Beyond · Remote locations caused severe problems · Major investments have been initiated by and overcapacity. foreign companies. · Focus was on heavy vehicles for military · All major Japanese companies now have purposes. operations in China. · Car demand increased rapidly and capacities · All major German producers now have were expanded to 160,000 units per year. operations in China. · By 1980, 58 carmakers, 192 assembly · French and Italian producers are nominally companies, and 2,000 spare parts producers present. were operating in China. · U.S. producers also are nominally present. · Currently, the expansion is rapid; capacity is 1981­98: Initial Fruits of the Open now close to 2.5 million units. Door Policy · Capacity is especially growing in coastal areas. · The Open Door policy adopted in 1978 kick-started the industry. automobile infrastructure such as roads, parking, China's Auto Industry and service facilities, and related factors that act as Because of China's own national and regional poli- constraints on growth of the sector. Even so, the industry itself expects continued strong growth.2 cies, the country's motor vehicle industry is highly fragmented and inefficient by global standards. Notwithstanding industry expectations about This situation is not just the result of the introduc- its prospects, what is really likely to happen once tion of Soviet-style industrialization in the 1950s, the competitive landscape has changed in critical when firms were merely production units and ways? The primary approach employed here to questions about efficiency were irrelevant because explore this question is general equilibrium, involv- the dictated output was all that mattered. As noted ing the application of a computational model. by Zhang and Taylor (2001), the First Automobile The next section discusses some basic issues about Works (FAW) provides ample evidence of the the structure of China's motor vehicle industry, impact evolving from the various policies in place in particular the impact of government interven- in China over the past 50 years. Between 1959 and tion on its efficiency. A brief description of the 1981, FAW produced "a mere 1,542 units, an aver- model framework and then the experiments them- age 67 units per annum." As Zhang and Taylor also selves follow. Conclusions are offered in the final note, in 1970 the production cost of a particular section. 194 China and the WTO FIGURE 12.1 China's Production of Motor Vehicles since Adoption of the Open-Door Policy Thousands 3,500 3,000 Total motor vehicles 2,500 2,000 1,500 1,000 Trucks and buses 500 Passenger cars 0 1984 1986 1988 1990 1992 1994 1996 1998 2000 2002 Sources: Bessum 2002; VDA, various issues. model (the CA72) was 220,000 yuan, but "the sales Actually, a great deal of variation in plant scale price was only 40,000 yuan. . . . In the absence of exists around this average (see table 12.2). The lead- competition, all production units ran at low levels ing producer, Shanghai-Volkswagen, made 248,000 of productivity and efficiency. . . . The legacy left by sedans in 2002 (and therefore operated at MES lev- Mao's regime loomed well into the 1980s. By 1980 els if producing one model). Several other plants the number of automotive enterprises had risen to had production runs of less than 20,000. In this 2,379, consisting of 56 vehicle manufacturers . . . sense, there are strong parallels to the situation [producing, among other things] 5,418 cars." in Mexico prior to implementation of the NAFTA in China's fragmented and inefficient automotive which protected, inefficient factories operated well industry is also the result of import substitution within the global technology frontier (see Lopez de policies, together with cooperation agreements Silanes, Markusen, and Rutherford 1994). Overall, made with foreign companies beginning in the China's motor vehicle industry has about 2,400 1980s that were meant to fill the increasing gap industrial enterprises. In 1998, these included between production and the rapidly expanding 122 motor vehicle manufacturing plants, 520 auto- demand for automobiles. After all, the major thrust refitting factories, 130 motorcycle factories, 62 car of policies until then had been to build trucks.2 engine factories, and 1,589 auto and motorcycle These trends are depicted in figure 12.1. spare parts factories. Annual production capacity Thus China's automotive companies are operat- exceeds 2.3 million motor vehicles and 10 million ing with cost structures well within the global fron- motorcycles. Since 1995, the general pattern has tier, and the plants are producing considerably been closure of the smaller plants (generally rele- below global standards for efficient scale. This gated to the "other" category in table 12.2) and observation is illustrated in table 12.2 for the sedan expansion of production runs in the larger plants. industry, assuming that plants of minimum effi- With foreign investment and the rapid growth in cient scale (MES) for final assembly of cars are the industry, the number of plants producing at those that produce at least 200,000 units a year of least 25,000 vehicles rose from 3 in 1995 to 11 in one model (see Huang 2002, p. 543). China's entire 2002 (see U.S. and Chinese National Academies of sedan production in 1998 was 507,000 vehicles, Engineering 2003). produced by 13 factories. The fact that only two Import and domestic shipment data in value factories produced more than 100,000 sedans each terms are summarized in table 12.3 (the data are implies that fewer than 40,000 sedans were pro- for 1997, which serves as the "pre-accession" refer- duced by each of the other factories. ence point). Import tariff protection is summarized WTO Accession and the Structure of China's Motor Vehicle Sector 195 a 11 8 3 4,400 2,100 2,000 1,900 2002 93,000 78,000 68,000 64,000 60,000 47,000 38,000 32,000 20,000 16,000 16,000 13,500 -- 248,000 131,000 106,000 1,040,900 2001 8,000 7,800 6,210 5,686 8,000 4,663 1,253 -- -- -- 58,548 41,703 52,667 52,850 50,573 51,153 30,085 21,702 230,378 101,622 732,883 859 2000 2,767 3,159 3,059 2,423 5,380 1,000 4,867 -- -- 94,147 30,024 81,951 31,225 53,900 48,235 32,228 14,594 17,930 221,524 649,272 -- -- -- -- -- -- -- -- 1999 9,294 -- -- -- 81,464 15,731 40,200 44,583 10,008 31,312 230,946 101,828 565,366 -- 1998 2,246 -- -- -- -- -- -- -- 8,344 -- -- -- 8,013 66,000 15,000 36,240 36,239 235,000 100,021 507,103 -- -- -- -- -- -- -- -- -- 1997 1,000 -- 1,557 46,405 95,155 30,035 35,160 19,377 22,479 230,443 481,611 798 2002. -- 1996 9,228 -- -- -- -- -- -- -- -- -- 2,416 -- 44,825 88,232 16,420 26,051 200,222 388,192 Center opened. 3455589 2223347 1112212 1995­2002 yet -- 1995 3,797 -- -- -- -- -- -- -- -- 7,105 -- 6,698 24,553 65,258 19,350 17,770 25,127 22,570 not 352,298 China, were Documentation in or projections. Plant ehicleV closed (Subaru) by have Motor W160,070 (Daihatsu) 25,000> 50,000> 100,000> company Plant (Citroen) (Suzuki) Fengshen plants plants plants unqueY oyota of of of on W Xiali Group (Nainan-Mazda) Jeep)( plants Chinese Production Shanghai-V W-VAF Shanghai-GM ianjinT W-Audi-HongqiAF based Shenlong Chang'an Guangzhou-Honda Shanghai-Qirui Geely Dongfeng Haima uedo-KIAY Qinchuan Nanya Beijing Guizhou anjin-TiT producing producing producing Guangzhou-Peugeot Other otalT Number Number Number because are 2002; Auto values applicable. available Bessum 12.2 Not 2002 ces: Not Rank 2002/1995 1/1 2/4 3/n.a. 4/2 5/5 6/9 7/6 8/n.a. 9/n.a. 10/n.a. 11/n.a. 12/n.a. 13/n.a. 14/n.a. 15/n.a. 16/3 17/7 18/n.a. n.a./8 -- n.a. The a Sour ABLET 196 China and the WTO TABLE 12.3 Import and Domestic Shipment Values of China's Motor Vehicle Industry, 1997 (millions of US$) Value Imported motor vehicles and parts, world prices 3,607.71 Imported motor vehicles and parts, internal prices 4,849.31 Imported parts, internal prices 3,239.45 Imported motor vehicles, internal prices 1,609.86 Domestic intermediates and parts 32,812.46 Domestic intermediate parts 10,896.15 Industry consumption of motor vehicles 21,625.50 Final consumption of motor vehicles 290.81 Source: Global Trade Analysis Project (GTAP) version 5 database. TABLE 12.4 Tariffs on Chinese Motor Vehicles (percent) Base Rate Final Rate Finished motor vehicles 70.5 25 Motor vehicle parts 23.4 10 Electronic vehicle parts 12.0 10 Average, vehicles and parts 34.7 15 Sources: China's WTO accession schedule; GTAP data; Office of the U.S. Trade Representative. in table 12.4. China's pre-accession average tariff on WTO accession. In addition, quotas on automo- auto products (vehicles and parts and electronics) biles and on key components will be phased out by was 35 percent. The rate for vehicles averaged 70 2006, after being expanded by 15 percent a year percent, with sedans subject to tariffs of between 80 until then. Likewise, domestic content require- percent and 100 percent. Motor vehicle parts were ments have already been removed. (Both of these subject to an average tariff of only 23 percent. nontariff measures violate basic WTO rules.) Import shares were relatively low, averaging per- Clearly, these changes in the structure of protection haps 3 percent in the years 1995­2002. Officially, will have significant implications for the structure only 20,000 sedans were imported.3 Government of the automotive sector. Critically, other WTO policy encouraged the use of domestic parts, and, obligations imply free movement of imported better yet, locally produced parts. For new invest- autos (free of import quotas) within the China ments, domestic content rules applied, stipulating market. This situation implies tremendous pressure 80 percent domestic content by the third year. All for a breakdown of internal barriers for domestic this is reflected as well in the low share of automo- production and for a rationalization of the domes- tive parts imports in total production. As for the tic industry. The internal barriers to trade simply ownership of companies, even after full implemen- cannot be sustained if China's new WTO obliga- tation of the WTO, foreign ownership will be tions are to be taken seriously. limited to 50 percent.4 The government has itself recognized this situa- China's tariff rates are scheduled to come down tion. Official and industry sources have signaled substantially (table 12.4)--25 percent for vehicles their intention to support only a small number of and 10 percent for parts on an MFN basis as part of domestic production groups, perhaps including the WTO Accession and the Structure of China's Motor Vehicle Sector 197 Shanghai group (Volkswagen), China's First Auto national input-output, trade, and final demand Works (Volkswagen), Shanghai-GM (Buick), and structures. We modified the basic database. In par- the Dongfeng Group (Citroën). These groups with ticular, we updated the data set to better reflect their foreign partners already account for over actual import protection for goods and services. 70 percent of production in China. Such a sharp The basic data on current tariff rates are drawn rationalization would undoubtedly be painful, but from United Nations Conference on Trade and it could allow the industry to consolidate produc- Development (UNCTAD) and WTO data on the tion and work its way down the average cost curve schedules of applied and bound tariff rates, and for vehicle production. they are integrated into the core GTAP database. They are supplemented with data from the Office of the U.S. Trade Representative and the U.S. The Modeling Framework International Trade Commission on regional pref- The next section sketches a quantitative assessment erence schemes in the Western Hemisphere. For of the possible impact of WTO accession on China's agriculture, protection is based on Organisation motor vehicle industry--an assessment made by for Economic Co-operation and Development applying the computable general equilibrium (OECD) and U.S. Department of Agriculture (CGE) model. This section provides a brief (USDA) estimates of agricultural protection, as overview of the model. More technical details and integrated into the GTAP core database. Tariff and references are provided in Francois and Spinanger nontariff barrier estimates are further adjusted to (2001) and in the technical annex available for reflect remaining Uruguay Round commitments, downloading with the model files.5 For multi-sector including the phaseout of remaining textile and policy initiatives (such as WTO accession) in which clothing quotas under the ATC. Data on post­ interactions between sectors have a major effect on Uruguay Round tariffs are taken from recent esti- results, the use of CGE models has become a mates reported by Francois and Strutt (1999). relatively standard approach (see François 2000). These estimates are taken primarily from the We recognize that the results of these exercises are WTO's integrated database, with supplemental hampered both by the assumptions made and by information from the World Bank's recent assess- the quality of the data available. Even so, we believe ment of detailed pre­ and post­Uruguay Round their utility in estimating the possible overall pat- tariff schedules. All of this tariff information has tern of the effect of broad policy changes--both been matched to the current model sectors. direct and indirect--is well proven. Services trade barriers are based on the estimates described in the technical annex and are shown in table 12.6. (The basic GTAP database includes no The Model Data information at all on trade barriers for services, for The model data, taken from a variety of sources, example.) were organized into 23 sectors and 25 regions. Although the basic GTAP data set is bench- (Note that we included some detail on the value- marked to 1997 and reflects applied tariffs actu- added chain, linking fibers with textile and clothing ally in place in 1997, we want to work with a production to better capture the initial impact of representation of a post­Uruguay Round world. the Agreement on Textiles and Clothing [ATC] on To accomplish this, before conducting any policy our base scenario.) The sectors and regions for this experiments we run a "pre-experiment" in which 23 × 25 aggregation of the data are detailed in we implement the remaining Uruguay Round tar- table 12.5. iff cuts. Most of these cuts are already in place in Data on production and trade are based on the 1997 benchmark data set. At the same time, national accounting data linked through trade the data are adjusted to reflect regional preference flows and drawn directly from the Global Trade schemes in Latin America (not represented in the Analysis Project (GTAP) version 5 data set core GTAP database). The data set we work with (McDougall 2001). The GTAP version 5 data set is for actual experiments is therefore a representa- benchmarked to 1997, and it includes detailed tion of a notional world economy (values are in 198 China and the WTO TABLE 12.5 Regional and Sectoral Breakdown of the Model Model Regions/Economies Model Sectors Economy Description Sector Description Hong Kong Hong Kong (China) Wool Wool China (PRC) People's Republic of China Other natural fibers Natural fibers (cotton, etc.) Chinese Taipei Chinese Taipei Primary food Primary food production Japan Japan Other primary production Other primary production Korea, Rep. of Korea, Rep. of Sugar Sugar ASEAN5 ASEAN5 member statesa Processed foods Processed foods, tobacco, and beverages Vietnam Vietnam India India Textiles Textiles Bangladesh Bangladesh Clothing Wearing apparel Other South Asia Other South Asian economiesb Leather goods Leather products Australia Australia Chemicals, rubber, and Chemicals, refinery products, refinery products rubber, plastics New Zealand New Zealand Primary steel Steel refinery products Canada Canada Primary nonferrous metals Nonferrous metal products United States United States of America Motor vehicles and parts Motor vehicles and parts Mexico Mexico Electronics Electronic machinery and equipment Brazil Brazil Other machinery and Other machinery and equipment equipment MERCOSUR, other MERCOSURc Other manufactures Other manufactured goods CBI Caribbean Basin Initiative economiesd ATP Andean Trade Pact Wholesale and retail trade Wholesale and retail trade services economiesd Chile Chiled Transport services Transportation services (land, water, air) Other Latin America Other Latin Americad Communications Communications services European Union(15) European Union, Construction Construction 15 economies Turkey Turkey Finance, insurance, and Finance, insurance, and real real estate estate services Africa, Mideast Africa and the Middle East Commercial services Other commercial services Rest of world Rest of world Other services Other services (public, health, etc.) aASEAN5 Philippines, Thailand, Indonesia, Singapore, and Malaysia. bPakistan, Sri Lanka, and Nepal. cMERCOSUR Argentina, Paraguay, Uruguay. Brazil is represented separately. dNot treated in tables and figures. Source: The authors. 1997 U.S. dollars), with full implementation of Model Structure Uruguay Round tariff cuts. We then examine both We turn next to the basic analytical features of the the ATC phaseout and Greater China accession model. Except for automobiles, we use a very stan- with reference to this post­Uruguay Round tariff dard CGE model structure. On the production side, benchmark. WTO Accession and the Structure of China's Motor Vehicle Sector 199 TABLE 12.6 Mainland China's Pre- and Post-WTO Accession Tariff Rates (as modeled) Sector Model Base Rate Accession Rate New Bound Rate Merchandise Wool 14.76 42.00 38.00 Natural fibers (cotton, etc.) 3.14 17.38 13.58 Primary food production 58.80 58.13 46.83 Other primary production 0.48 6.94 5.08 Sugar 29.49 30.00 20.00 Processed food, tobacco, and beverages 37.65 40.66 23.18 Textiles 25.09 25.43 10.21 Wearing apparel 31.75 32.80 16.05 Leather products 12.10 20.94 17.02 Chemicals, refinery products, rubber, plastics 12.62 14.85 7.17 Steel refinery products 9.68 8.92 5.10 Nonferrous metal products 7.83 8.20 5.52 Motor vehicles and parts 34.42 38.65 15.41 Motor vehicles 70.50 70.50 25.00 Parts 23.40 23.40 10.00 Electronic machinery and equipment 11.93 16.90 9.62 Other machinery and equipment 12.83 15.37 10.14 Other manufactured goods 14.51 21.99 16.29 Services Wholesale and retail trade services 0.00 n.a. 0.00 Transportation services (land, water, air) 3.97 n.a. 1.99 Communications services 9.18 n.a. 4.59 Construction 13.68 n.a. 6.84 Finance, insurance, and real estate services 8.08 n.a. 4.04 Other commercial services 47.92 n.a. 23.96 Other services (public, health, etc.) 25.74 n.a. 12.87 n.a. Not applicable. Note: Services barriers are based on gravity equation estimates. Accession rates reflect an assumed 50 percent drop in cross-border trading cost estimates. Sources: China WTO accession schedule; GTAP data; Office of the U.S. Trade Representative. Gravity estimates are based on trade and macroeconomic data and cross-country regressions. See François and Spinanger 2001. all firms are assumed to minimize costs, employing in which all markets clear. Although we model domestic production factors (capital, labor, and changes in gross trade flows, we do not model land) and intermediate inputs from domestic and changes in net international capital flows. (This foreign sources to produce goods and services. approach does not by any means preclude intro- These technologies are modeled as constant elastic- ducing changes in the level of gross capital flows.) ity of substitution (CES) technologies defined over Trade liberalization in the goods sectors involves a primary inputs and Leontief processes defined reduction of tariffs. This reduction involves in turn over intermediate inputs. Products from different a shift from the model base rates in table 12.6 to regions are assumed to be imperfect substitutes in the new bound rates. The new bound rates are gen- accordance with the so-called Armington assump- erally quite close to our calculations of average tion. Prices on goods and factors adjust until all accession rates, also shown in table 12.6. Service markets are simultaneously in (general) equilib- sector liberalization is modeled as a reduction in rium, which means that we solve for equilibria trading costs, reflecting the barrier reductions 200 China and the WTO reported in table 12.6. These are modeled as cost savings, with expectations even higher--in the Samuelson iceberg costs. range of 25­30 percent cost reductions. The World For the motor vehicle sector, we want to model Bank (1993: 57) has described quite succinctly the the status quo in a stylized though representative expected gains from reaching MES: "If this cost- way. One option is to implement imperfect compe- volume relationship is applied to the Chinese auto- tition in the model, but such an approach does not motive industry, the passenger car segment has a really deal with the primary issue at hand. Govern- cost disadvantage of 20 to 30 percent compared with ment policy has led to market segmentation, as well the international producers having MES. This cost as to price setting and regulation. Although the best disadvantage could be an understatement, however, approach to use in this situation is ultimately some- as there are already eight producers in the market." thing of a judgment call, we have chosen to focus on These efficiencies are illustrated in figure 12.2, realized cost efficiency for the sector. The current where we take sedan assembly as representative. In cost structure of the motor vehicle industry reflects figure 12.2, individual plants are plotted relative to the net effect of a basket of policies. Like clothing in the cost:output mapping for an efficient plant. The India and automobiles in Mexico before the North figure depicts an expectation that significant cost American Free Trade Agreement (NAFTA) the savings may be realized if the government takes structure of the auto sector in China is a product of seriously its commitments to eliminate internal regulated efficiency--that is, the general regulatory barriers, and if it responds by encouraging a and administrative environment. The critical issue restructuring of the domestic industry around a is actually these collective inefficiencies, which fol- limited number of motor vehicle groups. low from the full set of industrial policies. At This net cost effect is stressed here and sets the the same time, an implication of the reforms that treatment of motor vehicles apart from that of are being implemented alongside WTO accession other sectors in the model. We work with the lower seems to be restructuring and consolidation, lead- bound of these cost effect estimates. In particular, ing to an improvement in regulated efficiency. we focus on potential cost savings in the final What shape will changes in regulated efficiency assembly of autos (stemming from consolidation gains take? The industry, through rationalization, and rationalization at the plant level), which yields may move collectively down the relevant cost a higher regulated efficiency level for the industry. curves. A comparison of current average plant scale In addition, the differential treatment of parts and (table 12.2) with a global norm closer to 350,000 finished vehicles in the tariff schedule is tracked. units per plant implies that average costs are roughly To set these issues in context, it may help to draw 20 percent higher simply because of inefficient on developments in the automobile industry at an scale. The 20 percent figure is based on the distri- earlier point in time. That large gains can be bution of current plants in table 12.2. In particular, achieved in rationalizing production and accord- if we apply the formula ln (Average Cost) = ingly reducing costs in the automobile sector was -CDR · ln(Quantity), where CDR, the cost- most clearly demonstrated toward the beginning of disadvantage ratio, is the inverse elasticity of scale, the twentieth century. Henry Ford changed the defined as parameters of producing automobiles radically by introducing the assembly line in 1910. By 1914, CDR = -Average Cost - Marginal Cost "13,000 workers at Ford were producing 260,720 Average Cost cars. By comparison, in the rest of the industry, it and where CDR is between 0.125 and 0.135 (the took 66,350 workers to make 286,770 [cars]."6 But range of values found in engineering studies), we it was not just the shift in cost parameters that was can calculate an average cost index for the industry. significant in transforming the sector. Demand fac- If such an index is 100 at 350,000 units per plant tors were also important. Cars in the United States, (the approximate norm for European and North as a result of Ford's new production methods, American plants), current plant structure yields a moved from being a scarce good to one affordable cost index of roughly 120. by large segments of the U.S. population. China is Data from interviews with industry representa- already in the process of moving into this phase, as tives (Feenstra and others 2001) point to similar is evident from figure 12.3. WTO Accession and the Structure of China's Motor Vehicle Sector 201 FIGURE 12.2 Scale and Average Cost (with a CDR of 0.125) Average cost index 2.4 2.2 Beijing Jeep 2.0 1.8 Tianjin Xiali 1.6 Daihatsu 1.4 FAW Shanghai Volkswagen Volkswagen 1.2 Geely Group 1.0 0 50,000 100,000 150,000 200,000 250,000 300,000 350,000 Production per plant Note: Moving from the current range of output to plants with an average of 300,000 units implies a 20 percent cost savings. Source: Author's calculations and data from table 12.2. FIGURE 12.3 Motor Vehicle Production in China and United States 4,500 3,500 4,000 3,000 3,500 2,500 3,000 2,500 2,000 2,000 1,500 1,500 1,000 1,000 500 500 0 0 1901 1905 1909 1913 1917 1921 1925 1984 1988 1992 1996 2000 United States 1901­1925 China 1984­2002 Source: Bessum 2002; VDA, various issues. Experiments and Results · Industry rationalization will implicitly involve the elimination of internal regional barriers and The experiments are directed at full accession to the will allow for consolidation and rationalization WTO for Greater China (Mainland China and within the domestic market. Small, inefficient Chinese Taipei). The basic accession package covers factories will close. the changes in tariffs detailed in table 12.3. For automobiles, we model the following effects:7 To quantify the last effect, we take sedan produc- · Tariffs on motor vehicles will decrease to tion as representative. Given the typical scale of 25 percent. domestic production, we estimate that auto plants · Tariffs on auto parts will decrease from an aver- will realize a 20 percent cost savings in assembly if age of 23.4 percent to an average of 10 percent. we move plants to efficient scale (see note 2 and the 202 China and the WTO TABLE 12.7 Impact of WTO Accession of Greater China on Output (percent change) A B C D A B E A C Impact of Impact of Impact of Eliminating Mainland Mainland ATC Quotas China and China and for WTO Chinese Chinese Members, Taipei Taipei Mainland Accession, Accession, Total Impact, Total Impact, China, and without with Auto without with Auto Chinese Auto Sector Sector Auto Sector Sector Taipei Restructuring Restructuring Restructuring Restructuring Wool 12.80 18.26 16.84 33.40 31.79 Other natural fibers 12.11 17.86 16.41 32.13 30.51 Primary food 0.43 1.03 0.92 1.46 1.34 Other primary production 2.60 3.57 3.33 6.07 5.84 Sugar 2.26 7.93 8.48 10.01 10.55 Processed foods 1.02 4.66 4.74 5.63 5.71 Textiles 13.93 32.00 30.57 50.39 48.75 Clothing 50.26 75.46 73.03 163.65 159.98 Leather goods 7.18 5.36 3.51 2.20 3.92 Chemicals, rubber, and 2.03 4.53 4.27 6.46 6.21 refinery products Primary steel 3.99 9.13 7.86 12.76 11.54 Primary nonferrous metals 5.42 9.24 8.94 14.16 ­13.87 Motor vehicles and parts 4.11 36.68 7.99 39.28 3.54 Electronics 5.06 3.91 4.43 8.77 9.26 Other machinery 3.80 5.39 4.84 8.98 8.46 and equipment Other manufactures 2.16 0.34 0.14 2.49 2.02 Wholesale and retail trade 0.25 1.39 1.93 1.14 1.68 Transport services 1.94 1.95 1.39 3.85 3.31 Communications 0.51 0.06 0.99 0.45 0.47 Construction 0.75 2.81 4.17 3.58 4.95 Finance, insurance, 0.65 0.40 0.22 1.05 0.44 and real estate Commercial services 0.78 5.85 5.41 6.58 6.15 Other services 0.00 0.46 1.23 0.46 1.23 Source: Model estimates. discussion in this overall section). This savings is experiment are drawn from other parts of the econ- modeled at the assembly level. omy, including the motor vehicle sector. The overall sectoral effects of the experiments The results in columns B and C of table 12.7 are are presented in table 12.7. This table reports very important for the motor vehicle sector. They changes in the quantity of output under our alter- reflect the incremental impact of China's market native scenarios. Thus, as expected, the extension of access commitments made as part of accession. the ATC phaseout to China and Chinese Taipei Column B is a "business as usual" scenario, without implies a rather dramatic expansion of the textile the restructuring discussed elsewhere in this chap- and clothing sectors--13.9 and 50.3 percent, ter. It reflects a domestic motor vehicle industry respectively. There are important general equilib- that continues to be fragmented, with favored pro- rium effects, because the resources needed for this ducers in each region, small production runs, and WTO Accession and the Structure of China's Motor Vehicle Sector 203 TABLE 12.8 Mainland China Motor Vehicle Market (millions of 1997 US$) Impact of Impact of Mainland China Mainland China and Chinese and Chinese Taipei Accession, Taipei Accession, without Auto with Auto 1997 Sector Sector Benchmark Restructuring Restructuring Values Imported motor vehicles and parts, world prices 3,607.71 10,595.68 6,967.97 Imported motor vehicles and parts, internal prices 4,806.39 12,080.71 7,995.72 Imported parts, internal prices 1,609.86 2,827.93 5,535.24 Imported motor vehicles, internal prices 3,196.53 9,252.78 2,460.48 Domestic autos, intermediates and parts 32,812.46 19,401.89 24,249.56 Domestic intermediate parts 10,896.15 4,493.95 5,189.12 Industry consumption of motor vehicles 21,625.50 14,698.79 18,785.03 Final consumption of motor vehicles 290.81 209.15 275.41 Indexes and shares Import share of total auto parts (percent of value) 12.87 38.62 51.61 Index of vehicle production 100.00 67.98 102.78 Index of parts production 100.00 41.22 56.28 Sources: Author's calculations and GTAP base data. FIGURE 12.4 Import Share of Total Auto Parts Percent of value 51.61 38.62 50 12.87 40 30 20 10 0 1997 benchmark Accession without Accession with restructuring restructuring high costs. Such an industry is simply unable to realized efficiency gain of roughly 20 percent as compete with imports, and indeed is very hard-hit scale economies are realized. This industry is much by imports, with domestic production falling by different from the one in column B. Production 36.7 percent. When that result is combined with the actually goes up slightly (3 percent) in total, and the initial impact of the ATC phaseout, a rather dra- industry emerges as a relatively competitive one, matic retrenchment of the uncompetitive domestic despite the loss of protection. industry occurs in the face of imports (column D). More information is provided on the differences The contrast is offered in column C and the cor- between the two scenarios in table 12.8 and fig- responding total in column E. Column C represents ure 12.4. The table expands on the information the elimination of internal barriers, rationalization originally provided in table 12.3, with a compara- of plants (with smaller plants being closed), and a ble breakdown corresponding to columns D and E 204 China and the WTO TABLE 12.9 Impact of Greater China Accession on Value Added (percent change) A B C D A B E A C Impact of Impact of Impact of Eliminating Mainland Mainland ATC Quotas China and China and for WTO Chinese Chinese Members, Taipei Taipei Total Total Mainland Accession, Accession, Impact, Impact, China, and without with Auto without with Auto Chinese Auto Sector Sector Auto Sector Sector Taipei Restructuring Restructuring Restructuring Restructuring Wool 11.55 12.97 10.62 26.02 23.40 Other natural fibers 10.87 12.58 10.21 24.82 22.20 Primary food -1.53 -5.46 -6.19 -6.91 -7.63 Other primary production -3.68 -7.88 -8.47 -11.27 -11.84 Sugar -3.34 -12.05 -13.36 -14.99 -16.25 Processed foods -2.11 -8.92 -9.81 -10.85 -11.72 Textiles 12.67 26.09 23.62 42.07 39.28 Clothing 48.60 67.61 63.81 149.06 143.42 Leather goods -8.20 0.64 -2.00 -7.61 -10.04 Chemicals, rubber, and ­3.11 -8.80 -9.37 -11.64 -12.19 refinery products Primary steel -5.06 -13.20 -12.76 -17.59 -17.17 Primary nonferrous metals -6.46 -13.31 -13.79 -18.91 -19.36 Motor vehicles and parts -5.17 -39.51 2.24 -42.64 -3.05 Electronics ­6.11 -8.21 -9.52 -13.82 -15.04 Other machinery and -4.86 -9.62 -9.91 -14.02 -14.29 equipment Other manufactures -3.24 -4.80 -5.19 -7.88 -8.26 Wholesale and retail trade -1.35 -3.15 -3.50 -4.46 -4.80 Transport services -3.03 -6.34 -6.64 -9.17 -9.47 Communications -1.61 -4.42 -4.39 -5.96 -5.93 Construction -0.36 -1.79 -1.38 -2.15 -1.73 Finance, insurance, and -1.75 -4.86 -5.12 -6.53 -6.78 real estate Commercial services -1.87 -10.07 -10.45 -11.75 -12.13 Other services -1.10 -4.04 -4.16 -5.10 -5.22 Source: Model estimates. in table 12.7; table 12.9 portrays the results just for vehicle production. Under the second scenario, the value added. The most striking difference between final assembly sector is rationalized, allowing the two scenarios is the different impacts on inter- the sector to compete more directly with imports. mediate parts production and final auto produc- The result is a shift to imported intermediates tion, as illustrated in figure 12.4. Under the first (rising to a market share of over 50 percent), a fall scenario, characterized by a domestic policy status in domestic parts production (as they are displaced quo, imports of parts rise slightly and their share of by imports), but a steady overall demand for parts. the domestic parts market rises substantially. At the For the industry overall, although ground is still same time, there is a dramatic surge in imports of lost to parts imports, sales of domestic vehicles motor vehicles, which displace more than one-third remain relatively steady in the face of imports. of existing domestic production. The overall mar- One last view of the effect of accession is ket for parts falls because of the decline in domestic presented in tables 12.10, 12.11, and 12.12, which WTO Accession and the Structure of China's Motor Vehicle Sector 205 TABLE 12.10 Impact of Greater China Accession on Value of Exports (percent change) A B C D A B E A C Impact of Impact of Impact of Eliminating Mainland Mainland ATC Quotas China and China and for WTO Chinese Chinese Members, Taipei Taipei Total Total Mainland Accession, Accession, Impact, Impact, China, and without with Auto without with Auto Chinese Auto Sector Sector Auto Sector Sector Taipei Restructuring Restructuring Restructuring Restructuring Wool ­9.52 -18.73 -20.14 -26.47 -27.75 Other natural fibers -4.67 -16.12 -16.67 -20.04 -20.56 Primary food -4.16 -5.05 -7.29 -9.00 -11.14 Other primary production -3.33 -2.98 -4.23 -6.21 -7.42 Sugar -3.29 -5.19 -7.05 -8.31 -10.11 Processed foods -4.99 -4.15 -5.96 -8.93 -10.65 Textiles 6.18 32.74 32.01 40.93 40.16 Clothing 87.81 80.03 77.48 238.11 233.33 Leather goods -7.66 3.85 1.76 -4.11 -6.04 Chemicals, rubber, and -3.29 -0.30 -1.01 -3.59 -4.27 refinery products Primary steel -5.56 -3.00 -3.99 -8.39 -9.33 Primary nonferrous metals -5.87 -4.05 -5.36 -9.69 -10.92 Motor vehicles and parts 12.33 16.17 392.33 1.85 331.64 Electronics -4.56 -0.86 -1.75 -5.39 -6.24 Other machinery and -6.69 -3.69 -4.52 -10.13 -10.91 equipment Other manufactures -5.74 -2.57 -3.69 -8.16 -9.22 Wholesale and retail trade -4.79 -5.19 -5.82 -9.73 -10.34 Transport services -3.31 -2.28 -2.44 -5.52 -5.67 Communications -5.05 -4.38 -4.90 -9.20 -9.70 Construction -4.92 -4.34 -5.62 -9.05 -10.26 Finance, insurance, and -5.33 -6.01 -7.10 -11.01 -12.05 real estate Commercial services -5.17 -3.00 -4.17 -8.02 -9.13 Other services -5.59 -5.98 -7.04 -11.24 -12.24 Source: Model estimates. reveal the export impacts observed under the vari- the Chinese industry. The export share for the ous modeling scenarios. It is logical to expect some Republic of Korea is 10 times as large. China's trade export response, both because of the general liber- is therefore well below global integration standards alization in trade and because pressure from as measured by exports. imports may force firms to seek other markets. Tables 12.10 and 12.11 compare export volume China exports less than 4 percent of its production effects, as a percentage of base exports and in dollar in the motor vehicle sector, based on 1997 values. terms. Clearly, restructuring accelerates the export Of US$32 billion in production,8 only $1.3 billion orientation of the industry, with a rapid growth in in products is exported. To put this figure in exports. Exports rise by about 300 percent, or perspective, Australia has a comparable level of $3.8 billion, reaching roughly 10 percent of pro- exports for an industry only one-third the size of duction by value. Although this expansion seems 206 China and the WTO TABLE 12.11 Impact of Greater China Accession on Value of Exports (millions of U.S. dollars) A B C D A B E A C Impact of Impact of Impact of Eliminating Mainland Mainland ATC Quotas China and China and for WTO Chinese Chinese Members, Taipei Taipei Total Total Mainland Accession, Accession, Impact, Impact, China, and without with Auto without with Auto Chinese Auto Sector Sector Auto Sector Sector Taipei Restructuring Restructuring Restructuring Restructuring Wool -4.95 -9.74 -10.47 -13.76 -14.42 Other natural fibers -0.18 -0.63 -0.65 -0.78 -0.80 Primary food -237.46 -288.39 -416.09 -513.86 -636.25 Other primary production -153.94 -138.02 -195.70 -287.36 -343.13 Sugar -4.58 -7.22 -9.82 -11.57 -14.08 Processed foods -360.07 -299.33 -430.38 -644.47 -768.99 Textiles 1,356.02 7,186.55 7,026.81 8,986.47 8,816.86 Clothing 4,3997.88 40,096.19 38,820.21 119,303.97 116,907.51 Leather goods -1,503.44 755.88 345.26 -805.49 -1,184.64 Chemicals, rubber, and -560.97 -51.79 -171.59 -611.06 -726.91 refinery products Primary steel -225.93 -121.72 -162.30 -340.88 -379.21 Primary nonferrous metals -130.28 -89.89 -118.97 -214.90 -242.26 Motor vehicles and parts 141.58 185.76 4,505.86 21.28 3,808.80 Electronics -1,365.27 -257.74 -524.98 -1,611.25 -1,866.29 Other machinery and -2,145.66 -1,183.94 -1,450.56 -3,250.42 -3,499.21 equipment Other manufactures -2,183.35 -977.14 -1,403.76 -3,104.39 -3,506.52 Wholesale and retail trade -338.09 -365.87 -410.81 -686.43 -729.21 Transport services -278.13 -191.84 -205.06 -463.62 -476.40 Communications -19.84 -17.22 -19.27 -36.20 -38.14 Construction -25.88 -22.84 -29.53 -47.59 -53.96 Finance, insurance, and -26.63 -30.02 -35.49 -55.05 -60.23 real estate Commercial services -62.46 -36.25 -50.40 -96.84 -110.26 Other services -78.30 -83.81 -98.61 -157.43 -171.40 Source: Model estimates. dramatic, it should be kept in perspective. Export WTO, and its impact on China's motor vehicle sec- shares are shown in table 12.12. Currently, auto- tor. The approach we take is general equilibrium, mobiles and parts are a small share of exports requiring application of a global general equilib- (0.6 percent in 1997), and they remain small (up to rium model. 2 percent) even with the growth in automobile We argue that regulatory reform and internal exports. In addition, most of the restructuring restructuring are critical to any realized effect on remains focused on the domestic market. the auto sector. Such restructuring is represented here by a cost reduction that follows from consoli- dation and rationalization. This representation is Summary and Conclusions supported by a comparison of scale in a typical In this chapter we examine the interaction between auto plant in China with that in typical plants in regulated efficiency and China's accession to the North America and Europe, and by firm survey WTO Accession and the Structure of China's Motor Vehicle Sector 207 TABLE 12.12 China Export Shares: Baseline and Scenario Total Impact, Total Impact, without Auto with Auto 1997 Sector Sector Export Shares Baseline Restructuring Restructuring Primary foods 0.046 0.033 0.033 Textiles 0.084 0.098 0.097 Clothing 0.102 0.303 0.298 Motor vehicles and parts 0.006 0.004 0.019 Electronics 0.133 0.100 0.099 Other machinery and equipment 0.146 0.104 0.103 Other manufactures 0.397 0.294 0.290 Services 0.087 0.062 0.062 Sources: Author's calculations; GTAP base data. responses. We also draw on earlier similar estimates date.9 To the extent that this potential can be of the benefits to be had by achieving minimum tapped, the pressure on automotive firms to be efficient scale and by radically structuring produc- more productive and thus more competitive will tion more efficiently. be all the greater. This, then, would be another The net result is movement of costs toward factor helping to ensure that the welfare gains cal- global norms. Without such restructuring, the culated will come about. domestic industry remains uncompetitive, and The shortcomings of the analysis also need to be WTO accession will lead to a surge in imports of highlighted. We have worked with a very stylized final vehicles, though imports of parts will fall as model, even though we feel it captures important production moves offshore. With restructuring, the elements of the real world. Although restructuring final assembly industry will become competitive by has positive overall implications for the motor world standards, while the parts industry further vehicle industry, significant adjustment costs not integrates with the global industry through exports pointed to explicitly in the model are bound to (and through higher imports of parts). As high- emerge. Even if value added is preserved within the lighted in figure 12.5, most automobile firms are sector, there will most likely be a dramatic reloca- located along the coast and thus are well placed to tion of jobs toward a limited number of plants, take advantage of global markets. with job losses in the other, smaller plants. The cur- Two additional issues also require attention. rent regional scattering of final auto production First, as seen in table12.1, China's ratio of popula- (figure 12.5) will be replaced by a more geographi- tion to motor vehicles is far higher than that found cally concentrated pattern. Parts production will in many other countries with similar income levels. also tend to concentrate. To the extent that parts Because this ratio reflects the impact of the existing suppliers are able to supply regional markets, this policies, significant changes in these policies will concentration is likely to mean an intensification of shift demand back to what could be viewed as the clustering in the coastal regions, with parts a normal pattern of consumption of cars, given shipments to Japan, Korea, the United States, and China's geographic attributes. Second, improved other regional centers of production.10 This result access to car financing could further strengthen the is broadly consistent with the findings and recom- demand for cars. Whereas about 75 percent of U.S. mendations of Takayasu and Mori (2004), who and European automobile purchases are financed stress that the future of the automobile industry in through loans, only 15 percent of car purchases the entire region, and not just China, hinges on the in China are financed this way (KPMG 2003, p. 7). ability of local governments to adopt policies that Although China's protocol of accession to the WTO help to better integrate national manufacturers into stipulates that automobile finance will be liberal- regional and global supply chains. Overall, a rela- ized, only draft legislation has been presented to tively large share of value added is kept intact with 208 China and the WTO FIGURE 12.5 Location of Automobile Production in China Location of foreign production Production capacities in provinces Producer Foreign producer Capacity Production, Capacity Production, (cars/year) 2002 (cars/year) 2002 1 SAIC VW VW 450,000 278,890 Anhui 60,000 49,397 2 SAIC GM GM 100,000 111,623 Bejing 115,000 10,408 3 FAW VW VW 270,000 158,654 Fujian 80,000 16,935 4 FAW Toyota Toyota/Mazda 70,000 30,165 Guandong 120,000 97,921 5 Dongfeng PSA PSA/Citroën 150,000 84,378 Guangxi Zhuang 150,000 n.a. 6 Dongfeng Honda Honda 60,000 59,024 Guizhou 10,000 1,831 7 Dongfeng Yulong Nissan/Yulong 60,000 38,897 Hainan 50,000 11,989 8 Tianjing Toyota Toyota 30,000 2,147 Heilongjang 30,000 14,577 9 JIangsu Nanya Fiat 100,000 23,393 Henan 30,000 n.a. 10 SAIC Chery Daewoo 60,000 49,397 Hubei 180,000 84,378 11 Zehjiang Jili Daewoo (geplant) 150,000 47,443 Jiangsu 130,000 38,460 12 Chongqing Chang'an Suzuki Suzuki/Yanjin 150,000 67,846 Jilin 340,000 188,819 13 Chang'an Ford Ford 50,000 n.a. Liaoming 230,000 3,751 14 Dengfeng Yueda Kia Kia 50,000 20,080 Shandong 80,000 n.a. 15 FAW Hainan Mazda 50,000 11,989 Shanghai 550,000 390,513 16 Beijing Hyundai Hyundai 30,000 1,356 Shanxi 50,000 20,080 17 China Guizhou Aviation Ind. Wanhong/Chenchang 10,000 1,831 Sichuan 205,000 67,846 18 Shenyang Brilliance Junbei BMW (mid-2003) 200,000 n.a. Tianjing 50,000 2,147 19 Harbin Hafei Mitsubishi 30,000 14,577 Zehjiang 150,000 47,443 20 Shangdong Yantei GM 50,000 n.a. Total 2,380,000 146,495 21 Southeast Zhonghua 60,000 16,935 22 Beijing Jeep DaimlerChrysler 85,000 9,052 Other foreign Employees No. of 23 Jinbei GM GM 30,000 3,751 firms plants 24 Hunan Changfeng Mitsubishi 30,000 15,067 25 Zhengzhou Nissan Nissan 30,000 n.a. Bosch 3,600 6 26 Rongcheng Huatai Hyundai 20,000 n.a. Kolbenschmidt 1,500 2 27 Jiangxi Fuqi Golden Lion 20,000 n.a. Michelin 4,000 2 28 Tianjing Huali Golden Lion 20,000 n.a. ZF/Sachs 2,100 2 29 SAIC GM Wuling GM 150,000 n.a. Total 11,200 12 30 Sanjiang Renault Renault 30,000 n.a. 31 Chengdu FAW Toyota 5,000 n.a. 32 Yizhong SAIC/RDS 10,000 n.a. n.a. Not applicable or available. Sources: VDA, various issues; Bessum 2002. WTO Accession and the Structure of China's Motor Vehicle Sector 209 restructuring. From an employment perspective, (October 6, 2003), China has opened up this sector in line with value added results closely track the impact of its WTO commitments. 10. European manufacturers have already established restructuring on employment. Table 12.9, there- 12 plants in China, and one large American company (Delphi) is fore, identifies a range of effects on auto sector shifting from Mexico. employment, from 43 percent (without restruc- 11. Again, this is a message not unique to China. As empha- sized by Doner, Noble, and Ravenhill (2004), the industries in turing) to 3 percent (with restructuring). Need- China, Malaysia, the Republic of Korea, Taiwan (China), and less to say, it is important that the structure of the Thailand all face significant pressure to restructure. In China's industry be rationalized.11 case, like that of the Republic of Korea, the relatively large local market offers the scope for increased presence of foreign pro- ducers through joint ventures and foreign direct investment. If Notes managed properly (that is, if realized in a context of integration of China's internal markets), net efficiency boosts as modeled 1.Smith and Venables (1991) dealt with the benefits arising should be realized, with results as modeled in this chapter. from such a situation, where access across the domestic market is improved, when examining the implications of the formation of the European Union's common internal market for domestic References and foreign suppliers. 2. See, for example, China Online (2001). As WTO member- The word processed describes informally reproduced works that ship approached in late 2001, the opinions of the industry and may not be commonly available through libraries. related ministries, as reflected in the Chinese press, hinged criti- Bessum, F. 2002. "Global Car Production Statistics." March. cally on whether restructuring of the domestic industry would Available at www.geocities.com/MotorCity/Speedway/4939/ be allowed to proceed. Thus a report in Touzi Yu Hezuo (summa- carprod.html. rized in China Online 2001) stressed the expected injury to the China Online. 2001. "How WTO Membership Could Affect industry, while the industry itself was indicating optimism that China's Auto Industry." Available at www.chinaonline.com/ it could realize significant cost reductions, allowing it to remain issues/wto/NewsArchive/secure/2000/january/b200010319- competitive with imports (Feenstra and others 2001). In the 3-SS.asp. meantime, price cuts by foreign producers in China have Chinese Motor Vehicle Documentation Center. 2002. Catalogue become commonplace, with some of them induced by increased of the Present Chinese Motor Car Production. 2d ed. import competition and others by more intense domestic com- Aldeboarn: Netherlands. petition. Buick, for example, reduced its prices on its domesti- Doner, R., G. Noble, and J. Ravenhill. 2004. "Production cally produced models by 12 percent, and VW lowered Passat Networks in East Asia's Automobile Industry." World Bank, prices by 6.5 percent (indiacar.net, May 3, 2002). But, even more Washington, D.C. Processed. important, nearly all major foreign producers have announced Feenstra, R., D. Sperling, L. Branstetter, E. Harwitt, and W. Hai. plans to sizably establish or increase production capacities. A 2001. "China's Entry to the WTO: A View from the Auto recent major manufacturer to do so was DaimlerChrysler; in Industry." Mimeo, University of California Davis. Processed. September 2003 it finally ratified plans to establish facilities to François, J. F. 2000. "Assessing the Results of General Equilib- produce Mercedes C and E models in China (International rium Studies of Multilateral Trade Negotiations." UNCTAD/ Herald Tribune, September 9, 2003). ITCD/TAB/4, UNCTAD Policy Issues in International Trade 3. According to unnamed sources in Hong Kong, 100,000 or and Commodities Study Series. UN Conference on Trade more sedans may have been imported into the country. Many of and Development, Geneva, October. the smuggled cars tend to be luxury models. François, J. F., and D. Spinanger. 2001. "Greater China's Acces- 4. In the past, foreign investors seeking partners frequently sion to the WTO: Implications for International Trade/ had to go to provinces other than those on the coast (see Production and for Hong Kong." Paper prepared for Hong overview of policies). These provinces often tried to ensure that Kong Trade Development Council, December. "buy local" conditions prevailed for local authorities. In ------. 2004. "Regulated Efficiency, WTO Accession, and the Shanghai, taxi regulations "by chance" stipulated specifications Motor Vehicle Sector in China." World Bank Economic that could be filled only by a Shanghai-VW model. Review; forthcoming. 5. The model files themselves, along with the technical François, J. F., and A. Strutt. 1999. "Post Uruguay Round annex describing the model, can be downloaded from www. Tariff Vectors for GTAP Version 4." Erasmus University, intereconomics.com/francois. The model is implemented in Rotterdam. GEMPACK. Huang, Y. 2002."Between Two Coordination Series: Automotive 6. See http://inventors.about.com/gi/dynamic/offsite.htm? Industrial Policy in China with a Comparison to Korea." site=http://www.wiley.com/products/subject/business/forbes/ Review of International Political Economy 9(3): 538­73. ford.html. KPMG. 2003. China Automotive and Component Parts Market. 7. We did not model the impact of the elimination of quotas Hong Kong, August. for imported cars. Data on the value of quota rents could not be Li, Shantong. 2003. "Survey Report on Regional Protection." In produced. Assuming that positive quota rents did exist the Workshop on National Market Integration, proceedings of welfare gains would be all the greater. However, the qualitative Development Research Centre Workshop, September 6. message would not change. Lopez de Silanes, F., J. Markusen, and T. F. Rutherford. 1994. 8. All dollar amounts are 1998 U.S. dollars. "Complementarity and Increasing Returns in Intermediate 9. Nonetheless, some major car companies (VW and Inputs." Journal of Development Economics 45: 133­51. Ford) have reached agreements with Chinese banks (KPMG, McDougall, R., ed. 2001. The GTAP Database--Version 5. Global 2003, p. 7). According to the International Herald Tribune Trade Analysis Center, Purdue University. 210 China and the WTO Smith, A., and A. Venables. 1991. "Economic Integration and World Bank. 1993. China Industrial Organization and Efficiency Market Access." European Economic Review 35: 388­95. Case Study: The Automotive Sector. Report no. 12134-CHA. Takayasu, K., and M. Mori. 2004. "The Global Strategies of Washington, D.C. Japanese Vehicle Assemblers and the Implications for the ------. Various issues. World Development Indicators. Thai automobile Industry." World Bank, Washington, D.C. Washington, D.C. Processed. Zhang, W., and R. Taylor. 2001. "EU Technology Transfer to U.S. and Chinese Academy of Engineering. 2003. Personal Cars China: The Automobile Industry as a Case Study." Journal of and China. National Academies Press. Washington, D.C. Asia Pacific Economy 6(2): 261­74. VDA (Verband der Automobilindustrie). Various issues. Tatsachen und Zahlen. Frankfurt. 13 Economic Impacts of China's Accession to the WTO Elena Ianchovichina and Will Martin Trade policy reforms such as those flowing from of basic information such as trade flows and patterns accession to the World Trade Organization (WTO) of production and consumption. What standard lead directly to changes in policy instruments such models such as GTAP (Hertel 1997; www.gtap.org) as tariffs, nontariff barriers, and coverage of trade do not do, however, is incorporate the nonstandard rules. The main policy concerns, however, are with features of China's partially reformed economy, the impacts on such economic variables as prices; where many imports enter duty-free if used in output, employment, and trade volumes; factor the production of exports and labor market poli- returns; and household incomes. In this chapter we cies result in serious barriers between urban and estimate the impacts on these key economic vari- rural areas. ables of China's accession to the WTO in 2001 as a Like Ianchovichina and Martin (2001) and guide to policy and as a basis for subsequent analy- Wang (2003), the analysis here explicitly allows for sis at the household level.1 the duty exemption arrangements that result in The obvious instrument for performing this close to half of China's imports entering as duty- type of analysis is the computable general equilib- free inputs into the production of exports. Recent rium (CGE) model. Many such models now exist, work by Sicular and Zhao (2004) is drawn on and a cottage industry has emerged in estimating to represent imperfect labor mobility and labor the impacts of trade reform in China (Gilbert and market distortions. This chapter extended our ear- Wahl 2001).2 The availability of the internationally lier work reported in Ianchovichina and Martin standard database of the Global Trade Analysis (2001) by moving to the GTAP version 5 database Project (GTAP) has facilitated such modeling and (Dimaranan and McDougall 2002) based on 1997 reduced the burden involved in obtaining estimates data rather than 1995 data; by incorporating We thank Kym Anderson, Hana Polackova Brixi, Louise Fox, Thomas Hertel, T. N. Srinivasan, Alan Winters, and three anonymous reviewers for helpful comments, Prashant Dave and Zhi Wang for their generosity in providing data. Elena Ianchovichina is Economist with the Economic Policy Unit of the Poverty Reduction and Economic Management Network at the World Bank; her email address is eianchovichina@worldbank. org. Will Martin is Lead Economist at the World Bank; his email is wmartin1@worldbank.org. 211 212 China and the WTO improved estimates of protection in agriculture allows export-oriented activities to use much more (Huang, Rozelle, and Min 2004)3 and services import-intensive means of production. It provides (Francois and Spinanger 2004); by using measures a reasonably realistic depiction of China's trade of liberalization based on the final, multilateral regime in the 1990s, when duty exemptions were agreement; by taking into account the restructuring used to facilitate exports while protection in the of the automobile sector (Francois and Spinanger rest of the economy remained fairly high.7 2004); and by simulating the consequences of The tax arrangements for export processing major labor market reform. This last issue is a par- (duty and value-added tax [VAT] exemptions on ticularly critical area for China, and there have been imported intermediate inputs and VAT refunds on few simulation studies.4 exports) discouraged export-oriented firms from This chapter first discusses the methodology using domestic intermediate materials and selling and then the policy changes associated with China's in the local market.8 Furthermore, the vast majority WTO accession and the results of the simulation of exports were produced using imported interme- analysis. It then considers some possible comple- diates that were either exempt from duties or eligi- mentary policy actions, such as reducing barriers ble for refunds on taxes paid.9 China Customs data to rural outmigration and expanding access to for 2000 show that 60 percent of imports entered education. duty free, 41 percentage points of which were imports used for export processing, 13 percentage points were capital goods, and 6 percentage points Methodology were goods in special categories, such as materials The standard GTAP model5 was adjusted to incor- used by research institutes.Rough calculations based porate China's important export processing on input-output information from the GTAP ver- arrangements. Ianchovichina (2003) documents sion 4 database (McDougall, Elbehri, and Truong the approach used and shows that failing to 1998) and data from China Customs suggest that account for China's duty exemptions in analyzing 26 percent of imports were used to produce for the WTO accession overstates the increase in China's domestic market and only about 3 percent were trade flows by 40 percent and the increase in used to produce ordinary exports.10 exports of selected sectors by 90 percent. The Intermediate inputs for domestic and export- adjusted model (GTAP-DD) also incorporates oriented activities were initially estimated by allo- some of China's key labor market mechanisms and cating them to each sector in proportion to sales in institutions that related research has shown may export and domestic markets. However, this yielded have a major influence on the impacts of WTO unsatisfactory results, with the database showing accession (Sicular and Zhao 2002, 2004). much less use of imported inputs in the export sec- tor than the reported imports of duty-free interme- diate inputs for export production obtained from Export Processing Arrangements China Customs (Li Yan, China Customs, personal Export processing arrangements in China take communication). To deal with this, increased use of many forms. Most arrangements allow firms pro- imported intermediates was allowed for in export ducing goods for export to import intermediate activities in accordance with the price changes inputs at world prices. These arrangements were involved in providing duty exemptions, and the incorporated in the GTAP-DD model used in this elasticities of substitution between domestic and study by creating two activities for each sector. For intermediate goods in the model.11 This modifica- sectors covered--or potentially covered--by export tion more than doubled the import-intensity of the processing arrangements, one activity is specialized exporting activities and reduced that of the domes- in production for export, and one is specialized in tically oriented activities. production for the domestic market.6 This separa- tion is preferable to representations based on a China's Labor Market Policies single sector producing differentiated products for domestic and export markets because it allows the Perhaps the central labor market issue for the two sectors to use different input mixes, and it analysis is the barriers to mobility between rural Economic Impacts of China's Accession to the WTO 213 and urban activities. Taking up employment in an using a constant elasticity of transformation urban area is inhibited by the need to obtain an between workers in agriculture and workers in urban residence permit (hukou). In addition, work- other sectors in the following simple manner: ers tend to be reluctant to permanently cut their LNF/LF = (WNF/WF), (13.1) ties with the rural sector because it is not possible to sell the land to which a rural family has use rights where LNF is the number of nonfarm workers, LF is (Hussain 2004). Many workers move temporarily the number of farm workers, is a constant term, from rural to urban areas, although restrictions are WNF and WF are nonfarm and farm wages, respec- frequently imposed on such movements, and social tively, and is the elasticity of transformation. The welfare benefits such as health care and schooling value of was set at 1.32, based on Sicular and for children enjoyed by urban residents are typi- Zhao's (2002) estimates of this parameter.12 The tax cally not available to temporary migrants. While reflecting pure policy-induced barriers between it is sometimes possible to overcome these prob- rural and urban wages was estimated at 34 percent lems by purchasing an urban residence permit, this by Shi (2002). While there is some level of unem- imposes an additional cost on migrants from rural ployment in China, particularly associated with to urban areas, a group with particularly limited state enterprise reform and rural­urban migration, access to capital. Finally, as in all countries, rural­ much of this unemployment is transitional, because urban labor mobility is also inhibited by the sector- of the weakness of the social safety net (Hussain specific nature of farmers' human capital and a 2004). Given the long-run focus of the analysis here, reluctance to cut family ties by migration to urban total employment is treated as exogenous. areas. The per capita income of agricultural workers is Trade Policy Reforms only about one-third that of urban workers (World Bank 2002). Not all of this difference can be attrib- This section examines the implications of trade uted to barriers to mobility between rural and policy reforms in China and in its trade partners in urban areas, however. Urban workers typically have the years leading up to accession. higher skills, work more intensively, and face higher costs of living than rural workers (Sicular and Zhao Changes in China's Trade Policies 2004). To capture the effects of barriers to mobility, During the 1990s China made substantial progress the model allows for both imperfect transforma- in reducing the coverage of nontariff barriers, low- tion between unskilled workers in agricultural ering tariffs, and abolishing the trade distortions and unskilled nonagricultural employment, and created by the exchange rate regime. Lardy (2002) an implicit tax on nonagricultural employment. estimates that the number of tariff lines subject to The imperfect transformation representation is de- quotas and licenses fell from 1,247 in 1992 to 261 in signed to reflect the substantial differences in the 1999. By 2001, 257 tariff lines were covered by characteristics of unskilled workers in rural and a combination of licenses and quotas and 47 by urban areas, and the ability, at a cost, to transform licenses only, while 245 were subject to designated agricultural workers into nonagricultural workers trading and 84 to state trading. Tendering and through training, experience, and the creation of other registration requirements, primarily for ma- nonagricultural jobs in rural areas. The tax is de- chinery and electrical products, covered an addi- signed to reflect the pure policy-induced barriers tional 120 tariff lines. By 2001, nontariff barriers between rural and urban workers, such as the covered 664 tariff lines, or less than 10 percent of requirement for a residence permit in urban areas tariff lines (see Appendix table 13.A.1), with over a and the barriers to mobility created by the inability third of these being subject to designated trading, to sell farm land. It is specified as a barrier that one of the less intrusive forms of quantitative raises the cost of labor to urban employers, with restriction employed in China. urban workers receiving the tax-inclusive wage. Data on nontariff barriers frequency alone may The imperfect transformation between agricul- be misleading because of the enormous varia- tural and nonagricultural workers is represented tions in the importance of tariff lines. To give some 214 China and the WTO TABLE 13.1 Import Coverage of Nontariff number of factors such as the progressive phase- Barriers (percent) out of nontariff barriers, a standstill on new non- tariff barriers during the accession process, the gen- Barrier 1996 2001 eral WTO prohibition against nontariff barriers Licenses and quotas 18.5 12.8 after accession, and a likely reduction in the severity Tendering 7.4 2.7 with which many of these measures have been Licensing only 2.2 0.5 administered. A simple rule of thumb that protec- State trading 11.0 9.5 tion provided by nontariff barriers declines with Designated trading 7.3 6.2 their import coverage would suggest that their Any nontariff barriers 32.5 21.6 protective impact has fallen to about 5 percent. No nontariff barriers 67.5 78.4 However, given the large margin of uncertainty Total 100 100 associated with this measure, the analysis here focuses only on tariff liberalization, implying that Note: For 1996 nontariff barrier coverage, the the results should be taken as a lower bound to the trade weights used were for 1992, while for 2001 the trade weights used were for 2000. overall impact of liberalization. Source: For 1996, Lardy 2002; for 2001, Mei The pace of tariff reform in China was also rapid Zhen of the Ministry of Foreign Trade and during the 1990s. While average tariffs were very Economic Cooperation during an internship at high in the early 1990s, they fell sharply after 1994 the World Bank, using data from WTO. (table 13.2). A significant tariff reform in October 1997 reduced average tariffs well below 20 percent. indication of the potential importance of nontariff Three subsequent reductions at the beginning of barriers, the import coverage of the key nontariff 1999, 2000, and 2001 further reduced tariffs on a barriers was calculated using data on nontariff bar- wide range of items. The progressive reductions in rier coverage of tariff lines and on import data by tariffs between 1992 and 2001 lowered average tar- tariff line. iffs by two-thirds, with larger than average cuts The import coverage of all nontariff barriers in in the manufacturing sector, thereby ensuring that China fell from 32.5 percent in 1996 (World Bank the future reductions in tariffs required under the 1997, p. 15) to 21.6 percent in 2001 (see table 13.1). WTO accession agreement are much smaller than Coverage of import licensing and quotas fell from the reductions occurring before accession. Another 18.5 percent in 1996 to 12.8 percent in 2001, and important feature of the reforms has been a sub- coverage of state trading from 11 to 9.5 percent. stantial reduction in the dispersion of tariff rates-- The import coverage of tendering requirements with the standard deviation falling from 32.1 per- fell particularly rapidly, from 7.4 percent in 1996 to cent in 1992 to 10 percent in 2001. 2.7 percent in 2001. For more details on coverage Examination of weighted average applied tariffs by type and commodity, see Appendix table 13.A.2. for 1995 and 2001 and after implementation of the Oil was by far the most important import sub- final tariff bindings agreed in the accession sched- ject to nontariff barriers, accounting for almost half ule suggests13 that substantial merchandise trade the value of imports subject to any nontariff barri- liberalization occurred in China over the period ers (see Appendix table 13.A.3). Ferrous metals, 1995­2001 (table 13.3). Weighted average tariffs subject to designated trading arrangements, were dropped substantially for wheat, beverages and the second most important category. Imports of oil tobacco, textiles, apparel, light manufactures, and oil products accounted for 84 percent of total petrochemicals, metals, automobiles, and electron- imports subject to state trading. ics. Analysis by Huang, Rozelle, and Min (2004) The average protective impact of the complete suggests that some agricultural commodities such set of nontariff barriers in China was estimated as vegetables and fruits, livestock and meat, and rice (very crudely) to be 9.3 percent in the mid-1990s faced negative protection in 1995, generally as a (World Bank 1997), with most of the protective result of restrictions on exports. Protection on effect arising from licensing and quota-constrained these commodities rose (or negative protection goods. The protective effect of these nontariff fell) over the period 1995­2001. Accession is not barriers has clearly declined since then because of a expected to lead to a significant fall in protection on Economic Impacts of China's Accession to the WTO 215 TABLE 13.2. China's Average Statutory Tariff Rates (percent) All Products Primary Products Manufactures Year Simple Weighted Simple Weighted Simple Weighted 1992 42.9 40.6 36.2 22.3 44.9 46.5 1993 39.9 38.4 33.3 20.9 41.8 44.0 1994 36.3 35.5 32.1 19.6 37.6 40.6 1996 23.6 22.6 25.4 20.0 23.1 23.2 1997 17.6 18.2 17.9 20.0 17.5 17.8 1998 17.5 18.7 17.9 20.0 17.4 18.5 1999 17.2 14.2 21.8 21.8 16.8 13.4 2000 17.0 14.1 22.4 19.5 16.6 13.3 2001 16.6 12.0 21.6 17.7 16.2 13.0 After accession 9.8 6.8 13.2 3.6 9.5 6.9 Source: World Bank 1999, p. 340 to 1998. Authors' calculations for tariff lines with imports in 1999. Trade data come from COMTRADE. Protection data for 1999 to 2001 come from CDS Consulting Co. and after accession from China's WTO final offer. most agricultural commodities after 2001. Import these measures are represented here as barriers to protection is expected to remain unchanged for trade in services expressed in ad valorem terms. most commodities except oilseeds, sugar, and dairy Following Francois and Spinanger (2004), the products. impact of accession is represented as halving the Protection will continue to fall for all other mer- nontariff barriers to services trade. Efficiency chandise commodities, with especially big cuts for improvements in the services sectors are not mod- processed food, beverages and tobacco, automo- eled since there were no reliable estimates of the biles, electronics, and other manufactures. Francois likely productivity gains at the time this research and Spinanger (2004) conclude that the automo- was conducted.16 bile sector liberalization will be accompanied by massive restructuring to realize economies of scale Changes in China's Trade Partners' Policies and improve structural efficiency, perhaps increas- ing productivity by 20 percent during the accession The arrangements for textiles and clothing are a period (2001­07).14 particularly important element of China's acces- A key element of China's accession agreement is sion. Unlike most other developing economy ex- the abolition of agricultural export subsidies. porters, China was excluded from the liberalizing Huang, Rozelle, and Min (2004) estimate that there elements of the Uruguay Round Agreement on Tex- was a 32-percent export subsidy on feedgrains and tiles and Clothing. This means that prior to acces- a 10 percent export subsidy on plant-based fibers in sion China did not benefit from the integration of 2001 (particularly cotton). textile and clothing products into the General In addition to China's barriers on merchandise Agreement on Tariffs and Trade (GATT) or from trade, border measures and domestic regulations the increases in quota growth rates provided for on domestic service sectors and trade in these serv- under this agreement. That placed upward pressure ices have reduced the efficiency of services sectors. on the prices of these quotas in China, raising the China has made substantial commitments to open costs of exporters just as an equivalent export tax its services sectors,15 and critical services such as would.17 Under its accession agreement China ben- telecommunications, logistics, and finance are efited immediately from the integration of textiles likely to benefit from inward foreign direct invest- and clothing into the GATT, leading to the aboli- ment and rising productivity as they are restruc- tion of quotas and increases in quota growth rates tured. Based on work by Francois and Spinanger that have occurred since 1994 (WTO 1994). All (2001) reported in Francois and Spinanger (2004), quotas are to be phased out by 2005. Importing 216 China and the WTO TABLE 13.3 China's Import Protection before and after WTO Accession (tariff or tariff equivalent, percent) China Taiwan, China Product 1995 2001 Post-Accessionb 1997 2001 Post-Accessionb Agriculture Rice -5.0 -3.3 -3.3 2.2 0.0 0.0 Wheat 25.0 12.0 12.0 6.5 6.5 6.5 Feedgrains 20.0 32.0 32.0 1.0 1.0 0.0 Vegetables and fruits -10.0 -4.0 -4.0 35.7 36.9 16.0 Oilseeds 30.0 20.0 3.0 1.8 0.8 0.2 Sugar 44.0 40.0 20.0 21.9 25.8 22.7 Plant-based fibers 20.0 17.0 20.0 0.0 0.0 0.0 Livestock and meat -20.0 -15.0 -15.0 7.5 6.5 4.0 Dairy 30.0 30.0 11.0 16.6 9.3 5.9 Processed food 20.1 26.2 9.9 14.9 14.2 9.9 Beverages and tobacco 137.2 43.2 15.6 48.1 22.0 13.0 Total 4.8 7.6 3.6 9.1 6.9 4.6 Manufacturing Extractive industries 3.4 1.0 0.6 5.5 5.5 4.1 Textiles 56.0 21.6 8.9 6.1 6.3 5.6 Apparel 76.1 23.7 14.9 12.8 13.4 11.2 Light manufactures 32.3 12.3 8.4 4.0 4.1 3.4 Petrochemicals 20.2 12.8 7.1 4.2 4.2 2.9 Metals 17.4 8.9 5.7 4.0 3.8 1.5 Automobiles 123.1 28.9 13.8 23.9 21.5 13.3 Electronics 24.4 10.3 2.3 2.9 0.5 0.3 Other manufactures 22.0 12.9 6.6 4.4 3.3 2.1 Total 25.3 13.5 6.9 6.3 5.2 3.5 Total merchandise tradea 24.3 13.3 6.8 6.5 5.2 3.6 Services Trade and transport 1.9 1.9 0.9 1.3 1.3 0.7 Construction 13.7 13.7 6.8 5.9 5.9 2.9 Communications 9.2 9.2 4.6 9.2 9.2 4.6 Commercial services 29.4 29.4 14.7 3.7 3.7 1.9 Other services 24.5 24.5 12.7 7.1 7.1 3.5 Total 10.3 10.3 5.2 3.2 3.2 1.6 a Those estimates are based on trade weights for the respective years. If trade weights for 2000 at the six- digit level of the harmonized system are used, the total weighted average tariffs in 2001 and 2007 are 12.2 percent in 2001 and 6.3 percent in 2007 for China and 4.5 percent and 3.1 percent for Taiwan, China. b Applied rates at the end of the implementation period were estimated as the lesser of the bindings and 2001 applied rates. In virtually all cases the bindings were lower than the applied rates. Source: Authors' calculations based on agricultural protection data from Huang, Rozelle, and Min 2002; manufacturing protection data from GTAP in 1995, from CDS Consulting Co. in 2001 and from China's WTO final offer for protection after accession; services protection data from Francois and Spinanger 2004. economies will be allowed to introduce special tex- China's trading partners to take safeguard actions tile safeguards during the period 2005­07, but these under rules that are more permissive of protection will be effective for only one year at a time. than the usual WTO safeguard rules (Messerlin The accession agreement includes a Transitional 2004). These provisions have the potential for Product-specific Safeguard mechanism that allows introducing a new form of protection against Economic Impacts of China's Accession to the WTO 217 China. This potential danger needs to be weighed baseline broadly replicates World Bank projections against the substantial gains to China from her for overall growth in each region and uses projec- trading partners being required to follow WTO tions of factor input growth and a residually deter- rules in implementing contingent protection meas- mined level of total factor productivity growth to ures against China. For simplicity, these gains and ensure consistency between the two. losses are assumed to cancel each other out. The model considers the effects of the WTO China's accession also triggered a liberalization accession agreement signed at the Doha ministerial of its partners' trade policies. The average tariffs of meeting in November 2001 on protection pre- Taiwan (China) were estimated to fall by almost vailing in 2001. In addition, it separately takes 1.5 percentage points, from 4.5 percent in 1997 to into account the liberalization in China during 3.1 percent after accession.Taiwan (China) commit- 1995­2001 since much of the liberalization during ted to tariff reductions on thousands of industrial the 1990s was influenced by China's desire to pre- and agricultural product lines, a phase-out of tariffs pare for the type of trade regime needed for WTO on a number of products as part of the Zero-for- accession and to establish the credibility of its com- Zero program of the Uruguay Round, and reduc- mitment to an open economy. These sharp reduc- tions in tariffs on chemical products as part of the tions in protection are unlikely to have occurred Chemical Harmonization program. Tariffs on the without the prospect of accession to WTO, and vast majority of products related to information they have been locked in by China's WTO commit- technology were also reduced in 2000 and, once ments. Thus results are presented for both the WTO accession commitments are implemented, accession period (2001­07) and the entire liberal- the tariff on electronic products will fall to 0.3 per- ization period (1995­2007). cent (see table 13.3). The analysis starts with 1995 tariff levels be- Taiwan (China) made horizontal and sector- cause 1995 was a major turning point in the specific commitments in the following service negotiations--the closing of the door on China's sectors: business, communication, construction, attempt to enter the world trading system by engineering, distribution, education, environmen- resuming its status as a Contracting Party to the tal, financial, health, social, transport services, and GATT. As Long (2000, p. 43) emphasized, China tourism and recreation.Like Francois and Spinanger focused more strongly on commercial considera- (2002), we treat the WTO accession commitments tions after 1995 than it had previously done--and of Taiwan (China) as halving nontariff barriers to its trading partners also strongly emphasized the trade in services. commercial aspects of the negotiations. To capture the implications of WTO accession, 1997 protec- tion data for China in the benchmark data (GTAP Experimental Design version 5) are adjusted to 1995 levels to obtain the The impact of accession is evaluated here in the initial base.18 For Taiwan (China) liberalization is dynamic context of the growth and structural considered to have begun in 1997, the year for change expected in China and its trading partners which tariff data are available in GTAP version 5. during the period up to 2007, when almost all of Two experiments are conducted to evaluate the the changes associated with accession will have impact of WTO accession. The first assesses the come into effect. A baseline scenario is constructed impact of the fall in tariffs from 1995 to 2001 levels under which the economies of the world grow and the restructuring of the automobile sector and experience the manifold structural changes accompanying the reductions in tariffs on automo- associated with economic growth up to 2007 biles and automobile parts during this period. The (table 13.4 and Appendix table 13.A.5). The GTAP second assesses the impact of the fall in tariffs model includes key elements such as changes in from 1995 to post-accession (2007) tariff levels, demand patterns as incomes rise, changes in the liberalization of the services sectors, continued industrial structure associated with changes in the restructuring of the automobile sector, removal stock of capital per worker, and changes in world of quotas on China's clothing and textiles exports, prices resulting from changes in world supply and and removal of China's agricultural export subsi- demand that allow the model to capture key dies. The productivity shock designed to capture changes in the world economy over this period. The the restructuring of the automobile sector is 218 China and the WTO TABLE 13.4 Projected Growth in Factor Inputs and Total Factor Productivity 1997­2007 (percent) Manufacturing Unskilled Skilled Total Factor Trading Partner Population Labor Labor Capital Productivitya North America 11 11 12 49 High (1.05) (1.08) (1.11) (4.07) Western Europe 0 -1 1 30 High (0.03) (-0.08) (0.07) (2.69) Australia and New Zealand 10 12 10 55 High (0.98) (1.14) (0.99) (4.45) Japan 1 -2 -7 35 Medium (0.06) (-0.19) (-0.71) (3.02) China 8 13 50 174 High (0.81) (1.26) (4.15) (10.62) Taiwan (China) 9 11 14 96 High (0.86) (1.05) (1.36) (6.97) Other newly industrialized 10 -1 55 88 Medium countries in Asia (0.93) (-0.10) (4.47) (6.53) Indonesia 16 17 123 25 Low (1.50) (1.59) (8.36) (2.27) Vietnam 15 32 36 111 Medium (1.40) (2.79) (3.10) (7.78) Other Southeast Asia 18 22 134 60 Low (1.70) (2.04) (8.87) (4.83) India 18 23 78 88 Medium (1.67) (2.10) (5.92) (6.54) Other South Asia 25 30 80 72 Medium (2.22) (2.69) (6.06) (5.55) Brazil 14 19 72 31 Medium (1.31) (1.77) (5.60) (2.75) Other Latin America 18 6 90 54 Low (1.68) (0.57) (6.65) (4.42) Turkey 16 19 107 55 Low (1.47) (1.75) (7.55) (4.46) Other Middle East and North Africa 24 37 67 28 Low (2.16) (3.23) (5.24) (2.50) Economies in transition -1 6 9 33 High (-0.11) (0.56) (0.90) (2.88) South African Customs Union 15 31 47 34 Low (1.39) (2.76) (3.92) (2.94) Other Sub-Saharan Africa 30 40 54 38 Medium (2.65) (3.42) (4.42) (3.26) Rest of world 18 23 35 68 Low (1.63) (2.10) (3.05) (5.32) aLow = average annual growth rates of 0.1 percent, medium = 1.0 percent, high = between 2 percent and 4 percent. Note: Numbers in parentheses are annual growth rates. Source: World Bank; GTAP data. Economic Impacts of China's Accession to the WTO 219 proportionate to the fall in tariffs on automobiles within agriculture. However, barriers to labor in each simulation. mobility between rural and urban employment are The same macroeconomic closure is used for all incorporated. experiments--full employment,19 perfect mobility of skilled and unskilled workers between nonagri- Assessment of China's Accession cultural sectors, and perfect mobility of unskilled workers within agriculture. Based on the working This section assesses the impacts of China's acces- assumption that trade balances are not changed sion to the WTO on China and its trading partners. significantly by WTO accession, trade balances as shares of GDP were fixed for China and Taiwan Impacts on China (China). While trade balances can be expected to vary during the transition, particularly if there are The focus here is on the impacts of the trade policy substantial changes in foreign investment levels, changes remaining after 2001. (Detailed results for foreign investment levels are not determined the period before 2001 are presented in Appendix within the model.20 table 13.A.6). One important development in the Taxes lost because of trade liberalization are period after 2001 is the removal of the quotas on assumed to be replaced by a uniform, nondistor- apparel and textiles imposed against China and tionary consumption tax affecting both private and other developing country exporters by major government final consumption of all goods and industrial country importers. These quotas are services. This hypothetical tax is included to ensure scheduled for abolition in January 2005 for all WTO that any adverse impacts of trade reform on gov- members. Abolition gives a significant boost to the ernment revenues, and hence on the ability to textile and apparel sectors in China, which had been provide income transfers or public services, are one of the countries most tightly restricted by allowed for in the analysis of impacts of the reform the quotas.21 Output in these sectors rises substan- on households. Because the GTAP version 5 data- tially (table 13.5). That in turn stimulates the pro- base appears to represent the VAT on domestic pro- duction of plant-based fibers (mainly cotton), duction as an output tax, the model generates tax which increases by 16 percent. losses from the contraction of industries such as Output and employment in the other agricultural tobacco and alcohol. These inward-oriented indus- sectors, with the exception of livestock and meat, are tries have higher VAT rates than export-oriented expected to fall as unskilled agricultural labor moves sectors such as apparel because exports are exempt into the textile and apparel sectors and unskilled from the VAT. When the export-oriented sectors nonfarm real wages rise (table 13.6). Oilseeds and expand, the net impact of WTO accession is a sharp sugar contract more than other agricultural sectors contraction in tax revenues. In reality, such a con- as a result of falling protection.Tariffs on oilseeds fall traction will not be observed because VATs of the from 20 percent to 3 percent, while tariffs on sugar same magnitude are levied on imports. To offset fall from 40 percent to 20 percent. Protection in this impact, particularly in the poverty analysis, the other agricultural sectors is assumed to remain consumption tax had to be adjusted downward. almost unchanged. The automobile and electronics This was done by first computing the consumption sectors also expand slightly, creating employment tax that compensates for the loss in output taxes opportunities,particularly for skilled labor.22 Results (this tax as a share of the total replacement tax is suggest that approximately 6 million farm workers equal to the share of the output tax loss in the total in China will leave their farm jobs as a result of WTO tax losses) and then adjusting the consumption tax accession reform after 2001 in pursuit of employ- rate to eliminate the component due to the change ment in the non-agricultural sectors.23 in output taxes. For most merchandise goods real wholesale prices To reflect the long-run change in the stance of fall as a result of trade liberalization after accession. trade policy--phased in over many years--involved Retail prices reflect a uniform consumption tax in WTO accession, most of the analysis uses a increase of about 1.9 percent levied to compensate standard long-run specification, with capital and for the loss of tariff revenue.24 For some products, labor freely mobile between industrial sectors and such as beverages and tobacco, automobiles, and 220 China and the WTO TABLE 13.5 Changes in China's Key Economic Indicators after 2001 as a Result of WTO Accession (percent unless otherwise indicated) Trade Balance (Millions of Wholesale Consumer Product Output Employment Exports Imports US$) Prices Prices Agriculture Rice -2.1 -2.3 6.1 -7.1 64 -0.9 0.9 Wheat -2.0 -2.3 18.9 -10.1 174 -1.7 0.4 Feedgrains -2.3 -2.6 -77.8 -2.4 -596 -1.9 1.9 Vegetables and fruits -3.4 -3.7 14.6 -6.3 214 -1.9 -0.1 Oilseeds -7.9 -8.4 29.8 20.9 -789 -2.8 -4.7 Sugar -6.5 -7.4 13.9 24.1 -73 -1.9 -3.1 Plant-based fibers 15.8 16.4 -51.8 7.7 -189 0.1 3.1 Livestock and meat 1.3 1.1 15.5 -8.9 837 -1.6 0.2 Dairy -2.0 -2.4 13.5 23.8 -143 -1.5 0.2 Other food -5.9 -6.4 11.4 62.6 -3,460 -1.7 -1.8 Beverages and tobacco -33.0 -33.1 9.7 112.4 -14,222 -1.8 -6.9 Manufacturing Extractive industries -1.0 -1.3 7.5 -4.4 2,088 -0.7 1.2 Textiles 15.6 15.5 32.7 38.5 -10,366 -1.7 -3.2 Apparel 57.3 56.1 105.8 30.9 49,690 -0.5 -1.9 Light manufacturing 3.7 3.7 5.9 6.8 1,786 -0.9 0.0 Petrochemicals -2.3 -2.3 3.1 11.8 -8,810 -0.7 0.8 Metals -2.1 -2.1 3.7 6.8 -1,893 -0.4 1.3 Automobiles 1.4 -2.2 27.7 24.0 516 -3.9 -4.2 Electronics 0.6 0.4 6.7 6.8 453 -1.3 -1.7 Other manufactures -2.1 -2.2 4.1 18.9 -11,291 -0.5 0.8 Services Trade and transport 0.0 0.0 0.8 -0.4 493 -0.2 1.6 Construction 0.9 0.9 2.7 17.5 -436 -0.2 1.7 Communications -0.5 -0.5 -0.5 10.9 -56 0.1 1.9 Commercial services -2.0 -2.0 -0.4 35.4 -1,749 0.2 1.9 Other services -1.7 -1.8 1.4 33.6 -1,525 -0.1 1.6 Total 1.0 0.0a 16.8 17.3 717 -0.7 -0.2 aReflects the fixed labor supply assumption. Source: Authors' simulations with modified GTAP model; see details in text. sugar, the fall in real retail prices reflects a larger than some farmers may be hurt by WTO accession. Non- proportionate drop in protection. farm wages rise 1.2 percent and skilled labor wages Increased demand for nonagricultural labor rise 0.8 percent, implying that workers in urban means higher real nonfarm wages and higher centers--and farmers who are able to engage in returns to nonagricultural labor relative to agricul- nonfarm employment--are more likely to be better tural labor. Removal of protection on some agricul- off as a result of WTO accession.25 tural sectors additionally lowers the attractiveness Accession will make China a much bigger player of farming and implies falling returns to farm labor in world markets through three channels--the and land. Real farm wages fall 0.7 percent and the rapid growth and structural change of its economy; real rental price of land falls 5.5 percent. The the liberalization undertaken in preparation for decline in farm incomes and the rise in the real WTO accession; and the liberalization under- retail price of many nonfarm products mean that taken after accession in 2001. The liberalization Economic Impacts of China's Accession to the WTO 221 TABLE 13.6 Change in Real Factor Prices in China as a Result of Accession, 2001­07 (percent) Accession with Accession with Accession with Labor Accession Labor Market Increase in Skill Market Reform and Item Alone Reform only Level only Increase in Skill Level Farm unskilled wages -0.7 16.8 1.6 19.4 Nonfarm unskilled wages 1.2 -3.8 2.7 -2.5 Skilled labor wages 0.8 -1.7 -6.3 -8.7 Rental price of land -5.5 -9.7 -6.4 -10.5 Rental price of capital 1.3 -1.4 0.9 -1.8 Price of capital goods -0.9 -3.6 -1.1 -3.9 Migration from rural to 6 28 10 32.0 urban jobs (millions) National welfare 10.0 11.0 10.0 11.0 (billions of 1997 US$) Source: Authors' simulations with modified GTAP model; see details in text. undertaken after 2001 contributes to an increase in Impacts on China's Trading Partners China's share in world exports from 4.4 percent to Among China's trading partners the largest 7.8 percent upon completion of accession. Simi- absolute gains accrue to North America and larly, China's share in world import markets rises Western Europe with close to half of the gains com- from 5.8 percent in 2001 to 6.4 percent in 2007. ing from elimination of the quotas they impose on With the removal of textile and apparel quotas, China's exports of textiles and clothing--and thus apparel exports lead export expansion with an elimination of their efficiency losses and rent trans- increase in export volume of about 106 percent, fers to China. North America, Western Europe, and followed by textiles and automobiles. The dramatic Japan also gain from China's cuts in protection, fall in protection of beverages and tobacco results which increase China's efficiency as an export in imports more than doubling, followed by supplier and its demand for their exports. increases in imports of food products, textiles, agri- The welfare gain by Taiwan (China) from its and cultural products, automobile parts, and commer- China's accession to the WTO is estimated at $3 bil- cial services. lion per year--the second largest gain relative to the China's total welfare gain from WTO accession is size of the economy after China's (see table 13.7). estimated at $40.6 billion (in 1997 dollars), or 2.2 percent of per capita real income (table 13.7).26 About half of the gain ($1.6 billion) was realized as a result of the liberalization in China and in Most of the gain ($31 billion) was realized follow- Taiwan (China) during 1997­2001. Remaining ing the massive liberalization between 1995 and reforms will lead to an estimated real income gain 2001 and the ongoing restructuring of the automo- of $1.4 billion a year after 2001. Other newly indus- bile industry. The remaining reforms will lead to an trialized economies also benefit from China's acces- additional welfare gain of $9.6 billion. The largest sion. Most of these benefits are associated with part of this gain in welfare will come from fur- trade liberalization and removal of quotas on tex- ther merchandise trade liberalization ($4.7 billion, tile and apparel, which translate into gains from nearly half the $9.6 billion), followed by $2.4 billion terms of trade improvements after 2001. (25 percent) from the removal of quotas on textiles The world as a whole and key developing eco- and apparel and $1.2 billion (12 percent) from serv- nomies that trade directly with China benefit from ices liberalization. Continuing automobile sector China's accession, but developing economies in restructuring will generate $1.1 billion (11 per- Southeast Asia, South Asia, and Latin America cent), while the removal of agricultural export sub- that compete with China in third markets may sidies will provide only $275 million (3 percent) in lose from the removal of textile and apparel quotas additional benefits. 222 China and the WTO TABLE 13.7 Welfare Change and Sources of Welfare Change as a Result of China's WTO Accession (millions of 1997 US$) Total, Export 1995­ Tariff Quota Subsidy Liberalization Auto Sector Impact Country or Group 2007 Cuts Reductions Reductions of Services Restructuring 2001­07 North America 6,072 3,207 2,713 24 172 -44 5,259 (0.0)* Western Europe 18,189 9,724 8,285 -51 338 -107 14,200 (0.2) Australia/New Zealand 136 175 -47 2 18 -12 152 (0.0) Japan 5,694 5,522 291 -22 5 -102 2,553 (0.1) Chinaa 40,552 29,452 2,389 275 1,160 7,276 9,563 (2.2) Taiwan (China) 2,985 2,300 338 -4 265 85 1,376 (0.6) Other Newly Industri- 6,831 6,539 -82 -185 49 511 1,456 alized Countries (0.7) Indonesia -408 -167 -216 -10 1 -16 -310 (-0.2) Vietnam -453 -63 -395 0 6 0 -405 (-1.4) Other Southeast Asia -585 -109 -464 -46 16 18 -268 (-0.1) India -3,357 -1,087 -2,338 -5 -23 96 -2,999 (-0.4) Other South Asia -1,622 -176 -1,427 -7 1 -12 -1,619 (-0.8) Brazil -76 -76 3 4 5 -12 359 (-0.0) Other Latin America -32 59 -171 20 32 29 -36 (-0.0) Turkey -338 -50 -295 -2 7 2 -327 (-0.1) Other Middle East 368 675 -467 -13 57 116 -365 and North Africa (0.0) Economies in 19 318 -321 4 15 3 -185 Transition (0.0) South African 78 89 -18 0 5 2 13 Customs Union (0.0) Other Sub-Saharan -45 71 -159 4 15 24 -78 Africa (-0.0) Rest of World 155 330 -210 -15 27 23 -78 (0.0) World 74,166 56,733 7,409 -27 2,171 7,880 28,261 aImpacts exclude output tax losses because of a compensating value-added tax levied uniformly on both imported and domestic goods. Numbers in parentheses are percentage changes in per capita utility. The impact for 1995­2001 is the difference between the impact for 1995­2007 and the impact for 2001­07. Source: Authors' simulations with modified GTAP model; see details in text. Economic Impacts of China's Accession to the WTO 223 after 2001. The losses will be largest for Vietnam-- farm jobs if the government removed the policy an economy that is following in China's foot- barriers to labor movement from rural to urban steps and has a similar pattern of comparative areas28--several times the estimated 6 million peo- advantage in labor-intensive products. The wel- ple who would move as a result of WTO accession fare loss for Vietnam is estimated as a 1.4 per- reforms alone between 2001 and 2007. The impact cent drop in per capita income (see table 13.7). on the composition of Chinese industrial output The loss to India is estimated to be considerably would also be substantial (table 13.8). This would smaller as a share of per capita income, at 0.4 per- allow not only apparel production to expand more cent, while the percentage losses to other countries but also metals, automobiles, electronics, machin- are very small. ery, other manufactures, and construction, all at the expense of reductions in some agricultural sectors. Complementary Policy Reforms Impact of an Increase in Skill Levels While the overall effects of WTO accession on China's economy are generally positive, there are One of the central problems facing most rural work- some concerns that declines in real returns to ers is low levels of education. One way to get a sense farm labor may exacerbate poverty in rural areas. of the likely impacts of improving access to educa- Approaches that deal directly with these prob- tion is to consider the impact of resultant increases lems are more likely to succeed than approaches in the skill levels of rural workers on the perform- that attempt to water down China's trade policy ance of the Chinese economy.This experiment looks reforms. Two policy tools that lend themselves to only at the impact of improvements in education on analysis within the model framework used here are the skills of rural workers. It ignores any potential relaxation of the barriers to labor migration from benefits to rural households from improvements rural to urban areas and skills upgrading for work- in access to education for their children--such as ers in rural areas. reductions in school fees--and any changes in the government budget associated with increases in government spending on education.29 Impact of Reducing the Policy Barriers An increase in the provision of education that to Labor Mobility would boost the annual growth rate for skilled Abolishing policy barriers to labor mobility from labor from 4.15 percent to 5 percent and would rural to urban areas--such as residence permits, lead to a decline in the annual growth rates for differences in social insurance, and the inability to unskilled labor from 1.26 percent to 1.1 percent sell agricultural land--in conjunction with acces- was considered. This was found to have important sion leads to a nearly 17 percent increase in real impacts on the structure of the Chinese economy. returns to rural workers (see table 13.6).27 This An increase in skilled labor leads to a stronger contrasts sharply with the 0.7 percent reduction in expansion, or a smaller contraction, in the manu- real farm wages for accession without labor market facturing sectors that are skilled labor­intensive reform. Rents to farmland would decline, with than does accession with labor market reform but higher farm wages leaving a smaller residual return no change in education spending (table 13.8, to farmland. Real urban unskilled wages would columns 2 and 4). Metals, automobiles, electronics, decline by an estimated 3.8 percent. Clearly, there and other manufactures all expand. would be scope for partial reform of these arrange- While output in some sectors expands, the real ments that could leave both farm and nonfarm wages of skilled workers fall as the supply of skilled unskilled workers better off than in the absence of workers increases (see table 13.6) and world prices labor market reform. of the outputs they produce decline. This contrasts These results suggest that this reform would with the case of accession alone, which results in an have significant impacts on the number of people increase in the real wages of skilled workers. How- leaving their farm jobs for jobs in the nonfarm sec- ever, the real wages of generally much poorer tors and on the industry composition of China's unskilled workers rise with increased education, economy. Some 28 million people would leave their with the wages of unskilled nonfarm workers rising 224 China and the WTO more than those of unskilled farm workers (see undertaking the greatest reform and will gain the table 13.6). Of course, those who are able to trans- most. The North American and Western European fer from agricultural to nonagricultural employ- economies that abolish their export quotas on tex- ment as a result of increased educational opportu- tiles and clothing and increase their direct trade nities are likely to be substantially better off. with China will gain the most in absolute terms. Overall, it is clear that increasing access to train- Taiwan (China) will benefit substantially, both as a ing and skills could substantially offset the adverse consequence of its own liberalization and through impacts on rural labor of the trade reforms associ- strengthened trade links with China. Japan will ated with accession. Finally, increased education gain substantially because of increased export boosts the need for migration as demand for opportunities in China and China's increased unskilled workers increases in large urban areas. competitiveness as a supplier. Other industrializ- An estimated 10 million farm workers are expected ing and industrialized economies that are China's to exchange farm jobs for nonfarm ones (see largest trading partners will also be substantial table 13.6). The impact on consumer prices is gainers. small--with falling prices for farm products and China's WTO accession will have a noticeable rising prices for manufactured commodities. impact on the level and pattern of global trade. With accession, China is becoming a much bigger player in world markets. Apparel exports will lead Impact of Labor Market Reform and an Increase China's export expansion, followed by textiles and in Skill Levels automobiles. In addition to being an important The combination of removing labor market barri- source of traded goods, China will become an ers and increasing education spending creates the important destination for other economies' prod- most favorable scenario for unskilled farm labor, ucts. Imports of beverages and tobacco will more leading to the largest increase in real farm wages than double, followed by imports of food prod- (19.4 percent; see table 13.6). Farm output con- ucts, textiles, agricultural products, automobile tracts more than in the case of labor market parts, and commercial services. The expansion of reform alone, while skilled labor­intensive indus- textiles, light manufactures, petrochemicals, and tries such as metals, automobiles, electronics, other equipment exports from Taiwan (China) will be manufactures, and services expand more than in driven almost entirely by demand for these prod- the case of labor market reform alone or increased ucts in China. education spending alone (see table 13.8). Under Accession will have important distributional this scenario an estimated 32 million farm workers consequences for China. The wages of skilled work- would leave their farm jobs for jobs in urban areas ers and unskilled nonfarm workers will rise in real (table 13.6). terms and relative to the wages of farm workers. An These results suggest that to generate pro-poor estimated 6 million people will leave their farm jobs growth over the next decade, the government in pursuit of employment in industry and services. should consider both removing policy barriers to Real farm wages and land rental rates will decline. labor movement and changing the composition of The decline in farm incomes and the rise in the real spending to favor education. Not only would these retail prices of many nonfarm products suggest that policies facilitate the transformation of China's some farmers may be hurt by WTO accession after economy toward services and high-tech manufac- 2001--an issue explored by Chen and Ravallion in turing sectors, but they also have the potential to chapter 15. more than offset any negative impacts of accession To help offset these adverse impacts on farmers, on rural wages and incomes. the Chinese government might make changes in its labor market policies. Abolition of the hukou sys- tem and reform of the labor market more generally Conclusion would raise farm wages and allow 28 million people The analysis suggests that the reforming economies to migrate to nonfarm jobs in search of a better life. and their close trading partners will be the biggest It would lead to an even bigger expansion of the beneficiaries of accession to the WTO. China is labor-intensive manufacturing sector. An increase Economic Impacts of China's Accession to the WTO 225 TABLE 13.8 Change in Output and Employment in China as a Result of WTO Accession and Other Reforms (percent change over the period 2001­07) Output Employment With labor With Labor Market Without With With Market Without With Labor With Reform and Labor Labor Increase Reform and Hukou Market Increase in Increase in Market Market in skill Increase in Product Removal Reform Skill Level Skill Level Reform Reform Level Skill Level Agriculture Rice -2.1 -4.3 -2.4 -4.6 -2.3 -7.4 -3.1 -8.2 Wheat -2.0 -11.5 -3.3 -12.9 -2.3 -13.3 -3.9 -14.9 Feedgrains -2.3 -7.8 -3.1 -8.6 -2.6 -9.7 -3.7 -10.6 Vegetables and fruits -3.4 -7.1 -3.9 -7.7 -3.7 -8.9 -4.6 -9.7 Oilseeds -7.9 -18.4 -9.4 -19.8 -8.4 -20.4 -10.2 -22.0 Sugar -6.5 -17.1 -8.0 -18.4 -7.4 -22.4 -9.6 -24.2 Plant-based fibers 15.8 12.8 15.1 12.1 16.4 11.6 15.5 10.6 Livestock and meat 1.3 -3.3 0.6 -4.0 1.1 -7.0 -0.3 -8.2 Dairy -2.0 -9.4 -3.1 -10.5 -2.4 -14.4 -4.3 -16.0 Other food -5.9 -13.4 -7.0 -14.5 -6.4 -13.2 -8.9 -15.5 Beverages and tobacco -33.0 -38.7 -33.7 -39.5 -33.1 -37.6 -35.0 -39.5 Maufacturing Extractive industries -1.0 0.1 -1.2 -0.1 -1.3 0.2 -1.7 -0.2 Textiles 15.6 14.7 15.3 14.3 15.5 16.8 12.7 14.0 Apparel 57.3 61.4 56.7 60.7 56.1 62.6 52.7 59.1 Light manufacturing 3.7 -6.8 2.1 -8.5 3.7 -5.4 0.1 -8.9 Petrochemicals -2.3 -1.3 -2.3 -1.2 -2.3 0.7 -4.4 -1.4 Metals -2.1 0.8 -1.8 1.2 -2.1 2.4 -3.9 0.7 Automobiles 1.4 4.1 1.8 4.4 -2.2 2.3 -4.0 0.5 Electronics 0.6 4.5 1.1 5.1 0.4 6.3 -1.3 4.6 Other manufactures -2.1 0.3 -1.9 0.6 -2.2 2.2 -4.0 0.3 Services Trade and transport 0.0 0.8 0.1 1.0 0.0 3.4 -3.1 0.4 Construction 0.9 2.0 0.9 1.9 0.9 3.4 -1.4 1.0 Communications -0.5 0.6 -0.3 0.9 -0.5 3.4 -3.0 0.8 Commercial services -2.0 -1.4 -1.8 -1.2 -2.0 1.0 -4.7 -1.8 Other services -1.7 -0.5 -0.9 0.3 -1.8 1.5 -6.2 -2.9 Source: Authors' simulations with modified GTAP model. in education spending, to improve skills, would much of the variation in tariffs and the welfare gains have a positive impact on the structure of the from reducing this variation within the product Chinese economy. The real wages of skilled work- aggregates used in the analyses (Bach and Martin ers would fall, while the real wages of unskilled 2001; Martin, van der Mensbrugghe, and Manole workers would rise with increased education 2003). When Bach, Martin, and Stevens (1996) spending. Thus, on the income side, improved skills adjusted for this in a partial equilibrium context, would induce pro-poor growth and decrease gains to China almost doubled. The analysis here poverty and inequality. assumes flexible wages and full employment. How- A number of caveats are important. The gains to ever, trade liberalization and foreign competition China and Taiwan (China) are probably understated may worsen unemployment and put downward because tariff aggregation in the GTAP model hides pressure on the wages of unskilled workers in the 226 China and the WTO short run. Furthermore, while the analysis here liberalization improves returns to investment and improves on Ianchovichina and Martin (2001), the liberalization of rules on investment eases with better treatment of the extent of liberalization financial flows into previously restricted sectors in agriculture and services and the changes in the such as services and automobile production. The automobile sector, there are still areas that have substantial productivity gaps between local and been ignored. One is nontariff barriers in the foreign firms imply that new FDI will raise produc- manufacturing sectors other than the quotas on tivity.30 Walmsley, Hertel, and Ianchovichina apparel and textiles. Another is the impact of acces- (2004) take into account both the impact of FDI sion on foreign direct investment (FDI) and the and increased productivity growth in services. hard-to-measure efficiency gains in services that are They find that the impacts of accession are far associated with this increased investment. larger than those predicted by earlier studies, FDI has contributed significantly to China's eco- including this one, which ignore the potential nomic growth and will play an important role as productivity gains in the services sector and are China continues to reform its economy. WTO abstracted from capital accumulation and foreign accession is likely to increase FDI in China, as trade investment. Economic Impacts of China's Accession to the WTO 227 TABLE 13.A.1 Tariff Lines Subject to Import NTBs, China, 2001 Licenses & Licensing State Designated Any Total Quotas Tendering only Trading Trading Unrestricted NTB Tarifflines Paddy rice 0 0 3 3 0 0 3 3 Wheat 0 0 3 3 0 0 3 3 Cereal grains nec 0 0 1 2 0 9 2 11 Vegetables, fruit, nuts 0 0 0 0 0 109 0 109 Oil seeds 0 0 0 0 0 25 0 25 Sugar cane, sugar beet 0 0 0 0 0 1 0 1 Plant-based fibers 1 0 0 1 0 6 1 7 Crops nec 5 0 0 5 0 96 5 101 Cattle, sheep, goat, horses 0 0 0 0 0 6 0 6 Animal products, nec 0 0 0 0 0 62 0 62 Wool, silk-worm cocoons 3 0 0 1 2 9 3 12 Forestry 0 0 0 0 12 23 12 35 Fishing 0 0 0 0 0 57 0 57 Coal 0 0 0 0 0 6 0 6 Oil 0 0 2 1 0 1 3 4 Gas 1 0 0 1 0 1 1 2 Minerals nec 0 0 0 0 0 106 0 106 Meat: cattle, sheep, goats 0 0 0 0 0 26 0 26 Meat products nec 0 0 0 0 0 47 0 47 Vegetable oils and fats 0 0 12 7 0 32 12 44 Dairy products 0 0 0 0 0 24 0 24 Processed rice 0 0 2 2 0 0 2 2 Sugar 7 0 0 9 0 3 9 12 Food products nec 0 0 0 8 0 311 8 319 Beverages and tobacco 1 0 9 5 1 17 15 32 Textiles 39 0 0 3 25 711 46 757 Wearing apparel 0 0 0 0 0 289 0 289 Leather products 11 16 0 0 0 73 27 100 Wood products 0 0 0 0 18 106 18 124 Paper products, publishing 0 0 0 3 0 160 3 163 Petroleum, coal products 8 0 0 7 0 25 8 33 Chemical, rubber, plastic 35 0 15 21 5 1248 51 1299 Mineral products nec 0 0 0 0 0 198 0 198 Ferrous metals 0 0 0 0 181 49 181 230 Metals nec 0 0 0 0 0 190 0 190 Metal products 0 0 0 0 1 264 1 265 Motor vehicles and parts 64 0 0 0 0 93 64 157 Transport equipment nec 7 10 0 0 0 72 17 89 Electronic equipment 36 17 0 0 0 205 53 258 Machinery and equipment 39 77 0 2 0 1199 116 1315 Manufactures nec 0 0 0 0 0 219 0 219 Electricity 0 0 0 0 0 1 0 1 Gas manuf, distribution 0 0 0 0 0 1 0 1 Total 257 120 47 84 245 6080 664 6744 Source: WTO 2001. Commodity definitions are for GTAP-5, see www.gtap.org for concordances. 228 China and the WTO TABLE 13.A.2 The Import Coverage of Nontariff Barriers in China, 2001 Licenses & Licensing State Designated Any Unrestricted Quotas Tendering only Trading Trading NTB Trading Total % % % % % % % % Paddy rice 100 0 100 100 0 100 0 100 Wheat 100 0 100 100 0 100 0 100 Cereal grains nec 0 0 0 0 0 0 100 100 Vegetables, fruit, nuts 0 0 0 0 0 0 100 100 Oil seeds 0 0 0 0 0 0 100 100 Sugar cane, sugar beet 0 0 0 0 0 0 100 100 Plant-based fibers 93 0 0 93 0 93 7 100 Crops nec 48 0 0 48 0 48 52 100 Cattle, sheep and 0 0 0 0 0 0 100 100 goats, horses Animal products, nec 0 0 0 0 0 0 100 100 Wool, silk-worm cocoons 0 0 0 0 95 95 5 100 Forestry 0 0 0 0 94 94 6 100 Fishing 0 0 0 0 0 0 100 100 Coal 0 0 0 0 0 0 100 100 Oil 100 0 0 100 0 100 0 100 Gas 0 0 0 0 0 0 100 100 Minerals nec 0 0 0 0 0 0 100 100 Meat: cattle, sheep, 0 0 0 0 0 0 100 100 goats, horse Meat products nec 0 0 0 0 0 0 100 100 Vegetable oils and fats 59 0 60 59 0 60 40 100 Dairy Products 0 0 0 0 0 0 100 100 Processed rice 100 0 100 100 0 100 0 100 Sugar 85 0 0 85 0 85 15 100 Food products nec 1 0 0 1 0 1 99 100 Beverages and tobacco 20 0 16 20 0 36 64 100 products Textiles 9 0 0 0 8 14 86 100 Wearing apparel 0 0 0 0 0 0 100 100 Leather products 0 0 0 0 0 0 100 100 Wood products 0 0 0 0 55 55 45 100 Paper products, publishing 0 0 0 0 0 0 100 100 Petroleum, coal products 58 0 0 58 0 58 42 100 Chemical, rubber, 5 0 1 5 2 7 93 100 plastic prods Mineral products nec 0 0 0 0 0 0 100 100 Ferrous metals 0 0 0 0 85 85 16 100 Metals nec 0 0 0 0 0 0 100 100 Metal products 0 0 0 0 1 1 99 100 Motor vehicles and parts 32 0 0 0 0 32 68 100 Transport equipment nec 1 3 0 0 0 4 96 100 Electronic equipment 9 5 0 0 0 14 86 100 Machinery and 1 8 0 0 0 10 90 100 equipment nec Manufactures nec 0 0 0 0 0 0 100 100 Total Import Coverage 12.8 2.7 0.5 9.5 6.2 21.6 78.4 100 Note: Based on WTO 2001 and import data from China Customs for 2000. Economic Impacts of China's Accession to the WTO 229 TABLE 13.A.3 Commodity Import Shares by NTB Measure Licenses Licensing State Designated Any Unrestricted & Quotas Tendering Only Trading Trading NTB Trading % % % % % % % Total Paddy rice 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 Wheat 0.0 0.0 13.0 0.7 0.0 0.6 0.0 0.1 Cereal grains nec 0.0 0.0 0.0 0.0 0.0 0.0 0.2 0.2 Vegetables, fruit, nuts 0.0 0.0 0.0 0.0 0.0 0.0 0.2 0.2 Oil seeds 0.0 0.0 0.0 0.0 0.0 0.0 1.7 1.5 Sugar cane, sugar beet 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 Plant-based fibers 0.5 0.0 0.0 0.3 0.0 0.0 0.0 0.0 Crops nec 1.1 0.0 0.0 0.8 0.0 0.0 0.1 0.1 Cattle, sheep and goats 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 Animal products, nec 0.0 0.0 0.0 0.0 0.0 0.0 0.5 0.4 Wool, silk 4.3 0.0 0.0 0.0 4.7 0.7 0.0 0.1 Forestry 0.0 0.0 0.0 0.0 11.7 2.4 0.1 0.3 Fishing 0.0 0.0 0.0 0.0 0.0 0.0 0.1 0.1 Coal 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 Oil 0.0 0.0 0.0 69.4 0.0 48.9 0.0 5.6 Gas 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 Minerals nec 0.0 0.0 0.0 0.0 0.0 0.0 2.5 2.2 Meat: cattle, sheep, goat 0.0 0.0 0.0 0.0 0.0 0.0 0.1 0.1 Meat products nec 0.0 0.0 0.0 0.0 0.0 0.0 0.4 0.3 Vegetable oils and fats 0.0 0.0 55.0 2.8 0.0 0.7 0.2 0.3 Diary products 0.0 0.0 0.0 0.0 0.0 0.0 0.1 0.1 Processed rice 0.0 0.0 9.9 0.5 0.0 0.5 0.0 0.1 Sugar 0.8 0.0 0.0 0.5 0.0 0.4 0.0 0.1 Food products nec 0.0 0.0 0.0 0.1 0.0 0.0 1.4 1.2 Beverages and tobacco 0.0 0.0 2.9 0.2 0.0 0.2 0.1 0.1 Textiles 16.2 0.0 0.0 1.3 8.2 0.7 7.1 6.3 Wearing apparel 0.0 0.0 0.0 0.0 0.0 0.0 0.6 0.5 Leather products 0.0 0.0 0.0 0.0 0.0 0.0 1.6 1.4 Wood products 0.0 0.0 0.0 0.0 8.8 0.8 0.6 0.6 Paper products 0.0 0.0 0.0 0.1 0.0 0.0 4.0 3.5 Petroleum, coal prods 20.8 0.0 0.0 14.8 0.0 3.4 1.3 1.6 Chemical, rubber, plastic 17.0 0.0 19.1 8.1 4.2 0.5 17.9 15.9 Mineral products nec 0.0 0.0 0.0 0.0 0.0 0.0 1.1 1.0 Ferrous metals 0.0 0.0 0.0 0.0 62.1 29.9 0.9 4.3 Metals nec 0.0 0.0 0.0 0.0 0.0 0.0 5.2 4.6 Metal products 0.0 0.0 0.0 0.0 0.3 0.0 2.0 1.7 Motor vehicles and parts 8.7 0.0 0.0 0.0 0.0 2.4 1.6 1.7 Transport equipment nec 0.2 1.6 0.0 0.0 0.0 0.1 1.8 1.6 Electronic equipment 26.8 41.1 0.0 0.0 0.0 6.0 23.8 21.7 Machinery and equip 3.8 57.3 0.0 0.3 0.0 1.6 22.2 19.8 Manufactures nec 0.0 0.0 0.0 0.0 0.0 0.0 0.7 0.6 Total 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 230 China and the WTO TABLE 13.A.4 Elasticity of Substitution and Changes in Real Consumer Prices due to China's WTO Accession Elasticity of Substitution between Domestic Short Run Closure Long Run Closure Products and Imports* 1995­2007 2001­2007 1995­2007 2001­2007 Rice 4.4 2.2 0.7 2.5 1.0 Wheat 4.4 -0.8 0.7 0.4 0.4 Feed grains 4.4 13 2.1 13.5 1.9 Vegetables and fruits 4.4 0.9 -0.6 1.4 -0.1 Oilseeds 4.4 -6.7 -5.9 -4.8 -4.6 Sugar 4.4 -2.1 -3.5 -1.8 -3.1 Plant based fibers 4.4 2.1 4.1 3.1 3.1 Livestock & meat 5.0 3.8 0.7 2.4 0.2 Dairy 4.4 2 -0.5 2 0.2 Other food 4.4 0.3 -2.7 0.4 -1.8 Beverages & tobacco 6.2 -14.3 -7.7 -10.8 -6.9 Extractive industries 5.6 2.5 1.7 2.3 1.2 Textiles 4.4 -10.3 -1.5 -10.5 -3.1 Apparel 8.8 -6.7 0.8 -8.6 -1.9 Light manufacturing 8.8 -2 0.5 -2.6 0.0 Petrochemical industry 4.1 0.7 0.8 0.9 0.9 Metals 5.6 1.2 1.3 1.3 1.3 Automobiles 10.4 -23.6 -4.0 -22.2 -4.2 Electronics 5.6 -5.3 -1.4 -5.8 -1.7 Other manufactures 5.5 0.5 0.8 0.4 0.9 Trade and transport 3.8 3 1.7 2.6 1.7 Construction 3.8 2.8 1.7 2.5 1.7 Communication 3.8 3.6 1.7 2.8 1.9 Commercial services 3.8 2.7 0.9 3.2 1.9 Other services 4.0 2.4 1.3 2.5 1.6 Economic Impacts of China's Accession to the WTO 231 TABLE 13.A.5 Changes in China's Key Economic Indicators in the Baseline, without WTO Accession, 1995­2007 Output Employment Exports Imports (Percent) (Percent) (Percent) (Percent) Rice 63.8 -11.5 134.7 -8.8 Wheat 81.4 6.4 -15.2 126.3 Feed grains 109.5 23.8 -0.6 95.9 Vegetables and fruits 98.2 16.8 -10.8 122.1 Oilseeds 100.9 18.4 -36 151.7 Sugar 112.5 14.5 109.4 88.7 Plant based fibers 137.2 41.1 -8.5 146.1 Livestock & meat 121.9 25.6 12.8 135.3 Dairy 122.5 18.8 60.5 100.3 Other food 110.8 -1.5 76.8 58.5 Beverages & tobacco 114.6 -9 166.7 65.9 Extractive industries 77.8 67.2 -88 554.2 Textiles 142.2 -1.7 95 70.4 Apparel 110.5 -1.7 100.8 47.2 Light manufacturing 135.8 11 117.3 60.1 Petrochemical industry 126.7 -5.6 80.2 101.5 Metals 144.4 11.5 114.4 99.6 Automobiles 157.5 2.2 419.7 105.3 Electronics 186.5 13.1 168.1 112.1 Other manufactures 161.1 6.4 195.1 51.3 Trade and transport 129.4 -3.3 129.4 69.3 Construction 113.3 26.5 77.4 85.6 Communication 133.2 -12.4 452.9 11.9 Commercial services 133.4 1.8 211.7 55.1 Other services 113.6 21 101.1 75.1 Total 120.2 N/A 131.7 97 232 China and the WTO TABLE 13.A.6 Changes in China's Key Economic Indicators due to WTO Accession for the Period before 2001 (1995­2001) Trade Wholesale Consumer Output Employment Exports Imports Balance Prices Prices % % % % US$ m. % % Rice -0.7 -0.7 9.5 -11.3 101 0.5 1.5 Wheat -2.8 -2.9 14.9 39.1 -484 0.1 0.0 Feed grains 2.3 2.4 2.5 -20.9 244 0.8 11.4 Vegetables and fruits -0.6 -0.7 8.6 -26.5 486 0.5 1.5 Oilseeds -1.8 -1.9 10.9 16.6 -549 0.5 -0.2 Sugar -0.1 -0.2 9.1 3.3 8 0.6 1.3 Plant based fibers -11.6 -12.2 29.6 -9.8 264 -1.2 0.0 Livestock & meat 1.3 1.4 8.7 -27.3 1888 1.1 2.2 Dairy 2.0 2.2 10.3 2.6 -10 0.8 1.8 Other food 4.6 4.8 10.9 -26.3 3537 0.3 2.2 Beverages & tobacco -23.8 -23.8 16.8 614.9 -10656 -0.5 -4.2 Extractive industries -0.4 -0.5 16.6 1.9 -781 0.1 1.1 Textiles -12.1 -12.2 16.7 35.5 -9556 -3.0 -7.6 Apparel -3.4 -3.9 23.7 344.6 -344 -1.9 -6.8 Light manufacturing 3.7 3.6 14.6 48.9 2008 -0.5 -2.6 Petrochemical industry -0.7 -0.7 12.9 11.3 -4433 -0.6 0.0 Metals -0.5 -0.5 20.9 23.5 -3051 -0.6 0.0 Automobiles 2.8 -13.2 326.8 195.5 2935 -15.9 -18.8 Electronics 3.8 3.6 18.4 16.2 2657 -1.8 -4.2 Other manufactures 1.9 1.9 17.6 22.2 8511 -0.8 -0.5 Trade and transport 1.3 1.3 10.1 -7.3 5284 -0.2 0.9 Construction 2.2 2.2 14.8 -6.3 301 -0.2 0.8 Communication 1.0 1.0 11.1 -7.4 229 -0.1 0.9 Commercial services 0.8 0.8 10.5 -6.3 798 0.3 1.3 Other services 0.3 0.4 12.7 -8.1 749 -0.1 0.9 Total 2.0 N/A 20.3 22.0 138 -1.4 -0.8 Economic Impacts of China's Accession to the WTO 233 TABLE 13.A.7 Sensitivity Analysis with Respect to the Elasticity of Transformation (2.67) (percentage changes in China after 2001) Wholesale Consumer Output Employment Exports Imports Prices Prices (Percent) (Percent) (Percent) (Percent) (Percent) (Percent) Rice -2.2 -2.5 4.2 -6.3 -0.8 1.1 Wheat -2.4 -2.8 15.6 -9.2 -1.5 0.6 Feed grains -2.5 -2.9 -78.2 -1.9 -1.7 1.8 Vegetables and fruits -3.5 -3.9 12.3 -5.4 -1.7 0.1 Oilseeds -8.3 -8.9 26.8 21.3 -2.6 -4.5 Sugar -6.9 -8.1 11.8 24.4 -1.7 -3.0 Plant based fibers 15.7 16.3 -52.3 8.0 0.1 3.1 Livestock & meat 1.2 0.8 12.6 -7.6 -1.4 0.4 Dairy -2.3 -2.9 11.5 24.4 -1.3 0.3 Other food -6.2 -6.7 10.2 63.3 -1.7 -1.8 Beverages & tobacco -33.2 -33.2 8.8 112.9 -1.9 -7.0 Extractive industries -0.9 -1.2 7.6 -4.4 -0.8 1.1 Textiles 15.6 15.6 32.6 38.6 -1.8 -3.2 Apparel 57.4 56.4 106.1 30.7 -0.7 -2.0 Light manufacturing 3.3 3.3 5.5 6.9 -0.9 0.0 Petrochemical industry -2.3 -2.2 3.2 11.7 -0.8 0.7 Metals -2.0 -1.9 4.0 6.8 -0.5 1.2 Automobiles 1.5 -2.0 27.9 24.1 -4.0 -4.3 Electronics 0.8 0.7 6.8 6.9 -1.4 -1.8 Other manufactures -2.0 -2.0 4.3 18.8 -0.7 0.7 Trade and transport 0.0 0.1 1.0 -0.4 -0.3 1.5 Construction 0.9 1.0 3.0 17.4 -0.3 1.6 Communication -0.5 -0.3 -0.3 10.9 0.0 1.8 Commercial services -2.0 -1.9 -0.2 35.4 0.1 1.8 Other services -1.6 -1.7 1.7 33.5 -0.2 1.5 Total 1.0 0 17.5 17.2 -0.8 -0.1 Nonfarm unskilled wages 1.08 Migration 7* Unskilled wage 0.7 Farm unskilled wage -0.02 Land rent -5.7 Skilled wage 0.8 Price of capital goods -0.95 Capital rent 1.2 Welfare 9728** *1997 US millions. **Millions of workers leaving their farm jobs for nonfarm jobs. 234 China and the WTO Notes 10. According to GTAP version 4, 14 percent of imports were for final consumption and according to China's Customs 40 per- 1. Because China's accession was a necessary condition for cent of imports were ordinary imports that were not duty that of Taiwan, China, and because of the strong trade linkages exempt. This means that approximately 26 percent were ordi- between the two economies, the impact of Taiwan, China's nary imports used as intermediates. Also according to GTAP accession to the WTO is also considered. version 4, China's firms exported an average of 10 percent of 2. Among the studies using a general equilibrium approach their output, implying that only about 3 percent of imports were to quantify the impact of China's WTO accession are Lejour used for production of ordinary exports. (2000); Zhai and Li (2000); Li and others (2000); McKibbin and 11. The GTAP version 5 database (Dimaranan and Tang (2000); Ianchovichina and Martin (2001); Walmsley McDougall 2002) is the source for the elasticities of substitution and Hertel (2001); Deutsche Bank (2001); Wang (2002); between domestic and composite imported commodities in the Zhai and Wang (2002); Walmsley, Hertel, and Ianchovichina Armington production structure of a sector. The values for these (2004); and Ianchovichina and Walmsley (2003). elasticities are shown in column 1 of Appendix table 13.A.4. 3. Huang, Rozelle, and Min (2004) find that nominal rates of 12. In a more recent work Sicular and Zhao (2004) estimate protection on important agricultural commodities (rice, vegeta- the responsiveness of rural labor supply to changes in agricul- bles and fruits, livestock and meat) were negative in 2001 and are tural returns. They present two "push" elasticities--2.67 for likely to remain unchanged in the post-accession period. Conse- nonagricultural wage employment and 0.24 for nonagricultural quently, the reduction in agricultural protection is likely to be nonwage employment. Focusing on the push elasticity for nona- far less than presented in earlier studies. Nonetheless, greater gricultural wage employment and testing the sensitivity of the scope for imports is likely for a range of agricultural products results by replacing the elasticity of 1.32 used in our analysis (wheat, oilseeds, sugar, and dairy products), which are protected with 2.67 leaves the aggregate results largely unchanged (Appen- by tariffs that are scheduled to be reduced substantially and for dix table 13.A.7). The greater responsiveness of labor movement products (cotton and feedgrains) for which export subsidies are implied by the larger elasticity of transformation (2.67) trans- ruled out. These important findings were not incorporated in lates into better poverty and inequality outcomes since farm earlier studies. wages remain nearly unchanged and an additional 1 million 4. The current treatment differs from that in Zhai and Wang farm workers leave farming. (2002), who represent imperfect labor mobility in a single coun- 13. Applied rates at the end of the implementation period try model with endogenous urban unemployment but do not were estimated as the lesser of the bindings and 2001 applied differentiate between skilled and unskilled labor and employ a rates. In virtually all cases, the bindings were lower than the low level of elasticity of labor mobility between rural and urban applied rates. areas (0.25). The results here are similar to those in Zhai and 14. Francois and Spinanger base their estimate of the 20 per- Wang (2002) in the case of WTO accession with high labor cent productivity increase on the distribution of current plants mobility (elasticity of labor mobility is doubled) and fixed in China and apply the formula ln(AC) = CDR · ln(Q), urban unemployment (flexible urban wage). where AC is average cost, MC is marginal cost, Q is the quantity 5. GTAP is a standard global applied general equilibrium produced, and CDR is the inverse elasticity of scale, defined as model with perfectly competitive markets and constant returns CDR = -(AC - MC)/AC, and varies between .125 and .135 to scale technology. The model represents consumer demands (the range of values found in engineering studies). Then they through a constant difference of elasticities functional form and calculate an average cost index for the industry. Assuming that on the supply side emphasizes the role of intersectoral factor the index is 100 at 350,000 units per plant, current plant struc- mobility in the determination of sectoral output. Product differ- ture yields a cost index of roughly 120. entiation between imported and domestic goods and among 15. Mattoo (2004) argues that China's commitments on imports by region of origin allows for two-way trade in each services were the most comprehensive ever made in the WTO. product category, depending on the ease of substitution between 16. In a recent paper Mai, Horridge, and Perkins (2003) products from different regions. Land, capital, skilled and estimate productivity increases over 10 years at 1.8 percent a unskilled labor, and, in some sectors, a natural resource factor year for the strategic manufacturing industries and 2.7 percent a are used in production and are fully employed. year for the services sectors as reforms take place under WTO 6. Some sectors, particularly the service sectors, do not par- accession. ticipate in export processing arrangements and so are not 17. These quotas are represented in the analysis as an export exempt from duties on intermediate inputs used in the produc- tax. In some cases the proceeds of this implicit export tax are tion of exports. redistributed to quota holders, who may be quite different from 7. In a deterministic world, a producer of exports will always the producers and exporters of the goods. In other cases the quo- take advantage of duty exemptions or rebates unless the admin- tas are auctioned, with the quota rents accruing to the govern- istrative costs are excessive, which does not appear to be the case ment. In either case, the marginal return from additional output in China. Many studies have either ignored the problem or, as in of textiles and apparel is net of the quota rent/export tax. Lejour (2000), treated duty exemptions as simple reductions in 18. This adjustment was made with ALTERTAX (Malcolm initial tariffs, instead of exemptions on imports used specifically 1998), so that the consistency and the shares in the GTAP data- in the production of exports. base would be preserved. 8. Export processors can sell locally, but they have to pay 19. The fixed employment assumption may understate the duty/VAT on the imported inputs used in the production of out- costs of accession to some degree. Zhai and Wang, who explore put sold on the domestic market. the impact of WTO accession on migration and unemployment 9. The export processing arrangements did not prevent firms (2002), conclude that structural unemployment may rise follow- producing mainly for the domestic market from exporting. ing China's WTO accession as farmers move to urban areas. These firms produced exports, known as "ordinary" exports, 20. The assumption of fixed trade balance as a share of GDP using mainly domestic inputs and only a small portion of duty- is required when evaluating welfare impacts using a static trade or VAT-paid imported materials (Ianchovichina 2003). model such as GTAP. Economic Impacts of China's Accession to the WTO 235 21. This is a consensus finding supported by Ianchovichina References and Martin (2001), Deutsche Bank (2001), Wang (2002), and Ianchovichina and Walmsley (2003). Bach, C., and W. Martin. 2001."Would the Right Tariff Aggrega- 22. The model underestimates the potential expansion and tor for Policy Analysis Please Stand Up?" Journal of Policy efficiency increase in the service sectors.With its promise to elim- Modeling 23: 621­35. inate over the next few years most restrictions on foreign entry Bach, C., W. Martin, and J. Stevens. 1996. "China and the WTO: and ownership, as well as most forms of discrimination against Tariff Offers, Exemptions and Welfare Implications." foreign firms (Mattoo 2004), China has set the stage for increases Weltwirtschaftliches Archiv 132(3): 409­31. in foreign investment and productivity in these sectors. This in Bhattasali, D., Li, Shantong, and Martin, W., eds., 2004. China turn could lead to much larger income gains from WTO acces- and the WTO: Accession, Policy Reform, and Poverty Reduc- sion and larger increases in wages of skilled workers than shown tion Strategies. Washington, D.C.: World Bank and Oxford here (see Walmsley, Hertel, and Ianchovichina 2004). University Press. 23. This estimate represents the number of "effective" farm Chen, S., and M. Ravallion. 2004. "Household Welfare Impacts workers likely to migrate from rural to urban areas based on of China's Accession to the WTO." In Bhattasali, D., Li, employment data for 2000 from China Statistical Yearbook (NBS Shantong and Martin, W., eds., China and the WTO: Acces- 2001, pp. 111­12). sion, Policy Reform, and Poverty Reduction Strategies. Chap- 24. The consumption tax is close to nondistortionary, ter 15. Washington, D.C.: World Bank and Oxford University because it applies at the same rate to all components of private Press. and government consumption but not investment. Because the Claro, S. 2001. "Tariff and FDI Liberalization: What to Expect GTAP Version 5 database appears to represent the VAT on from China's Entry into WTO?" Paper presented at the domestic production as an output tax, the model generates tax Eighth Annual Conference on Empirical Investigations in losses from the contraction of some industries (e.g., the tobacco International Trade, Purdue University, West Lafayette, Ind., and alcohol industries). These inward-oriented industries have November 9­11. higher average VAT rates than export-oriented sectors such as Deutsche Bank. 2001. "Quantifying the Impact of China's WTO clothing because the VAT is exempted on exports. When the Entry." DB Global Market Research. [www.deutsche-bank. export-oriented sectors expand, the net impact of WTO acces- de/index_e.htm.] sion is a sharp contraction in tax revenues. In reality, such a con- Dimaranan, B., and R. McDougall. 2002. Global Trade, Assis- traction will not be observed, because VATs of the same magni- tance, and Production: The GTAP 5 Data Base. West Lafayette, tude are levied on imports. To offset this impact, particularly in Ind.: Purdue University, Center for Global Trade Analysis. our poverty analyses, we had to adjust the consumption tax in a Francois, J., and D. Spinanger. 2001. "Greater China's Acces- downward direction. We first computed the consumption tax sion to the WTO: Implications for International Trade/ that compensates for the loss in output taxes. This tax as a share Production and for Hong Kong." Paper prepared for the of the total replacement tax is equal to the share of the output Hong Kong Trade Development Council, Hong Kong, tax loss in the total tax losses. Second, we adjusted the consump- December 2001. tion tax rate to eliminate the component stemming from the ------. 2004. "Regulated Efficiency, WTO Accession and the change in output taxes. Motor Vehicle Sector in China." This issue. 25. High unemployment due to the restructuring of state- Gilbert, J., and T. Wahl. 2001. "Applied General Equilibrium owned-enterprises, privatization, and fierce competition in Assessments of Trade Liberalization in China." World China imply that WTO accession may dampen the effect on the Economy 25(5): 697­731. wages of unskilled workers. By assuming full employment, Hertel, T., ed. 1997. Global Trade Analysis: Modeling and Applica- the model overestimates the increase in wages of workers in the tions. Cambridge: Cambridge University Press. nonfarm sectors, underestimates the fall in the wages of farm Huang, J., S. Rozelle, and C. Min. 2004. "The Nature of Distor- workers, and overestimates the increase in total welfare. tions to Agricultural Incentives in China and Implications 26. These estimates are in agreement with findings in of WTO Accession." In D. Bhattasali, Shantong Li, and Ianchovichina and Martin (2001) and Wang (2002). These are W.Martin,eds.,China and theWTO: Accession, Policy Reform, conservative estimates since they do not reflect income increases and Poverty Reduction Strategies. Chapter 6. Washington, resulting from trade- and foreign direct investment­induced D.C.: World Bank and Oxford University Press. productivity gains, especially gains associated with liberalization Hussain, A. 2004. "Impact of WTO Membership on Inequality of China's service sectors (Walmsley, Hertel, and Ianchovichina and Poverty." In D. Bhattasali, Shantong Li, and W. Martin, 2004). Transaction cost savings from developing institutions eds., China and the WTO: Accession, Policy Reform, and compatible with an open and modern market could be very Poverty Reduction Strategies. Chapter 17. Washington, D.C.: large as well, but were not factored into the analysis. World Bank and Oxford University Press. 27. Zhai and Wang (2002) obtained similar results for a com- Ianchovichina, E. 2003. "GTAP-DD: A Model for Analyzing bination of WTO accession and full labor market reform. Trade Reforms in the Presence of Duty Drawbacks." GTAP 28. Since the tax on nonfarm employment of 34 percent rep- Technical Paper 21. Purdue University, West Lafayette, Ind. resents a bundle of policies that act as a barrier to rural-urban [www.gtap.org/resources/tech_papers.asp] and Journal of migration, this estimate is representative of the likely impact and Policy Modelling (forthcoming). could change depending on the policy mix the government Ianchovichina, E., and W. Martin. 2001. "Trade Liberalization in adopts. China's Accession to WTO." Journal of Economic Integration 29. The model does not track education spending as a com- 16(4): 421­45. ponent of the government budget constraint. Ianchovichina, E., and T. Walmsley. 2003. "Impact of China's 30. Claro (2001) estimates that in the apparel and footwear WTO Accession on East Asia." Policy Research Working industries,the adoption of foreign technology raises productivity Paper 3109. World Bank, Washington, D.C. by 30­62 percent in collective enterprises and 20­59 percent in Lardy, N., ed. 2002. Integrating China in the Global Economy. state enterprises. Washington, D.C.: Brookings Institution Press. 236 China and the WTO Lejour, A. 2000."China and the WTO: The Impact on China and Shi,Xinzheng.2002."Empirical Research on Urban-Rural Income the World Economy." The Netherlands Bureau for Economic Differentials: A Case of China."Peking University, Beijing. Policy Analysis. The Hague, Netherlands. Sicular, T., and Y. Zhao. 2002. "Employment, Earnings and Li, S., Z. Wang, F. Zhai, and L. Xu. 2000. WTO: China and the Labor Market Responses in Rural China." Mimeo, Peking World. Beijing: China Development Press. University and the University of Western Ontario. Long, Yongtu. 2000. "On the Question of Our Joining the World ------. 2004. "Labor Mobility and China's Entry to the WTO." Trade Organization." The Chinese Economy 33(1): 5­52. In D. Bhattasali, Shantong Li, and W. Martin, eds., China and Mai,Y., M. Horridge, and F. Perkins. 2003."Estimating the Effects the WTO: Accession, Policy Reform, and Poverty Reduction of China's Accession to the World Trade Organization." Strategies. Washington, D.C: World Bank and Oxford CoPS General Paper, no. G-137. Centre of Policy Studies and University Press. Impact Project, Clayton, Australia. Walmsley, T. L., and T. W. Hertel. 2001. "China's Accession to Malcolm, G. 1998. "Adjusting Taxes in the GTAP Data Base." the WTO: Timing is Everything." World Economy 24(8): GTAP Technical Paper 12. Center for Global Trade Analysis, 1019­49. Purdue University, West Lafayette, Ind. [www.gtap.org/ Walmsley, T. L., T. Hertel, and E. Ianchovichina. 2004."Assessing resources/tech_papers.asp.] the Impact of China's WTO Accession on Foreign Owner- Martin, W. 2001. "Implications of Reform and WTO Accession ship." Pacific Economic Review (forthcoming). for China's Agricultural Trade Policies." Economics of Transi- Wang, Z. 2002."WTO Accession,`Greater China' Free Trade Area tion 9(3): 717­42. and Economic Relations across the Taiwan Strait." Paper pre- Martin, W., D. van der Mensbrugghe, and V. Manole. 2003. sented at the Fifth Conference on Global Economic Analysis, "Keeping the Devil in the Details: Aggregation and the Bene- June 5­7, Taipei. fits of Trade Liberalization." Presented at the Sixth Annual ------.2003."The Impact of China'sWTOAccession on Patterns Conference on Global Economic Analysis, The Hague, June of World Trade."Journal of Policy Modeling 25(1): 1­41. 12­14. Available at http://www.gtap.agecon.purdue .edu. World Bank. 1997. China Engaged: Integration with the World Mattoo, A. 2004. "China's Accession to the WTO: The Services Economy. Washington, D.C. Dimension." In D. Bhattasali, Shantong Li, and W. Martin, ------. 1999. World Development Indicators. Washington, D.C. eds., China and the WTO: Accession, Policy Reform, and ------. 2002. China: Country Economic Memorandum. Poverty Reduction Strategies. Washington, D.C.: World Bank Washington, D.C. and Oxford University Press. WTO (World Trade Organization). 1994. The Results of the McDougall, R., A. Elbehri, and T. Truong. 1998. Global Trade, Uruguay Round of Multilateral Trade Negotiations. Geneva. Assistance, and Production: The GTAP 4 Data Base. Purdue ------. 2001. "Accession of the People's Republic of China." University, Center for Global Trade Analysis, West Lafayette, WT/L/432 (www.wto.org). Ind. Zhai, Fan, and S. Li. 2000. "The Implications of Accession to McKibbin, W., and K. Tang. 2000. "Trade and Financial Reform WTO on China's Economy." Paper presented at the Third in China: Impacts on the World Economy." World Economy Annual Conference on Global Economic Analysis, June 23(8): 979­1003. 28­30, Monash University, Melbourne, Australia. Messerlin, P. 2004. "China in the WTO: Antidumping and Zhai, Fan, and Z. Wang. 2002. "WTO Accession, Rural Labour Safeguards." In D. Bhattasali, Shantong Li, and W. Martin, Migration and Urban Unemployment in China." Urban eds., China and the WTO: Accession, Policy Reform, and Studies 39(12): 2199­217. Poverty Reduction Strategies. Washington, D.C.: World Bank and Oxford University Press. NBS (National Bureau of Statistics of China). 2001. China Sta- tistical Yearbook. People's Republic of China. China Statistics Press. Part III Impacts on Households and on Poverty 14 Earnings and Labor Mobility in Rural China: Implications for China's Accession to the WTO Terry Sicular and Yaohui Zhao The literature on trade liberalization in developing benefits agriculture and if land holdings are highly countries contains divergent views on the impact concentrated, then inequality could increase. Third, of liberalization on employment, incomes, and the distributional impact of trade liberalization poverty. Although most studies find aggregate wel- depends on the flexibility of domestic markets, fare gains, they disagree over the distribution of especially for labor. Gains from trade liberalization these gains among households. One view is that are less likely to be shared equally where labor mar- trade liberalization generates broad-based employ- kets are segmented and barriers hinder the move- ment gains across regions for skilled and unskilled ment of labor among sectors. labor, and that consequently income gains are These considerations are relevant to the impact shared widely (Dollar and Kraay 2001). Under of China's accession to the World Trade Organiza- these conditions, trade liberalization contributes tion (WTO) and its concomitant trade liberaliza- to reductions in income inequality and poverty. tion on employment and incomes in China. Indeed, The alternative view is that employment and the Chinese case has some interesting features. income gains go disproportionately to the already China's labor markets have historically (under better-off groups, with negative implications for socialism) been inflexible and highly segmented. inequality and perhaps also for the poor (Rodrik Domestic economic reforms have allowed greater 2000). labor mobility, but many observers believe that The reality probably lies somewhere in between, substantial institutional barriers to the movement with the distributional outcome depending on spe- of labor persist. Also, in China certain assets such as cific conditions in the country in question (Winters land and education, while not equally distributed, 2000). First, it depends on the level and structure of are nevertheless relatively equally distributed by pre-liberalization trade barriers, which determine developing country standards. These two features the sectors that gain and lose. A differential impact would have counterbalancing effects in that the on sectors holds implications for the distribution former would tend to cause the gains from WTO of gains among regions, skill levels, and income accession to be concentrated, while the latter would groups. Second, it depends on the preexisting dis- tend to cause the gains to be shared more broadly tribution of assets--that is, of land, capital, and compared with the gains from similar liberalization humancapital.So,forexample,if tradeliberalization in other countries. 239 240 China and the WTO This chapter examines the microeconomic deter- effects of trade liberalization on levels of employ- minants of rural employment and incomes in China. ment and earnings. The income generation func- More specifically, using survey data we estimate tions give estimates of the impact of agricultural income, wage, and labor supply functions for rural versus nonagricultural employment on income households. Because the households derive income from labor. We find that nonagricultural employ- from agriculture and sideline family businesses and ment generates substantially more income per hour since few households hire labor to undertake such worked than does agricultural employment. This activities(BowlusandSicular2003),wemustimpute result is consistent with the findings of other studies shadow wages derived from self-employment. (e.g., Knight and Song 1997; Meng 2000; Michelson Together, the income, wage, shadow wage, and labor and Parish 2000), which generally conclude that supply functions empirically describe household income inequality among households reflects dif- income generation from employment. ferences in access to higher paying, off-farm jobs. Our analysis fills a gap in the literature. Even We take this analysis one step further and though the literature examining employment and decompose the income gap between households in earnings in rural China is now quite substantial, richer and poorer subsamples. Our decomposition most studies of China's rural employment analyze reveals that the income gap is accounted for by the determinants of occupational status--that is, both differences in hours worked in different occu- whether or not individuals participate in different pations and differences in the estimated returns types of work such as wage jobs or nonagricultural to labor in different occupations--that is, poorer sidelines (examples are Hare 1994, 1999a, 1999b; households are poorer both because they supply Knight and Song 1997, 1999; Michelson and Parish less labor to higher-paying wage employment and 2000, Parish, Zhe, and Li 1995; Rozelle and others because the returns to the labor they supply in each 1999; Zhao 1999a, 1999b). Relatively few studies occupation, and their returns to education, are estimate the rural labor supply per se--that is, lower. These findings suggest that the impact of hours or days worked (examples are Knight and China's accession to the WTO on income distribu- Song 1997, and Yao 1999). Even fewer studies esti- tion will depend on how it affects the distribution mate labor supply as a function of wages, the rela- of employment, wages, and the returns to educa- tionship of greatest interest here. tion among households. To our knowledge, the only study that estimates Our labor supply estimates reveal how work labor supply for rural China and includes a measure hours, and the composition of work hours, would of the wage as an explanatory variable is by Meng respond to changes in wages for agricultural and (2000). Like Meng, we estimate time worked as a nonagricultural work. We find that wages do not function of wages and other variables.Unlike Meng, have a large effect on the total number of hours and indeed unlike the relevant literature for other worked. They do, however, significantly influence developing countries (e.g., Jacoby 1993; Skoufias the composition of hours worked. Our estimates of 1994), our labor supply functions allow for the pos- labor supply by type of employment--in house- sibility that wages or shadow wages, and labor's hold agriculture, household nonagriculture, and response to these wages,can differ depending on the off-farm wage jobs--indicate that most own- and type of wage. Thus it is possible for labor supply to cross-elasticities with respect to the different wages be more responsive to the market wage than to the are well below one. This finding suggests that agricultural shadow wage, or vice versa. Finally, although labor moves among these types of although other studies estimate total labor supply, employment, mobility is limited. we estimate total labor supply and its components-- A notable exception to this pattern is the cross- in this case, labor supply to household agricultural elasticity with respect to the agricultural shadow production, to household nonagricultural produc- wage of labor supplied to the market. This elasticity tion, and to market or wage employment. By esti- is significant and large, indicating that differences mating the components of labor supply, we obtain in labor supply for wage employment among information about how wages and other variables households is driven by differences in the returns to influence the composition of employment. labor in household farming. Put differently, high Our empirical results provide some of the market wages do not"pull"labor out of agriculture; underlying parameters needed to understand the rather, low marginal returns to work in agriculture Earnings and Labor Mobility in Rural China: Implications for China's Accession to the WTO 241 "push" labor into wage employment. We discuss grew slowly both overall and in rural areas. Because these results more fully later in this chapter. of population planning policies and a marked The next section of this chapter presents an decline in fertility in the 1970s, cohorts entering the overview of aggregate trends in China's rural labor force in the 1990s were smaller than those in employment, earnings, and labor markets, paying the 1970s and 1980s. New cohorts entering the special attention to the institutions and policies that labor force should remain small or even decline affect labor mobility and the distribution of earn- further in the coming years because of the strict ings. A description of the dataset follows. The next population control policies adopted in the 1980s. section examines household income generation On the demand side, macroeconomic growth and analyzes the income gap between richer and combined with policy liberalization generated new poorer households. We then turn to our economet- job opportunities for rural workers. Rural nonagri- ric estimates of wage and labor supply functions. cultural employment grew substantially. The offi- Throughout this chapter our focus is on rural cial data on such employment is problematic, but it households. China's accession to the WTO is affect- does provide a rough indication of trends. Rural ing both urban and rural households, but the insti- nonagricultural employment includes employment tutional setting and economic behavior of these in township and village enterprises (TVEs) and in two types of households are substantially different private and individual enterprises.2 During the and require separate analysis. Analysis of rural 1990s, employment in these enterprises grew, on households is important because the rural sector average, 5.4 percent annually. In absolute terms, contains most of China's population and also most during the 1990s the increase in rural enterprise of its poor. Indeed, a disproportionate share of employment exceeded 66 million jobs, and by 1999 China's poor population is located in rural areas.1 this employment was equivalent to one-third of the Recent estimates for 1999 by Chen and Wang rural labor force. (2001) report a poverty rate (using the $1/day A notable aspect of growth in rural enterprise poverty line) of 24.9 percent in rural areas versus employment is that nearly half of it stemmed from only 0.5 percent in urban areas. Despite increased the expansion of private and individual enterprises. urbanization, 64 percent of the population and By 1999 these enterprises were employing over 45 74 percent of employed persons are still classified million people, equivalent to about 10 percent of as rural (National Bureau of Statistics 2001, pp. 37, the rural labor force. Although some of this growth 39). Average per capita income in rural areas is only could be the result of reclassification of collective about one-third that in urban areas, and evidence TVEs as private businesses, it also reflects the grow- suggests a widening of the urban-rural income gap ing importance of household-based nonagricul- in recent years (Yang and Zhou 1999). tural activities. Also notable during the 1990s was a rise in rural­urban migration.After a relaxation of restric- Rural Employment and Earnings in tions on labor movement, rural­urban migration China in the 1990s: The Aggregate appears to have grown substantially. Data on Picture migration are spotty, and definitions of what con- A review of aggregate trends in rural employment stitutes migration differ (see Wu and Zhou 1996; and earnings in China provides a broad context in Rozelle and others 1999), but most estimates sug- which to interpret our microeconomic results. gest at least a doubling of the number of migrants Aggregate trends reflect the impact of major policy between the late 1980s and the mid- or late 1990s. reforms, including domestic market liberalization Sources suggest that by the mid- or late 1990s the as well as trade liberalization in advance of WTO number of migrants (excluding commuters) prob- accession, and so they also provide some clues ably exceeded 50 million, or about 10 percent of the about the potential impact of WTO accession on number of rural employed persons (Wu and Zhou employment and earnings. 1996; Zhao 1999b). Developments in the 1990s on both the supply Altogether, then, growth in employment by and demand side suggest movement toward fuller rural TVEs, by private and individual enterprises, employment and higher earnings in rural areas. On and through migration increased from perhaps 130 the supply side, in the 1990s China's labor force million in the early 1990s to roughly 230 million in 242 China and the WTO the late 1990s.By then,nonagricultural employment hours to different sectors can affect the impact of of rural residents had risen from less than 30 per- trade liberalization. cent to nearly 50 percent of the number of rural What have been the effects of the just-described employed persons. trends in employment on rural incomes and This substantial expansion of nonagricultural inequality? On average, the real per capita income employment has spurred some debate about the of rural households rose during the 1990s. Between nature of labor markets in rural China. Some 1990 and 1999, net income per capita rose about authors argue that labor markets are now fairly open 70 percent. This growth was derived from multiple and competitive, with considerable labor mobility sources. Wages contributed the largest share of (Rawski and Mead 1998). Others, however, argue the increase (38 percent), followed by agriculture that while open and competitive labor markets are (29 percent), and household nonagricultural emerging in some regional pockets, and while gen- sidelines and businesses (22 percent).3 Although erally the direction of change has been toward more nonagricultural sources dominated, the fact that open, competitive conditions, rural labor markets agriculture contributed nearly one-third of the nevertheless continue to be imperfect, and institu- increase in income is notable in view of the fact tional and administrative barriers persist (Parish, that, according to the official data, agricultural Zhe, and Li 1995; Knight and Song 1999). These employment supposedly declined. issues are relevant to our analysis and are discussed The average trends just outlined mask changes further later in this chapter. in distribution among poorer and richer house- Discussion of employment cannot be complete holds. The general consensus is that rural income without mention of the largest employer, agricul- inequality increased during the 1990s, while ture. Official statistics reveal cycles in agricultural poverty decreased. These apparently contradictory employment. The number of rural employed per- developments reflect income growth for the poor, sons rose to more than 340 million in 1991­92, fell but faster income growth for richer groups. to less than 325 million in 1995­97, and then rose A 2001 World Bank study outlines key aggregate again to almost 330 million in 1999­2000 (National factors underlying the recent decline in rural Bureau of Statistics, various years). Interestingly, poverty, some of which are relevant from the per- the upswings in these cycles occurred at the same spective of China's accession to the WTO (World time as downswings in rural enterprise employ- Bank 2001). First, aggregate growth in the gross ment, and vice versa, which suggests an inverse rela- domestic product (GDP) appears to be important. tionship between the two kinds of employment. Nationally, the reduction in poverty coincided with The official data on rural employment are, how- a period of rapid GDP growth, and poverty reduc- ever, problematic. They simply count the number tion occurred more quickly in those regions that of people by primary occupation. They do not cap- experienced the most rapid aggregate growth. Sec- ture the fact that many rural workers engage in ond, the composition of growth matters. The rate multiple occupations and that hours worked in any of poverty reduction has been faster in regions particular occupation can fluctuate over time. where agricultural growth has more or less kept Some studies indicate that changes in hours worked pace with growth in other sectors. This situation in agriculture have been significant (Rawski and reflects the fact that agriculture is the primary Mead 1998; World Bank 2001). Estimates of agri- source of employment and income for the poor. cultural employment based on hours worked indi- These points suggest that the effect of WTO acces- cate that in the mid-1990s, when rural enterprise sion on poverty will be influenced by its impact on employment was expanding rapidly, agricultural both aggregate growth and the sectoral composi- labor days worked actually rose (World Bank tion of that growth. 2001). These estimates indicate that growth in nonagricultural employment and agricultural The Survey Sample employment can occur concurrently. The relation- ship between these two kinds of employment For our empirical analysis, we use data from the is important to our analysis, because the willing- China Health and Nutrition Survey (CHNS). ness and ability of rural residents to supply labor CHNS data were collected through an independent Earnings and Labor Mobility in Rural China: Implications for China's Accession to the WTO 243 survey conducted by an international team of gives some comparable statistics from the 1997 researchers collaboratively sponsored by the National Bureau of Statistics (NBS) official rural Carolina Population Center at the University of household survey, where available. Household size North Carolina at Chapel Hill, China's Institute of and structure are similar for the NBS data and for Nutrition and Food Hygiene, and the Chinese CHNS subsample. Income levels for the CHNS Academy of Preventive Medicine. Data are available subsample are somewhat lower than for the NBS for four years: 1989, 1991, 1993, and 1997. The sur- survey, but this disparity could reflect differences in vey, which used a multistage, random cluster sam- how income is calculated. The structure of income pling method, covers about 3,800 households with also differs. In particular, agricultural income is 14,000 members in nine provinces with different noticeably higher in the NBS survey than in the geographic and economic characteristics.4 The CHNS samples. This difference could reflect the CHNS also offers information on community-level inclusion of households in suburban villages and variables such as market prices, health facilities, and county towns in the CHNS subsample but not in social services.5 the NBS rural survey. For both the NBS and CHNS The CHNS survey data are useful here because samples, the major sources of income are agricul- they provide detailed information on incomes and ture and wage employment. hours worked in different occupations as well as on a The CHNS data provide some information wide range of relevant individual, household, and about relative earnings in different sectors. Dividing community characteristics. The sample includes as average earnings by hours worked for each sector household members migrant workers who work and suggests that the returns to labor in agriculture are liveoutof town,butwhoseearningsandexpensesare lower than in other occupations. Average net earn- considered part of the household's. The data do not, ings per hour worked in agriculture are roughly however, allow us to distinguish between work and 1.1 yuan as compared with 2.2 yuan in nonagricul- income from migrant versus local employment. tural sidelines and 2.8 yuan in wage employment. Thus although we are able to investigate mobility When we use regression analysis to estimate the among types of work (in household agricultural returns to labor in the next section, these earnings production, household nonagricultural production, differentials become larger. and wage employment), we cannot investigate geographic mobility. Income Generation Functions For our estimations we use the most recent (1997) data, and we drop urban households--that A common approach to analyzing the determinants is, we use a subsample of households that includes of income is to estimate an income generation only those that reside in rural and suburban vil- function, where net income is a function of labor lages and in county towns. We include suburban inputs, land and capital assets, and other household villages and county towns because they have close or regional characteristics that contribute to the ties with rural areas, and a significant portion of the generation of earnings. This approach takes income population in these areas holds a rural residence generation as a simple, linear accounting relation- registration or hukou. This subsample covers 3,239 ship, where income equals the sum of household households with 8,590 working-age adults. Because labor and other assets times the returns to those we are interested in labor supply, we further restrict assets. The regression coefficients provide estimates our sample to households that contain at least one of the marginal returns to each asset. We use this household member who works. This reduces the approach both to examine income generation on subsample to 2,998 households with 8,326 working- average and to explore differences between poorer age adults. The number of observations actually and richer subgroups. used in our analysis varies among regressions, The first columns (Model I) of table 14.2 con- depending on the extent of participation in the tain results from income generation regressions for activity being analyzed and on the prevalence of all households with at least one working adult in missing values in relevant variables. the CHNS nonurban subsample. The dependent Table 14.1 contains descriptive statistics for the variable is the sum of net earnings from the three nonurban subsample of the CHNS survey. It also occupations--that is, net earnings (revenue minus 244 China and the WTO TABLE 14.1 Descriptive Statistics, China Health and Nutrition Survey, 1997 NBS Rural CHNS Nonurban Households Household Survey, Standard Number of Variable 1997 Mean Deviation Observations Household size 4.35 3.83 1.42 3,239 Number of adults 2.79 2.67 1.29 3,212 Number of dependents per capita 0.30 0.26 3,212 Number of male adults 1.36 0.80 3,212 Number of female adults 1.31 0.76 3,212 Number of male laborers (hours > 0) 1.18 0.73 3,239 Number of female laborers (hours > 0) 1.10 0.72 3,239 Mean age of working-age adults 35.39 14.30 3,238 Mean schooling of working-age adults 6.94 2.81 3,033 Ratio of male working-age adults to family 0.51 0.18 3,041 working-age adults Cultivated land (mua) 6.84 12.14 1,807 Household total income (yuan) 9,094 9,871 1,966 Household net earnings (yuan) 9,092 7,774 9,784 2,160 Household net earnings from agricultural 5,081 2,069 5,095 2,380 sectors (yuan) Household net earnings from nonagricultural 1,009 1,267 4,627 3,080 sectors (yuan) Household earnings from employed 2,240 3,830 7,835 3,069 work (yuan) Household nonlabor income (yuan) 1,397 2,868 2,808 Household total labor hours 3,786 2,874 2,904 Household total labor hours in household 1,803 2,203 3,054 agricultural production Household total labor hours in household 572 1,504 3,170 nonagricultural production Household total labor hours in wage 1,359 2,115 3,122 employment Heilongjiang 0.12 3,239 Jiangsu 0.12 3,239 Shandong 0.12 3,239 Henan 0.13 3,239 Hubei 0.12 3,239 Hunan 0.12 3,239 Guangxi 0.13 3,239 Guizhou 0.13 3,239 aThe mu is a Chinese measure of area. Fifteen mu equal 1 hectare. Note: The National Bureau of Statistics (NBS) data are taken from the China Statistical Yearbook. The NBS rural survey sample includes only households in rural villages but in all provinces, whereas the subsample of the China Health and Nutrition Survey (CHNS) data includes households in suburban villages and county towns but only in eight provinces. Note that income is calculated differently for the NBS and CHNS samples. Also, the NBS household income data are in per capita terms, and we calculate NBS household total income as mean household size times mean income. This method could create some bias, because household size and income per capita are typically inversely correlated. Nevertheless, comparison of the CHNS means with the NBS means gives a rough indication of how the CHNS sample compares with the official national rural survey sample. Sources: National Bureau of Statistics, various years; authors' calculations using CHNS data. Earnings and Labor Mobility in Rural China: Implications for China's Accession to the WTO 245 0.274 0.304 0.230 cent Error 37.720 26.446 57.421 n.a. 188.010 village in Standard Per 2,679.677 3,416.966 3,386.499 3,352.075 3,658.165 3,375.095 4,252.05 3,617.45 4,368.554 suppressed live 20 is opT average that term of III: . basis vey Model 0.510* 1.958*** 3.004*** 46.073 n.a. 0.757 constant the households 395.775** 447.157 173.030*** 109.221*** 349 sur Coefficient 4,998.626 3,594.468 3,304.311 8,191.833** 5,586.847* 1,270.767 2,781.757 5,023.888 the on all and lds tos the in es refer househo villag cent included, 0.132 0.212 0.153 for cent Error 48.873 58.782 38.577 n.a. per 116.527 are Per group Standard 1,597.452 2,331.068 2,369.045 3,374.844 2,288.171 2,242.68 2,279.672 2,502.903 2,185.661 20 20 we (top) variables bias, distribution Bottom II: bottom dummy income Model 0.043 1.266*** 2.113*** 57.601 n.a. 0.722 selection the earnings) 167.649 200.764*** 231.095*** 333 the Coefficient 3,709.768** 3,922.149* 3,473.656 4,332.018 4,466.833** 3,145.446 1,262.594 4,669.35* 2,520.564 Thus of net regional sample all cent III per 20 and minimize II household 0.101 0.137 0.111 n.a. oT questionnaire. Error 84.215 17.636 16.869 26.727 (top) 728.723 717.698 851.888 748.158 698.346 905.101 750.139 Standard 1,185.776 1,517.58 Models Households In bottom variable: I. All community I: a the decomposition. in in Model Model 0.226** 1.715*** 2.372*** n.a. 0.352 in 75.514*** 33.799 98.739 241.438*** 206.177 150.298*** 770.721 686.911 480.419 Oaxaca leaders income (dependent Coefficient 1,982.589*** 1,015.651 2,167.564** 1,083.532 1,671 variable the net in village by dummy data. Functions household hours to adults contribution reported CHNS hours .10. regional as capita labor adults adults < hours labor yuan) p* per using Generation adults regional labor omitted income working-age .05, the production of (1,000 vations < the of average estimates working-age is Income working-age p production of obser ** have male working-age R-square household of applicable. analysis that employment schooling Authors' 14.2 of equipment age .01, Guizhou Not < ce: age family p capita allow Agricultural Nonagricultural W Mean Ratio Land otalT Mean Heilongjiang Jiangsu Shandong Henan Hubei Hunan Guangxi Guizhou Constant Adjusted Number n.a. *** Note: to per villages Sour ABLET 246 China and the WTO TABLE 14.3 Descriptive Statistics for Variables Used in the Income Generation Functions, CHNS, 1997 All Households Used in the Income Generation Function Bottom 20 Percent Top 20 Percent Standard Standard Standard Variable Mean Deviation Mean Deviation Mean Deviation Household net earnings (yuan) 8,113.19 9,746.09 5,955.50 6,265.58 10,966.54 9,907.75 Household agricultural production labor 1,888.35 2,261.85 2,760.06 2,562.277 778.59 1,642.21 hours Household nonagricultural production 573.11 1,543.39 476.76 1307.40 529.67 1,484.76 labor hours Wage employment labor hours 1,578.73 2,093.83 1,127.63 1,883.53 2,525.19 2,217.55 Household total labor hours 4,040.18 2493.40 4,364.45 2,599.83 3,833.45 2,337.51 Mean schooling of working-age adults 7.20 2.83 6.77 2.91 8.17 2.55 Ratio of male working-age adults to 0.51 0.16 0.52 0.16 0.50 0.15 family working-age adults Cultivated land (mu) 4.80 12.16 4.85 6.75 5.72 13.75 Equipment (1,000 yuan) 2.24 11.64 1.06 4.36 2.92 15.18 Mean age of working-age adults 36.84 7.85 36.20 7.73 37.42 7.54 Heilongjiang 0.16 0.14 0.23 Jiangsu 0.15 0.08 0.23 Shandong 0.09 0.01 0.20 Henan 0.11 0.17 0.04 Hubei 0.14 0.24 0.14 Hunan 0.07 0.10 0.03 Guangxi 0.11 0.06 0.10 Guizhou 0.17 0.22 0.03 Number of observations 1,671 333 349 Source: Authors' estimates using CHNS data. nonlabor variable costs) from household agricul- These results indicate that income differences tural production, net earnings from household among households are generated more by nonagri- nonagricultural production, and wage income. cultural work than by agricultural work; further- Explanatory variables include actual labor hours in more,wage employment brings the highest returns.6 each of household agricultural production, house- Theseresultsareconsistentwiththefindingsof other hold nonagricultural production, and wage studies for China (Knight and Song 1997; Meng employment, as well as education, ratio of males to 2000; Michelson and Parish 2000). females, age, land and equipment assets to capture Education has a significant effect on household nonlabor inputs that contribute to income from income. An additional year of average schooling for household production, and provincial dummy working-age adults in the household increases variables. The mean values of these variables household net income by 241 yuan, roughly 3 per- appear in table 14.3. cent of average net income. Again, this result is Of central interest are the returns to labor in consistent with findings in other studies. different occupations. The estimated coefficient for The average age of household workers and the agricultural hours of work is small,0.23,and signifi- sex compositionof theworkforcedonothavesignif- cant at the 5 percent level. The coefficient for labor icant effects on household earnings. As expected, hours in nonagricultural household production is land and equipment assets have significant, posi- 1.72,and that for hours in wage employment is 2.37, tive effects. An additional mu of land (a Chinese both statistically significant at the 1 percent level. measure in which 15 mu equal 1 hectare) increases Earnings and Labor Mobility in Rural China: Implications for China's Accession to the WTO 247 net income by 76 yuan or 1 percent of mean income, education is 1.4 years longer. Land assets are 0.9 mu and an additional 1,000 yuan of equipment assets larger (although not corrected for land quality), increases net income by 150 yuan or 2 percent of and equipment assets are substantially larger, too. meanincome.Thepresenceof significant,largecoef- Employment also differs. Although both groups ficients on some provincial dummy variables indi- work in all three sectors, household agriculture cates that location of residence is also important. dominates for the poorer group and wage employ- We explore differences between poorer and richer ment dominates for the richer group. Overall, households using a modified Oaxaca-type decom- households in the poorer group work more hours. position. This method involves first estimating Total work time for households in the poorer group income generation functions for richer and poorer is 4,364 hours, about 14 percent higher than the subgroups.To minimize the problem of endogenous 3,833 hours total work time for households in the sample selection, we divide the sample using average richer group.7 village incomes reported by village leaders in a sepa- Table 14.4 gives the results of our Oaxaca-type rate community questionnaire. Models II and III in decomposition.Although conventional decomposi- table 14.2 contain the results of income generation tion combines the contribution of regional dummy equationsforhouseholdsinvillagesinthetop20per- variables together with that of other characteristics, cent and bottom 20 percent of the village average per we separate out the effects of regional differences on capita net income distribution. the income gap.The overall difference of 5,011 yuan The two groups have significantly different coef- in mean net earnings is thus decomposed into three ficients. The returns to labor and education are components: (1) the income gap due to differences substantially higher in the top 20 percent than the in levels of household productive characteristics bottom 20 percent. For the bottom 20 percent, other than regional location, (2) the income gap the returns to agricultural labor are not signifi- due to differences in returns to these household cantly different from zero. This finding would be productive characteristics, and (3) the income gap consistent with a standard surplus labor story. By due to regional factors. contrast, for the top 20 percent the coefficient for The results show that all three components con- agricultural labor is positive and significant, albeit tribute to the income gap, but the differences in small. The nonagricultural and wage labor groups productive characteristics are most important. This have positive and significant returns, but the component explains 82 percent of the income gap returns are lower for the poorer subsample (bottom between the two groups. The largest contributor 20 percent). here is the difference in wage employment hours, Schooling has positive coefficients everywhere, which accounts for 71 percent of the income gap. but the return to education is smaller and insignifi- The richer households have fewer labor hours in cant for the poorer subsample. The ratio of males agriculture (see table 14.4), which serves to reduce among working-age adults is insignificant except the income gap by 9 percent. Differences in educa- for the poorer group, for which it is large, positive, tion levels explain 8 percent of the gap. Differences and significant. The presence of male workers thus in mean endowments of land and productive appears to be most important in poorer areas. The equipment, and in household structure and age, are returns to land and equipment are all positive and relatively unimportant. significant, with the magnitudes higher for the bot- Differences in the returns to productive charac- tom 20 percent than the top 20 percent. This differ- teristics--that is, differences in the estimated coef- ence could stem from the greater scarcity of land ficients of the two groups--account for 34 percent and capital in poorer villages than in richer villages. of the income gap. The largest contributor here is The Oaxaca decomposition combines these esti- the difference in the returns to education, which by mated coefficients with the mean values of the vari- itself explains 34 percent of the income gap. Next ables for the two groups. The mean values of house- most important is the difference in returns to wage hold characteristics reveal substantial differences employment, which explains 32 percent of the between households in richer and poorer villages income gap. Differences in the returns to agricul- (table 14.3). Average net earnings of the rich group tural labor and sex composition of working-age are 1.8 times those of the poor group, and average adults are also important. 248 China and the WTO TABLE 14.4 Decomposition of Household Net Earnings Differentials between Households Belonging to the Top and Bottom 20 Percent Contributions to Income Gap between Households in the Top 20 Percent and Bottom 20 Percent of Villages Yuan Percent Household net earnings, total gap 5,011.04 100.00 Amount due to (1) Differences in productive characteristics 4,116.19 82.14 (2) Differences in returns to productive characteristics 1,723.47 34.40 (3) Differences in regional location and returns to regional 828.61 16.54 location (a) Agricultural production labor hours 462.40 9.23 (b) Nonagricultural production labor hours 85.28 1.70 (c) Wage employment labor hours 3,575.93 71.36 (d) Mean schooling of working-age adults 394.56 7.87 (e) Ratio of male working-age adults to family 19.45 0.39 working-age adults (f) Land 162.50 3.24 (g) Total equipment (1,000 yuan) 316.56 6.32 (h) Mean age of working-age adults 63.22 1.26 (2) Differences in returns to productive characteristics (total) 1,723.48 34.4 (a) Agricultural production labor hours 978.92 19.54 (b) Nonagricultural production labor hours 348.15 6.95 (c) Wage employment labor hours 1,627.69 32.48 (d) Mean schooling of working-age adults 1,703.70 34.00 (e) Ratio of male working-age adults to family 2,121.71 42.34 working-age adults (f) Land 146.48 2.92 (g) Total equipment (1,000 yuan) 242.43 4.84 (h) Mean age of working-age adults 424.36 8.47 (3) Differences in regional location and returns to 828.62 16.54 regional location (total) (a) Differences in regional location 93.21 1.86 (b) Differences in returns to regional location 921.83 18.40 Source: Authors' estimates using CHNS data. Regional location, like other household charac- richer households and because the returns to those teristics, contributes to the income gap in two ways. characteristics are lower. Here the key contributors First, the returns to region of residence or esti- to the income gap are a lower supply of wage labor mated coefficients for the provincial dummy vari- by households in poorer villages, as well as the lower ables differ between the rich and poor groups. And, returns received by poor households for education, second, the regional distribution of the poor differs wage employment, and agricultural labor. from the regional distribution of the rich. Together these regional factors explain about 17 percent of New Estimates of Labor Supply the overall income gap. Virtually all of this is due to and Allocation differences in the returns to location. In summary, our decomposition analysis shows The estimates and decomposition just described that households in poor villages are relatively poor indicate that returns to labor and pattern of both because they have different characteristics than employment are important factors underlying Earnings and Labor Mobility in Rural China: Implications for China's Accession to the WTO 249 income differences. This finding raises questions no longer be equal. Moreover, in such situations the about what explains the level and pattern of labor returns to labor in household production may not supply and why the returns to labor vary among be observable. These returns or shadow wages will sectors and households. Clearly, these two ques- be endogenous and a function of both production- tions are interrelated, because household labor sup- and consumption-side variables. Households thus ply may depend on the returns to labor, and the may simultaneously face three different prices of returns to labor (especially in household produc- labor: a shadow wage for labor in agricultural pro- tion) may depend on household production deci- duction (WA), a shadow wage for labor in nonagri- sions, including decisions on the allocation of cultural production (WN), and an observed wage labor. Here we pursue these questions by empiri- for market employment (WM). cally analyzing the determinants of the returns to Factor markets in China are likely to be imper- labor and of labor supply, and we employ instru- fect, so we adopt an empirical strategy that allows mental variable methods to address the endogene- for this possibility. Specifically, we use the approach ity of key explanatory variables. of Jacoby (1993), who notes that at the household We begin with a model of time allocation by optimum, the household's shadow price of labor rural households. The family is assumed to have would equal the value of the marginal product endowments of workers, land, and other assets. of labor in household production. Labor supply, The household may allocate its resources among then, can be modeled as a function of the shadow three possible income-generating activities: agricul- wages. We thus specify labor supply as a function of tural production, nonagricultural production, and shadow wages, as well as of other relevant variables. wage or market employment. In the first two activi- Also, because we are interested in labor supply to ties, the household organizes production using three activities, we have three different labor supply certain quantities of family assets and labor and functions. These take the form perhaps also purchased inputs, hired assets, and Hi = f (WA, WN, WM, I, X), i ^ ^ hired labor. In market employment the household (14.1) faces an exogenous wage rate set by the market. i = A, N, M Labor supply decisions are the outcome of where WA, WN, WM are the shadow wage in agri- ^ ^ household utility maximization, where utility is a cultural production, the shadow wage in nonagri- function of the consumption of leisure and goods cultural production, and the wage rate in market and the budget constraint depends on full income, employment, respectively; I is household nonlabor which includes profits from household production income; and X is a vector of household characteris- as well as the value of the household's time endow- tics. Total labor supply is the sum of these three ment. Where markets function well and factors are labor supplies. perfectly mobile, households maximize utility by first maximizing profits at market prices and then deciding on optimal levels of leisure and goods con- Empirical Specification sumption given market prices and their endow- The specification of our labor supply functions ments. Labor allocation to the different activities-- basically follows that of Jacoby and others,9 except hours worked in agricultural production (HA), that we include multiple wages. The functions for nonagricultural production (HN) and wage or mar- labor supply to household agricultural production, ket employment (HM)8--are the outcome of this household nonagricultural production, and wage decision process. In theory, the allocation of time or market employment are, respectively, among these activities should equalize the returns to labor in them. Optimal leisure consumption is log HA = 0 + A ln WA + N ln WN ^ ^ the difference between the household's time endow- + M ln WM + I I (14.2) ment and total time worked in all the activities. + X X + A Where markets are imperfect, where transaction ^ ^ costs or barriers to mobility are present, or where log HN = 0 + A ln WA + N ln WN households have preferences for certain types of + M ln WM + I I (14.3) work, the returns to labor in different activities may + X X + N 250 China and the WTO log HM = 0 + A ln WA ^ corrected estimation. In the IV or 2SLS estimations, + N ln WN + M ln WM ^ labor and variable cost inputs are treated as endoge- (14.4) nous. Instruments used in both the agricultural + I I + X X + M and nonagricultural production functions are Overall labor supply (H) is the sum of labor exogenous variables for the household and the com- supply into three sectors. Alternatively, it can be munity, including household composition variables estimated directly as (number of working-age adults, children, and eld- erly) and local market prices of vegetables, pork, log H = 0 + A ln WA + N ln WN ^ ^ chicken, and gasoline. In the nonagricultural pro- + M ln WM + I I (14.5) duction function we also include as instruments the + X X + local market prices of honey-combed coal bri- quette, coal lumps, coal powder, and liquefied natu- The signs of the effects of the three wage rates on ral gas, and the market wage for unskilled labor. The total labor supply depend on the relative magni- selectivity-bias corrected estimation treats whether tude of substitution and income effects. the household engages in the activity as an endoge- The estimation of equations 14.2­14.5 requires nous decision. The identification variable is the knowledge of wage or shadow wage rates in all number of dependents in the household,and we use activities. For those engaged in agricultural or the Heckman method (Heckman 1979).As noted in nonagricultural self-employment, shadow wages tables 14.5 and 14.6, the IV estimation does not can be calculated using estimated parameters from pass the Hausman joint exogeneity test for either household production functions (Jacoby 1993). We agricultural or nonagricultural production. Selec- therefore begin with estimates of production func- tivity terms in the selectivity-bias correction estima- tions in the next section. We then present our esti- tions (lambda statistics reported for Model III) are mates of the wage and shadow wage functions. We also statistically insignificant for both equations. use these estimates to project (shadow) wages for The first column of table 14.5 reports OLS esti- households that do not participate in one or more mates of the agricultural production function. The activities. These projected (shadow) wages are results indicate that variable inputs are the domi- employed in our analysis of labor supply. nant contributor to agricultural output, with an elasticity of 0.46. Land is the next most important contributor to agricultural production, with an Estimation of Production Functions elasticity of 0.32, and labor hours have an elasticity We use the standard Cobb-Douglas functional of 0.19. The coefficient for agricultural equipment form to estimate the production functions for agri- is also significant. These estimates are similar in cultural and nonagricultural activities. For each of magnitude to those in other studies (e.g., Yang these activities the production function takes the 1997; Li and Zhang 1998). The coefficients of these form inputs add up to 0.999, which indicates that agri- n m culture displays constant returns to scale. Educa- ln Q = j ln Mj + k Zk + (14.6) tion is insignificant. A comparison of the first three j=1 k=1 columns of the table reveals that the instrumental where Q is the total value of output produced by variable and selectivity-bias corrected estimates are the household; M is a vector of production inputs, very similar to the OLS estimates. which includes family labor hours, the value of The first column of table 14.6 reports OLS esti- variable costs (including materials and hired mates of the nonagricultural production function. labor),10 the value of fixed capital, and land used in The elasticity of labor inputs is 0.62, more than agricultural production; and Z is a vector of other triple the labor coefficient for agriculture. Output control variables, including average education of value is less responsive to variable costs in nonagri- household workers, average age of household cultural production than in agricultural produc- workers, and provincial dummy variables. tion. The coefficient of equipment is the same as in Eachproductionfunctionisestimatedusing three agriculture, 0.033. Education is statistically insignif- estimation methods--OLS, IV, and selectivity-bias icant. Interestingly, the nonagricultural production Earnings and Labor Mobility in Rural China: Implications for China's Accession to the WTO 251 TABLE 14.5 Agricultural Production Function (dependent variable: log of output value) Model III: Corrected for Model I: OLS Model II: IV Sample Selection Bias Standard Standard Standard Coefficient Error Coefficient Error Coefficient Error Log of labor hoursa 0.192*** 0.035 0.260 0.197 0.179*** 0.035 Log of variable costsa 0.458*** 0.031 0.422** 0.177 0.449*** 0.031 Log of equipment 0.035*** 0.009 0.034*** 0.011 0.034*** 0.009 Log of land 0.317*** 0.052 0.322*** 0.107 0.314*** 0.052 Mean schooling of working- 0.010 0.012 0.014 0.016 0.009 0.011 age adults Mean age of working-age adults 0.0001 0.004 0.0002 0.004 0.0001 0.004 Dummy: telephone present in 0.095 0.062 0.089 0.064 0.104* 0.062 local areas Heilongjiang 0.092 0.116 0.135 0.176 0.088 0.115 Jiangsu 0.069 0.097 0.026 0.154 0.081 0.097 Shandong 0.163 0.139 0.116 0.198 0.167 0.139 Henan 0.623*** 0.102 0.617*** 0.131 0.614*** 0.104 Hubei 0.023 0.090 0.005 0.108 0.046 0.090 Hunan 0.143 0.156 0.195 0.215 0.170 0.157 Guangxi 0.312*** 0.096 0.310*** 0.099 0.265*** 0.096 Constant 2.647*** 0.350 2.323** 1.146 2.811*** 0.353 Lambda n.a. n.a. 0.091 0.096 Adjusted R-square 0.455 0.453 n.a. Log-likelihood n.a. n.a. 1,393.699 Number of observations 1,066 1,066 1,095 n.a. Not applicable. *p < .01, **p < .05, ***p < .10. aIndicates endogenous variables. Instruments used are family composition variables (number of working-age adults, number of children, and number of elderly) and market prices for vegetables, pork, chicken, and gasoline. The Hausman statistic for the joint endogeneity test is 0.13, and the P-value is 0.937. Here and in later tables Guizhou is the omitted regional dummy variable. Source: Authors' estimates using CHNS data. function displays decreasing returns to scale. This the IV specifications to calculate shadow wages. We finding suggests that rural households may prefer the IV results because, in theory, output, encounter difficulties in expanding the scale of labor hours, and variable costs are jointly and household nonagricultural production (or perhaps endogenously determined, and so in principle IV higher-income households underreport such methods are needed to obtain unbiased results. income). Like in table 14.5, a comparison of the first three columns reveals that the selectivity-bias cor- Estimation of Wages and Shadow Wages rected estimates are similar to the OLS estimates. In the IV regression, the coefficients on labor and vari- We are interested in learning how rural households able costs differ from those in the other specifica- respond to differentials between agricultural and tions,but the Hausman test indicates that the overall nonagricultural wages. Although we observe the the results are not significantly different from those hourly wage rates received by employed workers, of OLS. the same is not true for workers engaged in house- Although the Hausman tests do not provide hold agricultural or nonagricultural production. strong support for preferring the IV results to those We follow the approach of Jacoby (1993) and derive of the OLS, we nevertheless use the results from shadow wages from the production function 252 China and the WTO TABLE 14.6 Nonagricultural Production Function (dependent variable: log of output value) Model III: Corrected for Model I: OLS Model II: IV Sample Selection Bias Standard Standard Standard Coefficient Error Coefficient Error Coefficient Error Log of labor hoursa 0.618*** 0.039 0.486*** 0.183 0.656*** 0.048 Log of variable costsa 0.167*** 0.012 0.253*** 0.069 0.154*** 0.015 Log of equipment 0.033*** 0.010 0.024* 0.014 0.030 0.032 Mean schooling of working-age adults 0.012 0.016 0.007 0.019 0.001 0.027 Mean age of working-age adults 0.005 0.006 0.007 0.006 0.012 0.008 Dummy: telephone present in local areas 0.075 0.098 0.041 0.149 0.100 0.114 Heilongjiang 0.031 0.206 0.125 0.229 0.376 0.350 Jiangsu 0.159 0.144 0.201 0.155 0.123 0.205 Shandong 0.271 0.188 0.354* 0.209 0.181 0.265 Henan 0.008 0.125 0.016 0.133 0.126 0.163 Hubei 0.028 0.143 0.092 0.159 0.143 0.214 Hunan 0.066 0.135 0.059 0.142 0.223 0.241 Guangxi 0.137 0.114 0.128 0.120 0.432*** 0.169 Constant 2.968*** 0.362 3.670*** 1.136 2.985*** 0.661 Lambda n.a. n.a. 0.450 0.239 Adjusted R-square 0.603 0.561 n.a. Log-likelihood n.a. n.a. 946.014 Number of observations 507 507 1,611 n.a. Not applicable. ***p < .01, **p < .05, *p < .10. aIndicates endogenous variables. Instruments used are family composition variables (number of working-age adults, number of children, and number of elderly); market prices of honey-combed coal briquette, coal lumps, coal powder, and liquefied natural gas; and local market wage rate for the unskilled. The Hausman statistic for the joint endogeneity test is 1.77, and the P-value is 0.413. Source: Authors' estimates using CHNS data. estimates. Specifically, we calculate the shadow labor allocation. After a selection is made, only wage rates, or marginal products, of family labor wages in the actually selected sectors are observed. hours from the IV estimates of the Cobb-Douglas To estimate the expected wage differential, we need production functions in tables 14.5 and 14.6 as the counterfactual wage rates for households not follows: participating in a sector. To obtain these rates, we regress the shadow wages computed as just out- Wj = ^ ^j Qj , lined against a range of household characteristics Lj (14.7) that potentially affect the productivity of workers. j = agriculture, nonagriculture These include all exogenous variables in the pro- where Q is the value of output and L is family labor duction functions, the variables used as instru- hours. ments for labor, and variable costs. We then use the Equation 14.7 allows us to calculate average results from these regressions to predict shadow shadow wages for households that engage in agri- wages for all households. cultural or nonagricultural production. These Note that the projected shadow wages are used shadow wages are derived from household produc- later in our labor supply functions, and so some of tion functions and thus are at the household, rather the explanatory variables here will serve as instru- than individual, level. ment variables in our labor supply regressions. Households compare expected wage differen- Therefore, because all second-stage variables tials among sectors when deciding their sectoral should be included in the first-stage regressions, we Earnings and Labor Mobility in Rural China: Implications for China's Accession to the WTO 253 also include all exogenous variables from the labor variables mentioned above, it is difficult to find supply equations in the shadow wage equations. additional identification variables for use in an IV These variables include nonlabor income, number or 2SLS regression. The agricultural shadow wage of dependents, education and age of workers, function has good explanatory power, and the health status of household members, and a few results are reasonable. As expected, equipment and community-level variables that measure the preva- land are significant in enhancing the marginal pro- lence of nonfarm activities. ductivity of agricultural labor. Average schooling of Table 14.7 reports estimates of the shadow wage working adults is also positive and significant. equations for household agricultural production The explanatory power of the nonagricultural and household nonagricultural production. We use shadow wage function is low, perhaps reflecting the the OLS regression because, after including all the smaller number of observations and heterogeneity TABLE 14.7 Shadow Wage Equations for Household Agricultural and Nonagricultural Production (dependent variable: log of shadow wage) Household Agricultural Household Nonagricultural Production (OLS) Production (OLS) Coefficient Standard Error Coefficient Standard Error Agricultural equipment (1,000 yuan) 0.020*** 0.006 0.037 0.025 Nonagricultural equipment (1,000 yuan) 0.014*** 0.005 0.018*** 0.005 Land 0.012*** 0.003 0.016 0.012 Number of family working-age adults 0.017 0.033 0.056 0.058 Number of dependents 0.045 0.036 0.104 0.064 Ratio of male working-age adults to 0.281 0.207 0.656 0.399 family working-age adults Family nonlabor income (1,000 yuan) 0.002 0.019 0.052 0.033 Mean schooling of working-age adults 0.036** 0.015 0.008 0.031 Mean age of working-age adults 0.004 0.005 0.013 0.011 Dummy: all family members are healthy 0.220*** 0.071 0.191 0.134 Dummy: telephone present in local 0.008 0.084 0.330** 0.148 areas Market price of kerosene 1.093*** 0.229 0.732* 0.402 Market price of rice most commonly 0.049 0.030 0.052 0.132 used Dummy: TVEs present in local areas 0.138* 0.076 0.084 0.136 Number of self-employed household 0.003 0.002 0.003 0.003 enterprises in local areas Heilongjiang 0.604*** 0.153 0.058 0.443 Jiangsu 0.063 0.153 0.266 0.296 Shandong 0.057 0.236 0.347 0.406 Henan 0.653*** 0.132 0.186 0.217 Hubei 0.0002 0.114 0.120 0.247 Hunan 0.949*** 0.275 1.035*** 0.360 Guangxi 0.458*** 0.135 0.265 0.211 Constant 0.169 0.509 0.206 0.990 Adjusted R-square 0.274 0.076 Number of observations 998 274 ***p < .01, **p < .05, *p < .10. Source: Authors' estimates using CHNS data. 254 China and the WTO of labor inputs in nonagricultural activities. Here ture, but adds some variables relevant to the China the most significant variable is nonagricultural case. The dependent variable is the log of hourly equipment. wages. Independent variables include variables cap- Not all households or individuals participate in turing individual human capital or labor quality wage employment. We therefore also need to pre- (education, age, age squared, and health status), dict market wages for nonparticipants. In addition, individual characteristics (marital status and sex), market wages are reported at the individual level, and regional dummy variables. We allow for possi- but to analyze household labor supply we need a ble sample selection bias by applying the Heckman measure of the expected wage at the household method, but the results show that selectivity is not level. To address these issues, we first estimate a important.11 wage equation for all workers participating in wage The employed wage function indicates that edu- employment.We then use these estimates to predict cation receives a rate of return of 3 percent. This wage rates for all workers in the sample, after which effect is statistically significant at the 1 percent level. we aggregate for each household the predicted indi- The age-earning profile has a concave shape, with vidual wage rates of family members to obtain a the maximum wage reached at age 48. Marriage has household-level wage. a positive but statistically insignificant effect on Table 14.8 gives the estimates of wage functions wage.Once given marital status,female workers earn for the employed workers. Our specification fol- 20.3 percent less than their male counterparts.Those lows the standard approach used in the labor litera- who have urban hukou status earn 11.0 percent more TABLE 14.8 Employed Wage Equation (dependent variable: log of hourly wages) Model II: Selectivity-Bias Model I: OLS Correcteda Independent Variable Coefficient Standard Error Coefficient Standard Error Female 0.203*** 0.029 0.203*** 0.033 Marriage status 0.012 0.046 0.012 0.058 Education level (years) 0.030*** 0.005 0.030*** 0.009 Age 0.029*** 0.011 0.029** 0.012 Age squared 0.0003** 0.0001 0.0003** 0.0001 Health status 0.066* 0.040 0.066* 0.039 Urban hukou 0.110*** 0.029 0.110*** 0.029 Heilongjiang 0.084 0.065 0.084 0.064 Jiangsu 0.181*** 0.058 0.181*** 0.069 Shandong 0.057 0.061 0.057 0.070 Henan 0.236*** 0.072 0.236*** 0.073 Hubei 0.043 0.064 0.043 0.063 Hunan 0.229*** 0.063 0.229*** 0.065 Guangxi 0.002 0.063 0.002 0.062 Constant 0.018 0.189 0.018 n.a. Lambda n.a. 0.00005 0.084 Adjusted R-square 0.116 n.a. Log-likelihood n.a. 5,493.72 Number of observations 1,910 8,123 n.a. Not applicable. ***p < .01, **p < .05, *p < .10. aThe identification variable is the number of dependents. Source: Authors' estimates using CHNS data. Earnings and Labor Mobility in Rural China: Implications for China's Accession to the WTO 255 TABLE 14.9 Summary Statistics for Wage Rates and Shadow Wages (yuan/hour) Number of Standard Variable Observations Mean Deviation Shadow wage for agricultural production (WA) 1,066 0.742 1.322 Shadow wage for agricultural production, predicted (WA) ^ 1,500 0.517 2.197 Shadow wage for nonagricultural production (WN) 507 2.383 3.243 Shadow wage for nonagricultural production, predicted (WN) ^ 1,500 1.760 6.871 Employed wage rate (individual level) (WM) 1,910 4.013 37.789 Employed wage rate, predicted (individual level) (WMi) ^ 8,270 2.239 0.554 Employed wage rate, predicted (household level) (WM) ^ 2,755 2.271 0.526 Source: Authors' estimates using CHNS data. than those who do not. Using the coefficients from we also run Tobit regressions and report the mar- the selectivity-bias corrected regression, we predict ginal effects for all households and the standard wages for all workers. To obtain a household-level errors of the marginal effects. wage, we calculate the weighted average of pre- Our results appear in table 14.10. The first dicted wages for the workers in each household, columns give estimated coefficients for total labor where the weights are based on each worker's hours supply per adult. These results indicate that total of work. labor supply is not sensitive to the marginal returns Table 14.9 shows the mean predicted shadow to nonfarm labor. The returns to agricultural labor wages and market wages of all households and, for have a coefficient that is negative, significant (at the comparison, the computed shadow wages and 10 percent level), and small. This coefficient reflects observed market wages of participating house- the net impact of the agricultural shadow wage on holds. Not surprisingly, the mean wages of partici- the different components of the labor supply. This pating households are higher than those predicted impact is positive for the agricultural labor supply for all households. This difference reflects that par- but negative for the market labor supply (this is dis- ticipation is more likely at higher wage rates. The cussed later in this chapter). The effect of nonlabor returns to labor are lowest for agricultural produc- income is negative and significant at the 10 percent tion, followed by nonagricultural production and level, indicating that leisure is a normal good. Inter- wage employment. These numbers are broadly estingly, total labor supply per adult is higher the consistent with the estimated returns to work time more dependents (elderly and children) in a house- from the income generation regressions reported hold. This finding suggests that the need to earn earlier. income to support dependents outweighs the need to spend non-earning time caring for dependents. The effect of dependents also indicates that depend- Estimation of Labor Supply Functions ents cause households to reallocate labor away from Our household labor supply functions follow the higher-paying wage jobs to lower-income house- form of equations 14.2­14.5. For the dependent hold production activities. variable, we use hours per adult rather than total Although the nonagricultural wage variables do hours to avoid possible correlation between house- not have a significant effect on total labor supply, hold labor supply and number of working-age they affect the allocation of labor among different adults in the household. All specifications are esti- activities. The shadow wage for work in household mated using IV (2SLS), with the wage and shadow agricultural production is significant for market wage equations reported earlier serving as the first- employment, where it has a fairly large, negative stage regressions. A large number of households coefficient. A 1 percent increase in the shadow agri- supply zero hours to nonagricultural production cultural wage decreases labor supply to market and wage employment, and so for these functions employment by 2.6 percent. 256 China and the WTO Error 0.485 0.296 0.414 0.021 0.077 0.040 0.527 0.011 0.171 0.204 0.004 0.627 0.063 0.903 0.529 0.696 0.247 0.253 0.792 0.244 n.a. Standard Hours Unconditional obitT Effect 2.137*** 1.183*** 0.485 0.111*** 0.310*** 0.249*** 2.078*** 0.026** 0.249 1.279*** 0.011*** 2.190*** 0.208*** 2.051** 1.916*** 1.767** 0.907*** 0.001 0.521 0.621** n.a. n.a. 0.073 1,421 Marginal Employment age W Error 0.335 0.214 0.448 0.022 0.077 0.040 0.494 0.011 0.177 0.171 0.004 0.595 0.061 0.440 0.335 0.437 0.336 0.252 0.621 0.341 1.048 Standard of IV Log n.a. 1.137*** 0.617*** 0.555 0.118*** 0.318*** 0.236*** 1.598*** 0.018 0.180 1.054*** 0.010** 1.376** 0.158*** 0.651 1.684*** 1.770*** 0.653* 0.111 0.220 0.542 1.466 0.175 1,421 Coefficient Error 0.376 0.247 0.476 0.018 0.061 0.035 0.455 0.010 0.145 0.151 0.003 0.501 0.070 0.176 0.199 0.176 0.369 0.145 0.322 0.410 n.a. Standard adult) Unconditional per Hours obitT n.a. n.a. Effect 0.008 0.389 0.791* 0.047*** 0.141** 0.077** 0.316 0.018* 0.075 0.226 0.0003 0.272 0.139** 0.944*** 0.429** 0.596*** 0.304 0.529*** 0.362 0.893** 0.058 1,430 Marginal hours Nonagricultural of Error 0.304 0.194 0.400 0.019 0.070 0.036 0.445 0.010 0.160 0.154 0.004 0.537 0.056 0.400 0.302 0.389 0.304 0.228 0.554 0.307 0.942 Log Standard Self-Employment working IV of n.a. 0.141 0.332* 0.577 0.052*** 0.199*** 0.088** 0.292 0.013 0.145 0.193 0.001 0.018 0.095* 1.204*** 0.387 0.761* 0.475 0.607* 0.385 1.234*** 0.931 0.105 1,430 log Coefficient Error 0.098 0.062 0.129 0.006 0.023 0.012 0.144 0.003 0.052 0.050 0.001 0.174 0.018 0.129 0.098 0.130 0.098 0.074 0.183 0.099 0.307 variable: Standard IV Agricultural Hours of n.a. Log 0.025 0.123** 0.120 0.032*** 0.140*** 0.049*** 0.261* 0.007** 0.101* 0.206*** 0.006*** 0.021 0.001 0.150 0.258*** 0.443*** 0.312*** 0.344*** 0.191 0.414*** 6.881*** 0.198 1,404 Coefficient (dependent Hours Error 0.084 0.054 0.113 0.005 0.020 0.010 0.125 0.003 0.045 0.044 0.001 0.151 0.015 0.111 0.085 0.113 0.085 0.064 0.161 0.086 0.268 Standard otalT IV Equations of Log 0.144* 0.051 0.067 0.009* 0.117*** 0.009 0.128 0.005* 0.012 0.005 0.001 0.030 0.031** 0.054 0.137 0.220* 0.172** 0.216*** 0.027 0.329*** 6.525*** 0.098 n.a. 1,360 Supply Coefficient data. to Labor adults adults CHNS areas .10. adults < family local household rice p* using adults for in areas working-age working-age .05, Household of < estimates income) health local kerosene common working-age present p family in of of ** dependents of self-employed of male working-age good TVEs of R-square price price R-square applicable. Authors' 14.10 schooling of age vations .01, ^W A^W N^W M (nonlabor Not < ce: family enterprises p ln ln ln Log Number Mean Ratio Mean Dummy: Dummy: Number Market Market Heilongjiang Jiangsu Shandong Henan Hubei Hunan Guangxi Constant Adjusted Pseudo Obser n.a. *** Sour ABLET Earnings and Labor Mobility in Rural China: Implications for China's Accession to the WTO 257 The effect of the shadow wage for work in Using these expressions and the results in household nonagricultural production is negative table 14.10, we calculate the marginal response of and significant for agricultural labor and positive log(HN/HA) to log(WN/WA) as 0.512 (5 0.389 1 ^ ^ and mostly significant for nonagricultural and 0.123), and the marginal response of log(HM/HA) market labor supply. In other words, households to log(WM/WA) as 0.605 (5 0.485 1 0.120). These ^ with higher returns to labor in nonagricultural numbers imply that a 1 percent increase in the ratio sidelines work fewer hours in agriculture and more of either nonagricultural wage to the agricultural hours in both nonagricultural and wage employ- wage raises the ratio of that nonagricultural work ment. Here the elasticities are small for agriculture to agricultural work by about 0.5­0.6 percent. and nonagriculture, and close to one for wage Note that these numbers assume that the change employment. The signs on the coefficient for the in the wage ratio stems from a rise in the nonagri- market wage are the same as those for the shadow cultural shadow or market wage, with the agricul- nonagricultural wage, but this variable is signifi- tural shadow wage remaining constant. Due to the cant in only one equation. presence of cross-wage effects, shifts caused by Overall, the coefficients on the wage variables changes in the agricultural shadow wage (holding reveal complementarity between employment in the nonagricultural and market wages constant) household nonagricultural production and off-farm would be different. Here the relevant calculations wage jobs, but substitution between agriculture and are as follows: the marginal response of both types of nonagricultural employment.Further- log(HA/HN) to log(WA/WN) is 0.017 ( ^ ^ 0.025 more, the estimated coefficients on wage variables 0.008), and the marginal response of log(HA/HM) are larger in magnitude and more significant in the to log(WA/WM) is 2.165 ( ^ 0.028 2.137). In off-farm wage labor supply equations than in the other words, a 1 percent increase in the ratio of the householdagriculturalornonagriculturallaborsup- agricultural shadow wage to the nonagricultural ply equations.In other words,labor supply to house- shadow wage would increase the ratio of labor in hold production activities does not respond much to agricultural to nonagricultural production by wage levels; the supply of labor to off-farm wage 0.02 percent. For an increase in the ratio of the employment does. agricultural shadow wage to the market wage, the From the wage elasticities we can calculate response would be much larger, 2.2 percent. the elasticity of labor transfers in response to These numbers indicate that, for the most part, relative wage differentials among sectors.12 Sub- the effects of relative wages on sectoral labor alloca- tracting equation 14.2 from 14.3 gives the following tion are not overly large or significant (relative to expression: the standard errors). The one exception is the case log(HN/HA) in which a change in the agricultural shadow wage = (^0 - ^0) + (^A - ^A)ln WA ^ causes a change in the ratio of the agricultural shadow wage to the market wage. In this case, shifts + (^N - ^N)ln WN ^ in labor allocation between agriculture and off-farm + (^M - ^M)ln WM + ··· ^ employment may be relatively large. = (^0 - ^0) (14.8) These findings suggest that labor movement out of agriculture into other sectors is not driven by the + (^A - ^A + ^N - ^N)ln WA ^ "pull"of rising nonagricultural wages, but would be + (^N - ^N)ln(WN/WA) ^ ^ driven by the "push" of lower agricultural wages. + (^M - ^M)ln WM + ··· ^ Why would these pull-and-push effects differ? One explanation is that for rural families income from Similarly, subtracting (14.2) from (14.4) gives agriculture provides the base income that ensures log(HM/HA) that the household achieves some acceptable or = (^0 - ^0) minimum standard of living. If agricultural income + (^A - ^A + ^M - ^M)ln WA ^ falls short--for example, in the event of drought, (14.9) + (^N - ^N)ln WN ^ which would reduce overall agricultural income as well as the returns to labor in agriculture-- + (^M - ^M)ln(WM/WA) + ··· ^ households may feel compelled to seek income 258 China and the WTO from other sources. They do not typically seek that Community-level variables also affect the com- additional income from nonagricultural self- position of the labor supply. We use a dummy employment, because such activities usually require variable for the availability of TVEs and the num- start-up time and some initial investment. Casual ber of self-employed household businesses as indi- wage jobs entail lower initial costs and generate cash cators of the prevalence of nonfarm activities in income more quickly. Field interviews with rural local communities. As expected, both variables are households that have experienced shortfalls in agri- significant in reducing labor allocation to agricul- cultural income indeed indicate that these house- tural production and raising participation in holds often send family members out to find casual nonagricultural production and wage employ- wage jobs, sometimes seasonal, in construction, ment. Also, the coefficients on the market prices services, and other unskilled occupations. for fuel and rice are in some cases negative and Several nonwage variables have significant and significant. interesting effects on labor allocation. Nonlabor In general, our estimates indicate that labor sup- income is significant and negative for labor supply ply to agriculture and labor supply to wage employ- to both agricultural and nonagricultural household ment have opposite responses to most explanatory production, but is significant and positive for wage variables. In other words, factors that increase wage employment. This result could indicate that higher employment would tend to decrease agricultural income increases access to wage employment by, employment, and vice versa. This conclusion is for example, providing the resources needed to consistent with aggregate statistics that show such cover initial and search costs associated with opposing trends in aggregate employment in the obtaining wage jobs. It could also indicate that two sectors. Labor supply to nonagricultural house- households gain some status or utility from off- hold businesses, however, moves with agriculture farm work. The magnitudes of these income effects, for some variables and with wage employment for however, are fairly small. Not surprisingly, the others. Thus, depending on which determinants of number of dependents significantly increases labor labor supply are driving aggregate trends, we might supply to both types of household production, but expect to observe concurrent growth in agricultural reduces labor supply to off-farm jobs. One more and nonagricultural household employment, or dependent in the household increases working growth in one and decline in the other. hours that an adult devotes to agricultural produc- tion by 14 percent, increases hours devoted to Conclusions nonagricultural production by 14 percent, and reduces working hours in wage employment by The impact of trade liberalization in China depends 31 percent. Age has a marginally significant and not only on the aggregate sectoral and price shifts positive effect on agricultural labor supply; for that follow liberalization, but also on the microeco- nonagricultural production and market employ- nomic response of households and individuals to ment, its effects are negative, and for the Tobit those shifts. In this chapter we have analyzed the regressions, its effects are significant. This result is microeconomicsof householdearningsandemploy- consistent with the observation that in rural China ment in rural China. Using household survey data younger adults are more mobile than older adults. with broad regional coverage, we have examined Interestingly, a few household characteristics household incomes, wage determination, and labor that do not significantly influence total labor supply supply in total and among sectors. have significant effects on the allocation of labor Our analysis of income generation provides among activities. Education reduces hours worked information about earnings from labor. We find in agricultural production by 5 percent, increases that the returns to nonagricultural labor hours are hours in nonagricultural production by 8 per- higher than for agricultural labor hours, a result cent, and increases hours in wage employment by a that is consistent with other studies in the litera- substantial 25 percent. The proportion of males ture. We then decompose the income gap between among working-age adults increases labor supply poorer and richer groups. Key contributors to the to agriculture, but reduces labor supply to wage income gap are the lower returns received by poor employment. households from education, agricultural labor, and Earnings and Labor Mobility in Rural China: Implications for China's Accession to the WTO 259 wage employment. Also important is the fact that Notes poorer households work fewer hours in higher- 1. Official Chinese estimates show the number of rural poor paid wage employment. These results suggest that declining from 65 million in 1995 to 42 million in 1998 (World analysis of the impact of trade liberalization should Bank 2001), and show the number of urban poor ranging from 10 to 15 million in the mid-1990s (UNDP 1998; NBS 1998). consider not only the fact that the level and compo- These numbers may not be entirely comparable, but they indi- sition of employment differ among income groups, cate that roughly 75­85 percent of the poor are rural. but also that the returns to employment in each 2. Official Chinese statistics distinguish between private and occupation can differ among groups. individual enterprises. Private enterprises are larger and rely on hired employees as the main source of labor. Individual enter- Our analysis of household labor supply provides prises are smaller and, although they may hire employees, rely some insights into how changes in wages after largely on family workers for labor. Both types of enterprises are China's accession to the WTO might affect patterns privately owned. 3. Income data in the text are deflated using the rural con- of employment and thus earnings. Total labor sup- sumer price index and are taken from the National Bureau of ply is not sensitive to changes in wages, which sug- Statistics (various years) and the National Bureau of Statistics gests that any changes in wages resulting from Rural Social and Economic Survey Team (2000). 4. The CHNS survey covers the provinces of Guangxi, WTO accession would not have much impact Guizhou, Heilongjiang, Henan, Hubei, Hunan, Jiangsu, on overall employment. We also find that, for the Liaoning, and Shandong. The 1997 survey, which we use for our most part, the allocation of labor among self- analysis, does not include Liaoning and so covers only eight of employment in agriculture, self-employment in the nine provinces. 5. Detailed information about the CHNS is available at nonagriculture, and off-farm wage jobs is not www.cpc.unc.edu/china/home.html. overly sensitive to changes in wages. In particular, 6. Note that for agricultural and nonagricultural household labor supply to both agricultural and nonagricul- production, these estimated returns are substantially lower than average earnings per hour worked given earlier (1.2 and tural household production is inelastic with respect 2.2 yuan, respectively). This finding could reflect the fact that in to most wages. Put differently, changes in wages do household production the average and marginal returns to labor not cause large changes in the amount of labor differ and that the income generation function controls for non- labor inputs. The estimated return to wage labor hours is also supplied to household production. somewhat lower than average earnings (2.8 yuan). The off-farm labor supply is more responsive, 7. Note that the difference in time worked per adult is similar at least to the agricultural shadow wage. A lower to that for total hours, because the number of working-age agricultural shadow wage is associated with sub- adults in the two groups is similar--on average, the poor house- holds had 2.74 adults and the rich households 2.71 adults. stantially increased employment in off-farm jobs. 8. Note that wage employment includes wage labor in agri- Thus if WTO accession raises wages for off-farm culture, but in our sample hired farm labor is a small proportion jobs, it would not "pull" labor into wage jobs; of total wage employment. 9. Other studies taking this approach include Skoufias however, if WTO accession reduces the returns to (1994) and Abdulai and Regmi (2000). agricultural labor, it might "push" labor into wage 10. The CHNS data do not distinguish hired labor from jobs. other variable inputs. 11. Estimated coefficients for Models I and II are nearly Although in general labor supply is not overly identical despite the significance of the number of dependents sensitive to wages, it is sensitive to other variables. in sector choice. Note that lambda is also highly insignificant. Household characteristics such as education, This indicates that there is no selectivity bias. household composition, and regional location have 12. We thank Martin Ravallion for suggesting this approach. significant effects on the level and sectoral composi- tion of labor supply. Higher education is associated References with less work time in agriculture and more work The word processed describes informally reproduced works that time in nonagricultural types of work. More may not be commonly available through libraries. dependents are associated with more work time Abdulai, Awudu, and Punya Prasad Regmi. 2000. "Estimating overall,as well as with more work time in household Labor Supply of Farm Households under Nonseparability: Empirical Evidence from Nepal." Agricultural Economics 22: production and less in off-farm wage employment. 309­20. These findings suggest that labor mobility among Bowlus, Audra, and Terry Sicular. 2003. "Moving toward sectors is significantly influenced by variables such Markets? Labor Allocation in Rural China." Journal of as education, health, and demographic structure. Development Economics 71 (2): 561­83. Chen Shaohua and Wang Yan. 2001. "China's Growth and Such variables are not directly affected by WTO Poverty Reduction: Recent Trends between 1990 and 1999." accession, at least in the short term. World Bank, July. Processed. 260 China and the WTO Dollar, David, and Aart Kraay. 2001. "Trade, Growth and 2000 [China Rural Household Survey Yearbook 2000]. Poverty." Policy Research Working Paper 2615. World Bank, Beijing: China Statistics Press. Washington, D.C. Parish, W. L., X. Zhe, and F. Li. 1995. "Nonfarm Work and Hare, Denise. 1994. "Rural Nonagricultural Activities and Their Marketization of the Chinese Countryside." China Quarterly Impact on the Distribution of Income: Evidence from Farm 143: 697­730. Households in Southern China." China Economic Review Rawski, T. G., and R. W. Mead. 1998. "On the Trail of China's 4 (1): 59­82. Phantom Farmers." World Development 26 (5): 767­81. ------. 1999a. " `Push' versus `Pull' Factors in Migration Out- Rodrik, Dani. 2000."Comments on `Trade, Growth, and Poverty,' flows and Returns: Determinants of Migration Status and by David Dollar and Aart Kraay."Available at http://ksghome. Spell Duration among China's Rural Population." Journal of harvard.edu/~.drodrik.academic.ksg/papers.html. Development Studies 35 (3): 45­72. Rozelle, Scott, Guo Li, Minggao Shen, Amelia Hughart, and John ------. 1999b. "Women's Economic Status in Rural China: Giles. 1999. "Leaving China's Farms: Survey Results of New Household Contributions to Male-Female Disparities in the Paths and Remaining Hurdles to Rural Migration." China Wage Labor Market." World Development 27 (6): 1011­29. Quarterly 158 (June): 367­93. Heckman, James. 1979."Sample Selection Bias as a Specification Skoufias, Emmanuel. 1994. "Using Shadow Wages to Estimate Error." Econometrica 47 (1): 153­62. Labor Supply of Agricultural Households." American Journal Jacoby, Hanan. 1993. "Shadow Wages and Peasant Family of Agricultural Economics 76 (May): 215­27. Labour Supply: An Econometric Application to the Peruvian UNDP (United Nations Development Programme). 1998. China Sierra." Review of Economic Studies 60 (4): 903­21. Human Development Report: Human Development and Knight, John, and Lina Song. 1997. "Chinese Peasant Choices: Poverty Alleviation, 1997. Beijing: UNDP. Migration, Rural Industry or Farming." Oxford Applied Winters, L. Alan. 2000. "Trade, Trade Policy, and Poverty: What Economics Discussion Paper No. 188. October. are the Links?" Centre for Economic Policy Research Paper ------. 1999. The Rural­Urban Divide: Economic Disparities No. 2382. Centre for Economic Policy Research, London. and Interactions in China. Studies on Contemporary China. World Bank. 2001. China: Overcoming Rural Poverty. Oxford and New York: Oxford University Press. Washington, D.C. Li Tianyou and Junsen Zhang. 1998. "Returns to Education Wu, Harry X., and LI Zhou. 1996 "Research on Rural to Urban under Collective and Household Farming in China." Journal Labour Migration in the Post-Reform China: A Survey." of Development Economics 56 (2): 307­35. Chinese Economy Research Unit Working Paper 96/4. Meng Xin. 2000. Labour Market Reform in China. Cambridge: University of Adelaide, February. Cambridge University Press. Yang, Dennis Tao. 1997. "Education and Off­Farm Work." Michelson, Ethan, and William Parish. 2000. "Gender Differen- Economic Development and Cultural Change 45 (3): 613­32. tials in Economic Success: Rural China in 1991." In Barbara Yang, Dennis Tao, and Hao Zhou. 1999."Rural-Urban Disparity Entwisle and Gail E. Henderson, eds., Redrawing Boundaries: and Sectoral Labour Allocation in China." Journal of Work, Households and Gender in China. Berkeley: University Development Studies 35 (3): 105­33. of California Press. Yao Yang. 1999. "Rural Industry and Labor Market Integration National Bureau of Statistics. 1998. Zhongguo Shehui yu Keji in Eastern China." Journal of Development Economics 59 (2): Fazhan Baogao [China Social and Science and Technology 463­96. Development Report]. Beijing: China Statistics Press. Zhao Yaohui. 1999a. "Labor Migration and Earnings Differ- ------. 2001. China Statistical Abstract 2001. Beijing: China ences: The Case of Rural China." Economic Development and Statistics Press. Cultural Change 47 (4): 767­82. ------. Various years. China Statistical Yearbook. Beijing: China ------. 1999b. "Leaving the Countryside: Rural-to-Urban Statistics Press. Migration Decisions in China." American Economic Review National Bureau of Statistics Rural Social and Economic Survey 89 (2): 281­86. Team. 2000. Zhongguo Nongcun Zhuhu Diaocha Nianjian 15 Welfare Impacts of China's Accession to the WTO Shaohua Chen and Martin Ravallion There has been much debate about the welfare within countries can add considerable noise to impacts of greater trade openness. Some argue that the measured levels and changes in poverty and external trade liberalization is beneficial to the poor inequality. It is unclear how much power cross- in developing countries, while others argue that the country data sets have for detecting any underlying benefits will be captured by people who are not effects of greater openness or other covariates. poor. Expected impacts on relative wages (notably There is also the issue of whether trade volume can between skilled and unskilled labor) and relative be treated as exogenous in these cross-country prices (such as between food staples and luxury im- regressions; it is clearly not a policy variable and ports) have figured prominently in debates about may well be highly correlated with other (latent) the welfare impacts. attributes of country performance independently of What does the evidence suggest? One might trade policy. The attribution of inequality impacts hope to provide a conclusive answer by comparing to trade policy reforms themselves is clearly changes over time in measures of inequality or problematic. The correlations (or their absence) poverty between countries that are open to external found in cross-country studies can also be decep- trade and countries that are not. A number of tive because starting conditions can vary so much attempts to throw empirical light on the welfare between reforming countries. Averaging across this effects of trade liberalization have been made using diversity in initial conditions can readily hide sys- aggregate cross-country data sets that combine tematic effects of relevance to policy (Ravallion survey-based measures of inequality or poverty 2001). with data on trade openness and other control vari- In principle, such problems in cross-country ables (see Bourguignon and Morisson 1990; comparative work can be dealt with by better data Edwards 1997; Barro 2000; Dollar and Kraay 2002; and methods. However, the concerns go deeper. Lundberg and Squire 2003). Aggregate inequality or poverty may not change However, caution in drawing implications from with trade reform even though there are both gain- such cross-country comparisons is appropriate. ers and losers at all levels of living. Survey data Concerns have been raised about data and econo- tracking the same families over time commonly metric specifications. Differences in survey design show considerable churning under the surface.1 and processing between countries and over time The data show that many people have escaped 261 262 China and the WTO poverty while others have fallen into poverty, even economywide impacts on markets. However, stan- though the overall poverty rate is unchanged. dard CGE models entail considerable aggregation Numerous sources of such diverse impacts can across household types, with rarely more than six be found in developing country settings. For exam- or so "representative households." Such models are ple, geographic disparities in access to human and crude tools for welfare-distributional analysis. physical infrastructure affect prospects for partici- The challenge for applied work is to find an pating in economic growth.2 For China the eco- approach that respects the richness of detail avail- nomic geography of poverty and how this interacts able from modern integrated household surveys with geographic diversity in the impacts of policy while ensuring that the price changes attributed to reforms are high on the domestic policy agenda. A reform are internally consistent with economywide policy analysis that simply averaged over such dif- equilibrium conditions. In principle, the CGE ferences would miss a great deal of what matters to model could be built onto the household survey, so the debate on policy. that the number of households in the model is the This chapter follows a different approach in number sampled in the survey.4 For this study, that which the attribution to trade policy changes is degree of integration would require an extraordi- unambiguous and the diversity of welfare impacts narily high dimensional CGE model, with 85,000 is not lost. The article examines the welfare impacts households. This is currently not a feasible route. at the household level of the changes in commodity The intermediate approach used here carries and factor prices attributed to a specific trade the reform-induced commodity and factor price policy reform, namely China's accession in 2001 to changes simulated from a general equilibrium the World Trade Organization (WTO). For China, model to the level of all the sampled households in this meant a sharp reduction in tariffs, quantitative the survey.5 The welfare impacts are measured restrictions, and export subsidies, with implica- using standard tools of analysis familiar from past tions for the domestic structure of prices and wages work on the welfare effects of price changes associ- and thus for household welfare. Drawing on esti- ated with tax and trade policy reform. This ap- mates by Ianchovichina and Martin (2002) of the proach imposes minimal aggregation conditions impacts of reform on prices (for both commodities on the survey data, within unavoidable data limita- and factors of production), the following analysis tions. In addition to calculating the trade reform's applies standard methods of first-order welfare overall effects on poverty and inequality, this analysis to measure the gains and losses at the approach provides a detailed socioeconomic map household level using large sample surveys col- of impact, showing how it varies with other nonin- lected by China's National Bureau of Statistics. come characteristics, such as location. This gener- ates better insights to the questions policymakers ask about who gains and who loses from reform. Measuring the Welfare Impacts of The general equilibrium analysis generates a set Trade Reform of price and wage changes. These embody both the Past approaches to studying the welfare impacts direct price effects of the trade policy change and of specific trade reforms have tended to be either second-round, indirect effects on the prices of non- partial equilibrium analyses, which measure traded goods and on factor returns, including ef- household-level welfare impacts of the direct price fects operating through the government's budget changes due to tariff changes using survey data constraint. Ianchovichina and Martin (2002) use a (typically) covering many thousands of randomly competitive market-clearing model from the Global chosen households, or general equilibrium analy- Trade Analysis Project.6 The revenue implications ses, which use a computable general equilibrium of the trade policy change are reflected in changes in (CGE) model to capture second-round responses.3 indirect tax rates.7 Since the price changes are based While partial equilibrium analysis requires little or on an explicit model, their attribution to the trade no aggregation of the primary household data, it policy reform is unambiguous, thus avoiding the misses potentially important indirect effects on identification problems common to past attempts prices and wages. General equilibrium analysis has to estimate distributional effects of trade policy the power to capture these effects by simulating reform using cross-country comparisons. Welfare Impacts of China's Accession to the WTO 263 The approach can be outlined as follows. Each monetary value of the change in utility for house- household has preferences over consumption and hold i is given by: work effort (under the standard assumption that s d goods have positive marginal utilities while labor m gi dui = dpij dpij s s d d supply has negative marginal utility) represented by vi pijqij s - pij qij + zij j=1 pij pij d the utility function ui(qi , Li), where qi is an m- d d n dimension vector of the quantities of commodities + wkLsik dwk (15.3) consumed by household i and Li is a vector of labor k=1 wk supplies by activity, including supply to the house- hold's own production activities. The household is where vi is the marginal utility of income for assumed to be free to choose qi and Li subject to its d household i (the multiplier on the budget con- budget constraint. Consistently with the general straint in equation 1) and Lik = Lik - Lik is the s o equilibrium model that generated the price and household's "external" labor supply to activity k. wage changes,there is no rationing at the household (Notice that gains in earnings from labor used in level; for example, involuntary unemployment is own-production are exactly matched by the higher ruled out. cost of this input to own-production.) The indirect utility function of household i is Equation 15.3 is the key formula used in calcu- given by: lating the household-level welfare impacts of the price changes implied by the general equilibrium vi pi , wi, i = max ui qi , Li d d pi qi d d analysis of the trade policy reform. The propor- (qi ,Li) d = wiLi + i tionate changes in all prices and wages are weighted (15.1) by their corresponding expenditure and income where pi is the price vector (of dimension m) for d shares. The weight for the proportionate change in consumption, wi is the vector of wage rates, and i the j'th selling price is pijqij, the revenue (selling s s is the profit obtained from all household enter- value) from household production activities in prises as given by: sector j. Similarly, -pij(qij + zij) is the (negative) d d i pi , pi , wi = maxo s d pi qi - pi zi - wi Li qij s s d o s weight for demand price changes, and wkLik is the s (zi,Li ) weight for changes in the wage rate for activity k. fij zij, Loij , j = 1, ..., m; The term pijqij - pij(qij + zij) is referred to as net s s d d revenue, which (to a first-order approximation) zij zi, Loij Lio (15.2) gives the welfare impact of an equiproportionate j j increase in the price of commodity j. where pi is the m-vector of supply prices, qi is the s s With the gain (or loss) to each household corresponding vector of quantities supplied, Li is o calculated based on equation 15.3, the covariates the labor input to own-production activities, of of those gains can now be examined. One covariate which Loij is used in producing good j, fij is the of obvious interest is income, needed to assess household-specific production function for good j impacts on aggregate poverty and inequality. Ide- (embodying fixed factors), and the zi terms are the ally, one would use a money metric of utility based commodities used as production inputs, of which on equation 15.1. However, that would require an zij is used in producing good j. explicit model of the demand and supply system Measurement of the welfare impacts is of course (that can be integrated back to obtain the indirect constrained by the data, which do not include ini- utility function). Again, feasibility becomes an issue tial price and wage levels.8 However, this data limi- because of the absence of complete data on price tation does not matter in calculating a first-order and wage levels. Thus there is little choice but to use approximation to the welfare impact in a neighbor- income as the money metric of utility, in effect hood of the household's optimum. Taking the dif- ignoring all geographic differences in the prices ferentials of equations 15.1 and 15.2 and using the faced or in the extent to which border price changes envelope property (whereby the welfare impacts in are passed on locally. However, we make a seem- a neighborhood of an optimum can be evaluated ingly plausible allowance for urban­rural cost of by treating the quantity choices as given), the living differences in this setting. 264 China and the WTO Two further limitations of this approach should Household Survey by China's National Bureau of be noted. First, applying the calculus in deriving Statistics (NBS). The sample size is 67,900 house- (15.3) implicitly assumes small changes in prices. holds for the rural survey and 16,900 households Relaxing this requires more information on price (out of the survey total of 40,000 households) for levels and the structure of the demand and supply the urban survey.10 Over the past 15 years, NBS has system.9 This would entail considerable further worked to improve both surveys, focusing on sam- effort, and the reliability of the results will be ques- ple coverage, questionnaire design, methodology, tionable given the problem of incomplete price and and data processing.11 The number of variables in wage data. the surveys has increased dramatically, with addi- Second, as already noted and consistent with the tional details on income, expenditure, savings, general equilibrium analysis, this approach also housing, and productivity, among others. NBS also rules out rationing in commodity or factor markets provided micro data for three provinces (Liaoning, or nonconvexities in consumption or production. Guangdong, and Sichuan--the "test provinces"). In principle, these problems can also be handled The computer program to implement the estima- through a completely specified demand model, tion method was written for these data, after which which can be used to estimate the virtual prices at the program was run by NBS staff on the entire which the rationed demand or supply would be national data set. chosen. This is not feasible without data on price A number of problems remain in the 1999 sur- and wage level. veys. For a sample frame, the rural survey relies on its sampled counties from 1985, which may no longer be representative. The urban survey excludes Setting and Data rural migrants, since the base of the sample frame is While the official date of China's WTO accession is the legal registration system (hukou). As in other 2001, it is clear that the Chinese economy had countries the rural survey gives data on the remit- already started to adapt to this expected change tances of migrants workers, but it does not provide well before 2001. The trade reform can thus be information about the migrant workers themselves, thought of as having two stages, a lead-up period in who (unlike in other countries) are not sampled in which tariffs started to fall in anticipation of WTO the urban survey either. This makes it difficult to accession and the period from 2001 onward. measure impacts through labor mobility and rural- Ianchovichina and Martin (2002) argue for 1995 as urban transfers. a plausible beginning of the lead-up period. Their Comparisons between the rural and urban sur- estimates of the price changes induced by WTO veys also present problems. For example, income in accession for the periods 1995­2001 and 2001­07 the rural survey includes in-kind income (such as are used in this analysis. While the primary focus is from own-farm production and other household on the second period, welfare impacts are also esti- enterprises), but income in the urban survey ig- mated for the lead-up period. nores some in-kind components, notably subsidies The measure of welfare impacts given by equa- from the government. tion 15.3 is calibrated to survey data for 1999, two years before official WTO accession and a few years Sampling Weights after the likely beginning of the lead-up period. The The population census puts the 1999 urban popu- choice of 1999 was partly made for data reasons, lation share at 34 percent while the sample-based since it was the most recent year for which the urban population share is 20 percent. To correct micro data were available. Choosing a year near the the rural and urban sampling weights, the urban middle of the lead-up period (rather than a survey population share from the China Statistical Year- at the beginning or end) should also diminish book (NBS 2000) was used to replace the survey biases due to any nonlinearity in the welfare sample weights to form the national figures. impacts of price and wage changes. Matching the Global Trade Analysis Project Model Survey Data and the Surveys The survey data used in this study are from the There are 57 sectors in the Global Trade Analysis 1999 Urban Household Survey and the 1999 Rural Project (GTAP) model. The China GTAP model Welfare Impacts of China's Accession to the WTO 265 used in this study regroups these 57 sectors into 25: transportation and trade, and the like are consid- rice, wheat, feed grains, vegetables and fruits, ered income from skilled nonfarm labor. oilseeds, sugar, plant-based fibers, livestock and meat, dairy, processed food, beverages and tobacco, Land extractive industries, textiles, apparel, light manu- Under China's economic reforms, which began in factures, petrochemicals, metals, automobiles, elec- 1978, all farmers have land use rights but not the tronics, other manufactures, trade and transport, right to sell, although they can subcontract the allo- construction, communications, commercial serv- cated land to other farmers. Therefore, the change ices, and other services. To these are added land, in land prices from the GTAP model affects only capital, and three types of labor (see below). the value of land rentals paid and received. China's rural and urban surveys have about 2,000 categories for consumption and production. The Household Income variables from the household surveys are matched For assessing the overall impacts on poverty and to the closest category in the GTAP model. For inequality, rural and urban households are com- example, corn, millet, and potatoes are placed in the bined. There is no cost of living index between category"feed grains"and cotton and fiber crops are urban and rural areas of China; urban and rural placed in the category of "plant-based fibers." The consumer price indexes are both indexed to 100 at working paper version, Chen and Ravallion 2003 the base date. The urban price level is assumed to gives details on how the variables from the surveys be 15 percent higher than the rural price level. This are matched to the GTAP model sectors. differential is less than for other developing coun- Definitions of Labor and Labor Earnings tries because subsidies to urban households in China help to compensate for higher housing and The China GTAP model defines three types of food costs than in rural areas. labor: unskilled farm labor, unskilled nonfarm Income per person is used as the welfare indica- labor, and skilled nonfarm labor.12 Since the rural tor (so that all households are ranked by per capita and urban surveys have different questionnaires, income, from the poorest to the richest). This is rural and urban labor earnings are treated differ- termed "net income" in the rural survey and "dis- ently. In the urban survey three variables--sector, posable income" in the urban survey. Post-reform occupation, and education--are used to determine income is then income plus the estimated gain de- labor types. However, "sector" or "occupation" fined by equation 15.3. alone cannot indicate whether a person should be classified as skilled labor. For example, the financial Measured Welfare Impacts of WTO sector may hire some unskilled labor while the Accession services sector may hire some skilled labor. Simi- larly, a train driver in the occupation category Based on the predicted relative price and wage "workers and staff-members in production and changes from the GTAP model for 1995­2000 transportation"counts as skilled labor. Accordingly, (table 15.1) and 2001­07 (table 15.2) and produc- education is also taken into account. Workers who tion and consumption shares from the 1999 rural have received education at the senior high school and urban household survey data, equation 15.3 level or above are considered skilled labor. Others can be used to compute the net gain for each are classified as unskilled labor. household. The first panel in table 15.3 gives the It is more difficult to determine the type of labor mean gains for 1995­2001 and 2001­07, split by income for rural areas. There is no information on urban and rural areas. The second panel gives the how much each person earns and from what work. Gini indices, both actual (for the baseline year, Consequently, labor earnings can be classified only 1999) and simulated. The two simulated income roughly by income source. For instance, all labor distributions are obtained by subtracting the remuneration from agriculture is considered in- estimated gains over 1995­2001 from the 1999 come from unskilled farm labor; earnings from incomes at the household level and by adding the industry or construction, grain processing, and household-specific gains from 2001­07 to the 1999 the like are considered income from unskilled incomes. Thus the first simulation shows the nonfarm labor; earnings from the services sector, distributional impact of the price changes during 266 China and the WTO TABLE 15.1 Predicted Price Changes from GTAP Model and Per Capita Net Gain or Loss for Rural and Urban Households, 1995­2001 Rural Urban Wholesale Consumer Net Mean Welfare Net Mean Welfare Expenditures and Prices Prices Revenue Change Revenue Change Income Sources (Percent) (Percent) (Yuan) (Yuan) (Yuan) (Yuan) Expenditures Rice 0.5 1.5 73.66 0.15 -109.33 -1.64 Wheat -1.7 -1.5 40.86 -0.74 0.00 0.00 Feedgrains 2.6 10.7 117.04 2.15 0.00 0.00 Vegetables and fruits 0.5 1.5 123.41 0.13 -378.69 -5.68 Oilseeds -0.6 -0.8 37.05 -0.24 -1.04 0.01 Sugar 0.7 1.4 13.74 0.05 -174.06 -2.44 Plant-based fibers -3.6 -1.9 36.84 -1.34 0.00 0.00 Livestock and meat 2.0 3.1 194.62 2.59 -500.65 -15.52 Dairy 1.5 2.5 2.50 0.02 0.00 0.00 Other food 1.2 3.1 -81.60 -3.39 -343.13 -10.64 Beverages and tobacco -4.6 -7.2 -72.98 5.25 -197.20 14.20 Extractive industries -0.2 0.8 17.99 -0.44 -173.03 -1.38 Textiles -5.0 -8.9 -11.08 0.99 -53.50 4.76 Apparel -2.7 -7.4 -64.13 4.75 -394.30 29.18 Light manufacturing -0.3 -2.5 -16.15 0.40 -82.96 2.07 Petrochemical industry -0.7 -0.1 -325.39 0.33 -398.23 0.40 Metals -0.7 -0.1 -15.30 0.02 -24.02 0.02 Automobiles -17.7 -20.4 -52.27 10.66 -37.76 7.70 Electronics -1.5 -4.0 -24.27 0.97 -162.69 6.51 Other manufactures -0.6 -0.3 -264.61 0.79 -431.16 1.29 Trade and transport 0.2 1.3 -18.70 -0.24 -110.53 -1.44 Construction 0.1 1.1 0.00 0.00 -31.11 -0.34 Communication 0.9 1.9 -16.72 -0.32 -152.04 -2.89 Commercial services 0.8 1.8 -61.37 -1.10 -533.33 -9.60 Other services 0.1 1.1 -414.45 -4.56 -680.99 -7.49 Income sources Farm unskilled labor 1.7 1.7 313.58 5.22 0.00 0.00 Nonfarm unskilled labor 1.7 1.7 287.19 4.78 1227.51 20.44 Skilled labor 2.0 2.0 360.87 7.09 3391.11 66.64 Land 1.3 1.3 17.08 0.22 0.00 0.00 Capital 1.3 1.3 21.14 0.27 126.01 0.77 Sources: Ianchovichina and Martin 2002; author's calculations from China National Bureau of Statistics 1999 Rural Household Survey and 1999 Urban Household Survey. the first stage of the trade reform (what the baseline National Bureau of Statistics and for the $1 a day distribution would have looked like without the and $2 a day poverty lines from Chen and Ravallion reforms) while the second shows the impact of (2001). the post-2001 price changes (how the changes are There is an overall gain of about 1.5 percent of expected to affect the baseline distribution, looking mean income.All of this gain is in the period leading forward). The third panel gives the headcount up to WTO accession. There is almost no impact on index of poverty for the official poverty line inequality, either in the period leading up to WTO based on the poverty lines used by China's accession or predicting forward. The aggregate Gini Welfare Impacts of China's Accession to the WTO 267 TABLE 15.2 Predicted Price Changes from GTAP Model and Per Capita Net Gain or Loss for Rural and Urban Households, 2001­07 Rural Urban Wholesale Consumer Net Mean Welfare Net Mean Welfare Expenditures and Prices Prices Revenue Change Revenue Change Income Sources (Percent) (Percent) (Yuan) (Yuan) (Yuan) (Yuan) Expenditures Rice -1.4 0.7 73.66 -1.39 -109.33 -0.75 Wheat -1.5 0.7 40.86 -0.92 0.00 0.00 Feedgrains -3.7 2.1 117.04 -4.90 0.00 0.00 Vegetables and fruits -2.6 -0.6 123.41 -4.02 -378.69 2.24 Oilseeds -5.7 -5.9 37.05 -2.10 -1.04 0.06 Sugar -2.8 -3.5 13.74 -0.34 -174.06 6.01 Plant-based fibers 1.6 4.1 36.84 0.56 0.00 0.00 Livestock and meat -1.5 0.7 194.62 -5.21 -500.65 -3.40 Dairy -2.4 -0.5 2.50 -0.09 0.00 0.00 Other food -3.1 -2.7 -81.60 2.04 -343.13 9.32 Beverages and tobacco -5.6 -7.7 -72.98 5.62 -197.20 15.09 Extractive industries -0.4 1.7 17.99 -0.86 -173.03 -2.92 Textiles -0.2 -1.5 -11.08 0.17 -53.50 0.82 Apparel 2.6 0.8 -64.13 -0.51 -394.30 -2.98 Light manufacturing -0.6 0.5 -16.15 -0.08 -82.96 -0.43 Petrochemical industry -1.1 0.8 -325.39 -2.60 -398.23 -3.19 Metals -0.6 1.3 -15.30 -0.20 -24.02 -0.31 Automobiles -3.8 -4.0 -52.27 2.09 -37.76 1.52 Electronics -1.2 -1.4 -24.27 0.34 -162.69 2.20 Other manufactures -0.8 0.8 -264.61 -2.12 -431.16 -3.46 Trade and transport -0.4 1.7 -18.70 -0.32 -110.53 -1.85 Construction -0.4 1.7 0.00 0.00 -31.11 -0.52 Communication -0.4 1.7 -16.72 -0.28 -152.04 -2.54 Commercial services -1.1 0.9 -61.37 -0.55 -533.33 -4.72 Other services -0.7 1.3 -414.45 -5.39 -680.99 -8.76 Income sources Farm unskilled labor -0.3 -0.3 313.58 -0.85 Nonfarm unskilled labor 1.0 1.0 287.19 2.96 1227.51 12.64 Skilled labor 0.4 0.4 360.87 1.55 3391.11 14.58 Land -4.7 -4.7 17.08 -0.80 Capital 0.6 0.6 21.14 0.13 126.01 0.80 Sources: Ianchovichina and Martin 2002; authors' calculations from China National Bureau of Statistics 1999 Rural Household Survey and 1999 Urban Household Survey. index increased slightly, from 39.3 percent without by the remaining tariff changes during that period. WTO accession to 39.5 percent after accession. The impacts on rural and urban poverty for a wide The incidence of poverty would have been range of poverty lines can be seen in figures 15.1 slightly higher in 1999 if not for the trade policy and 15.2, which give the cumulative distributions of changes in the lead-up to WTO accession, while income for both the baseline and the two simulated poverty is predicted to increase slightly during distributions for the poorest 60 percent in rural 2001­07 due to the expected price changes induced areas and 40 percent in urban areas, respectively. 268 China and the WTO TABLE 15.3 Summary Statistics on Aggregate Welfare Impacts, 1995­2001 and 2001­07 Item Rural Urban National Mean gains (yuan per capita) 1995-2001 34.47 94.94 55.49 (1.54%)a 2001-07 -18.07 29.45 ­1.54 (­0.04%)a Inequality impacts (Gini index as percentage) Baseline, 1999 33.95 29.72 39.31 Simulated: Less gains 1995­2001 33.90 29.68 39.27 Simulated: Plus gains 2001­07 34.06 29.65 39.53 Poverty impacts (headcount index, percentage)b Official poverty line Baseline, 1999 4.38 0.08 2.92 Simulated: Less gains 1995­2001 4.56 0.08 3.04 Simulated: Plus gains 2001­07 4.57 0.07 3.04 $1/day (1993 purchasing power parity) Baseline, 1999 10.51 0.29 7.04 Simulated: Less gains 1995­2001 10.88 0.28 7.28 Simulated: Plus gains 2001­07 10.81 0.28 7.23 $2/day (1993 purchasing power parity) Baseline, 1999 45.18 4.07 31.20 Simulated: Less gains 1995­2001 46.10 4.27 31.88 Simulated: Plus gains 2001­07 45.83 3.97 31.60 a. Percentage of mean income. b. Official poverty line is from China National Statistics Bureau; poverty lines of $1 a day and $2 a day are from Chen and Ravallion 2001. Sources: Authors' computations from China National Bureau of Statistics 1999 Rural Household Survey and 1999 Urban Household Survey. FIGURE 15.1 Poverty Incidence Curves, Rural Percentage of population below poverty line 60 50 40 30 20 10 0 400 600 800 1,000 1,200 1,400 1,600 1,800 2,000 Annual per capita income (yuan) Pre-WTO Post-WTO Baseline distribution Sources: China National Bureau of Statistics; Chen and Ravallion 2001. Welfare Impacts of China's Accession to the WTO 269 FIGURE 15.2 Poverty Incidence Curves, Urban Percentage of population below poverty line 40 35 30 25 20 15 10 5 0 500 1,000 1,500 2,000 2,500 3,000 3,500 4,000 4,500 Annual per capita income (yuan) Pre-WTO income Baseline distribution Post-WTO income in 1999 Source: Authors' computations based on data from China National Bureau of Statistics 1999 Rural Household Survey and 1999 Urban Household Survey. While there is virtually zero aggregate impact will experience a real income loss. Farm income is when predicting forward from WTO accession, the predicted to drop by 18 yuan per person while disaggregated results show a more nuanced picture. urban income rises by 29 yuan per person. The The analysis focuses on three measures of impact breakdown by sectors in table 15.2 shows that at the household level: the absolute gain or loss, gi; the decline in rural income is due to the drop in the proportionate gain or loss, gi/yi; and whether wholesale prices for most farm products, plus there is a gain or not, I (gi), where I is the indicator higher prices for education and health care. Farm- function. This third measure helps to determine ers will also benefit from the drop in some con- where there might be high concentrations of losers, sumer prices and from the increase in nonfarm in specific areas or socioeconomic groups. labor wages. In urban areas residents will enjoy The results by provinces ranked by mean income lower prices for most farm products and higher per person are plotted in figure 15.3 for mean wages, but they will also be hit by increases in serv- absolute gains (gi in yuan per capita), in figure 15.4 ice fees for education and health care. for proportionate gains (gi/yi, as a percentage), Impacts differ considerably across regions (see and in figure 15.5 for the proportion of households figures 15.3­15.5 and appendix tables 15.A.1 and that registered positive gains. The average gain or 15.A.2).Themeanabsolutegainstendtobehighestin loss by province for urban and rural areas and the the richest provinces in both urban and rural areas number of gainers in each case are shown in appen- (figure 15.3), though there is no correlation between dix tables 15.A.1 and 15.A.2, respectively; Chen and the proportionate gains and mean income of Ravallion 2003 give the province rankings. provinces (figure 15.4). One spatially contiguous The same results are also plotted in figures 15.6­ region--the northeast provinces of Inner Mongolia, 15.8 against percentiles of the income distribution. Liaoning,Jilin,and Heilongjiang--stands out as hav- So, for example, to see the mean impact in yuan per ingthelargestlossfromthereform.Bothabsoluteand capita at the median income one looks at the 50th proportionate impacts are highest in this region-- percentile of figure 15.6. (Notice that figure 15.6 more than 90 percent of farmers in Heilongjiang and gives the horizontal differences in figures 15.1 and Jilin are predicted to experience a net loss. 15.2 plotted against the point on the vertical axis.) Notice that these geographic differences in In the aggregate, about three-quarters of rural welfare impacts arise entirely from differences in households and one-tenth of urban households consumption and production behavior. In reality, 270 China and the WTO FIGURE 15.3 Mean Gains by Provinces: Absolute Gain in Yuan Per Capita Net gain/loss per capita (yuan) 60 40 20 0 20 40 60 80 1 6 11 16 21 26 31 Provinces ranked by provincial per capita income Urban Total Rural Source: Authors' computations based on data from China National Bureau of Statistics 1999 Rural Household Survey and 1999 Urban Household Survey. FIGURE 15.4 Mean Gains by Provinces: Proportionate Gains in Percent Percentage net gain/loss per capita 2 1 0 1 2 3 1 6 11 16 21 26 31 Provinces ranked by provincial per capita income Urban Total Rural Source: Authors' computations based on data from China National Bureau of Statistics 1999 Rural Household Survey and 1999 Urban Household Survey. Welfare Impacts of China's Accession to the WTO 271 FIGURE 15.5 Mean Gains by Provinces: Percentage of Gainers by Provinces Percentage of gainers 100 90 80 70 60 50 40 30 20 10 0 1 6 11 16 21 26 31 Provinces ranked by provincial per capita income Urban Total Rural Note: Urban prices are assumed to be 15 percent higher than rural prices. Source: Authors' computations based on data from China National Bureau of Statistics 1999 Rural Household Survey and 1999 Urban Household Survey. FIGURE 15.6 Mean Gains in Yuan by Income Percentile Net gain/loss per capita (yuan) 40 30 20 10 0 10 20 30 40 50 60 0 10 20 30 40 50 60 70 80 90 100 Percentage of population ranked by per capita income Urban National Rural Note: Urban prices are assumed to be 15 percent higher than rural prices. Source: Authors' compilations based on data from China National Bureau of Statistics 1999 Rural Household Survey and 1999 Urban Household Survey. 272 China and the WTO FIGURE 15.7 Mean Percentage Gain by Income Percentile Percentage net gain or loss per capita 2.0 1.0 0.0 1.0 2.0 3.0 4.0 5.0 6.0 0 10 20 30 40 50 60 70 80 90 100 Percentage of population ranked by per capita income Urban National Rural Note: Urban prices are assumed to be 15 percent higher than rural prices. Source: Authors' compilations based on data from China National Bureau of Statistics 1999 Rural Household Survey and 1999 Urban Household Survey. FIGURE 15.8 Percentage of Gainers by Income Percentile Percentage of gainers 100.0 90.0 80.0 70.0 60.0 50.0 40.0 30.0 20.0 10.0 0 10 20 30 40 50 60 70 80 90 100 Percentage of population ranked by per capita income Urban Total Rural Note: Urban prices are assumed to be 15 percent higher than rural prices. Source: Authors' compilations based on data from China National Bureau of Statistics 1999 Rural Household Survey and 1999 Urban Household Survey. Welfare Impacts of China's Accession to the WTO 273 differential impacts on local prices are likely, due to naturally spend more on food, so if the relative transport or other impediments to internal trade. price of food changes, the welfare impacts will be The analysis here does not incorporate such differ- correlated with this aspect of household demo- ences, and doing so would pose a number of data graphics. Similarly, there may be differences in and analytic problems. This might, however, be a tastes associated with stage of the life cycle and edu- fruitful direction for future work where the neces- cation. There are also likely to be systematic covari- sary data on prices and wage levels are available by ates of the composition of income. geographic area. Generically, the gain can be written as: When households are ranked by initial income, gi = g pi , pi , wi, x1 , x2 d s i i there is a notable difference between urban and m s rural households, with absolute gains tending to be = pijqs pi , pi , wi, x2 dpij s d s i higher for higher income households in urban j=1 pij s areas, but lower for higher income households in - pij qd pi , wi, i, x1 d d i rural areas (see figure 15.6). Nationally (combining d urban and rural areas with the corrected weights +zij pi , pi, wi, x2 dpij d s i d discussed above), there is the hint of a U-shaped pij n relationship, though still with the highest absolute + wk Lik pi , wi, i, x1 d i gains for the rich. k=1 This picture is reversed for proportionate gains, o d s dwk (15.4) which tend to fall as income rises in urban areas, - Lik pi , pi, wi, x2 i wk but to rise with income in rural areas and nation- Notice that the gain from reform is inherently non- ally (see figure 15.7). In the aggregate the propor- separable, in that it cannot be written as a function tion of gainers rises with income, a result that is d driven by the rise in the number of gainers as solely of pi , x1 and i because the gain also i depends on production choices. income increases in rural areas (see figure 15.8). However, as noted above, household-specific wages and prices are not observed, so further assumptions are required. In explaining variations Explaining the Incidence of Gains across households in the predicted gains from trade and Losses reform, wage rates are assumed to be a function The way the problem of measuring welfare impacts of prices and characteristics as wi = w(pi , pi , d s was formulated in section II allows utility and x1 , x2 ), and differences in prices faced are i i profit functions to vary between households at assumed to be adequately captured by a complete given prices. To explain the heterogeneity in meas- set of county-level dummy variables. ured welfare impacts, these functions can instead Under these assumptions, and the linearization be supposed to vary with observed household char- of equation 16.4 with an additive innovation error acteristics. The indirect utility function becomes term, the following regression model applies for the vi(pi , wi, i) = v(pi , wi, i, x1 ) d d i where i = gains: (pi , pi , wi, x2 ) for vectors of characteristics x1 s d i i and x2 that shift the utility functions in equation i gi = 1x1 + 2x2 + i i k Dki + i (15.5) 15.1 and the profit functions in 15.2. Note that k the characteristics that influence preferences over where Dki = 1 if household i lives in county k and consumption (x1 ) are allowed to differ from those i Dki = 0 otherwise and i is the error term. that influence the outputs from own-production The characteristics considered include age and activities (x2 ). i age-squared of the household head, education and The gain from the price changes induced by demographic characteristics, and land (interpreted trade reform, as given by equation 15.3, depends on as a fixed factor of production, since it is allocated the household's consumption, labor supply, and largely by administrative means in rural China). production choices, which in turn depend on prices Also included are dummy variables describing and characteristics, x1 and x2 . For example, house- i i some key aspects of the occupation and princi- holds with a higher proportion of children will ple sector of employment, such as whether the 274 China and the WTO TABLE 15.4 Regression Results for Level of Gain (Yuan) in Rural Areas of Three Provinces, 2001­07 Variable Liaoning Guangdong Sichuan Log of household size 37.642 (6.42) 28.822 (2.64) 4.958 (2.16) Age of household head -2.425 (-3.11) -1.783 (-2.60) -0.548 (-1.51) Age of household head squared 0.026 (3.36) 0.017 (2.66) 0.005 (1.30) Agriculture household -10.942 (-3.31) -42.850 (-6.45) -37.723 (-6.54) Number of employees/ 12.665 (4.10) -6.932 (-0.29) 12.652 (3.02) household size Number of township and 10.768 (3.13) 29.466 (3.06) 15.327 (4.26) village enterprise workers/ household size Number of migrant workers/ 5.399 (1.73) 7.798 (2.35) 7.067 (3.79) household size Area of cultivated land -0.027 (-5.73) -0.002 (-1.00) -0.001 (-0.28) Area of hilly land 0.000 (-0.05) -0.001 (-0.87) 0.002 (1.94) Area of fishpond land -0.001 (-0.94) -0.070 (-2.85) 0.000 (0.04) Highest education level Illiterate or semi illiterate 7.926 (1.04) 19.016 (1.25) 8.387 (0.92) Primary school 0.071 (0.01) -2.148 (-0.13) 9.694 (1.06) Middle school -0.755 (-0.11) -4.261 (-0.26) 7.669 (0.84) High school 2.125 (0.31) 2.806 (0.18) 9.675 (1.03) Technical school -3.096 (-0.44) -36.482 (-1.09) 4.270 (0.38) College (default) Labor force/household size 0.576 (0.08) 2.877 (0.15) -4.995 (-1.16) Children under 6/household 46.999 (2.71) 8.109 (0.35) -2.291 (-0.45) size Children ages 6-11/household 1.414 (0.11) 2.247 (0.10) -9.011 (-1.50) size Children ages 12-14/household -0.155 (-0.01) -24.489 (-1.20) -9.606 (-1.51) size Children ages 15-17/household -2.592 (-0.22) -23.390 (-1.02) -5.485 (-0.73) size Constant -17.851 (-0.82) -17.742 (-0.65) -17.220 (-1.43) R-square 0.278 0.116 0.116 Note: Numbers in parentheses are t-statistics. Source: Authors' computations based on data from China National Bureau of Statistics 1999 Rural Household Survey and 1999 Urban Household Survey. household is a registered agricultural household, rural areas in each of the three test provinces whether there is wage employment, whether there (Liaoning, Guangdong, and Sichuan) for which is state-sector employment, and whether there is complete micro data are available. participation in a township and village enterprise. There are some differences in the explanatory There are endogeneity concerns about these vari- variables between rural (tables 15.4 and 15.5) and ables, but they appear to be minor in this context, urban areas (tables 15.6 and 15.7). Results are pre- especially when weighed against the concerns about sented for both absolute gains (gi) and proportion- omitted variable bias in estimates that exclude these ate gains (gi/yi ). Recall that these are averages characteristics. Under the usual assumption that across the impacts of these characteristics on the the error term is orthogonal to these regressors, consumption and production choices that deter- equation 5 is estimated by ordinary least squares. mine the welfare impact of given price and wage The model is estimated separately for urban and changes. This makes interpretation difficult. These Welfare Impacts of China's Accession to the WTO 275 TABLE 15.5 Regression Results for Percentage Gains in Rural Areas of Three Provinces, 2001­07 Variable Liaoning Guangdong Sichuan Log of household size 0.768 (2.46) 0.022 (0.20) 0.030 (0.40) Age of household head -0.108 (-2.17) -0.007 (-0.34) -0.004 (-0.31) Age of household head squared 0.001 (2.19) 0.000 (0.40) 0.000 (-0.02) Agriculture household -0.896 (-2.98) -1.365 (-14.85) -1.420 (-7.58) Number of employees/ 0.630 (2.76) 0.271 (2.57) 0.444 (3.61) household size Number of township and 0.669 (4.27) 0.585 (4.47) 0.548 (6.11) village enterprise workers/household size Number of migrant workers/ 0.655 (3.59) 0.187 (3.59) 0.346 (7.08) household size Area of cultivated land 0.000 (-1.77) 0.000 (-0.73) 0.000 (-1.61) Area of hilly land 0.000 (-0.48) 0.000 (-0.35) 0.000 (2.20) Area of fishpond land 0.000 (-0.17) -0.001 (-2.23) 0.000 (0.55) Highest education level Illiterate or semi illiterate 1.393 (2.18) 0.507 (1.26) -0.013 (-0.05) Primary school -0.634 (-2.01) -0.154 (-0.90) 0.069 (0.30) Middle school -0.891 (-3.08) -0.023 (-0.14) -0.011 (-0.05) High school -0.660 (-2.42) 0.010 (0.06) 0.006 (0.02) Technical school -0.573 (-1.87) -0.229 (-1.18) 0.038 (0.14) College (default) Labor force/household size 0.456 (0.85) 0.323 (1.81) -0.099 (-0.71) Children under age 6/household 3.730 (3.61) 0.461 (1.49) -0.169 (-0.78) size Children ages 6-11/household 1.557 (1.41) 0.173 (0.72) -0.275 (-1.48) size Children ages 12-14/household 1.625 (1.54) -0.477 (-1.60) -0.343 (-1.85) size Children ages 15-17/household 1.325 (1.80) -0.289 (-0.91) -0.192 (-0.88) size Constant 0.788 (0.69) -0.709 (-1.39) -0.584 (-1.68) R-square 0.108 0.217 0.171 Note: Numbers in parentheses are t-statistics. Source: Authors' computations based on data from China National Bureau of Statistics 1999 Rural Household Survey and 1999 Urban Household Survey. regressions are mainly of descriptive interest, to more township and village enterprise workers, with help isolate covariates of potential relevance in more migrant workers, and with less cultivated thinking about compensatory policy responses. land (though significant only in Liaoning). The only For rural areas the results show that the predicted strongdemographiceffectisthatyoungerhouseholds gain from trade reform tends to be larger for larger (with a higher proportion of children under six) tend households in all three provinces. There is also a U- to be gainers in Liaoning. While the results for shaped relationship with age of the household head: the county dummy variables are not shown (to save thegainsreachaminimumaround50yearsof age(47 space), losses were significantly higher than average in Liaoning, 52 in Guangdong, and 55 in Sichuan). in six counties in Liaoning,seven in Guangdong,and The gains are lower for agricultural households and six in Sichuan. Table 16.8 gives the mean losses in higher for households with more employees and thesecountiesforagriculturalhouseholds. 276 China and the WTO TABLE 15.6 Regression Results for Level of Gain (Yuan) in Urban Areas of Three Provinces, 2001­07 Variable Liaoning Guangdong Sichuan Log of household size -5.627 (-1.81) 5.289 (0.27) -19.441 (-4.09) Single household head -1.366 (-0.4) -37.216 (-2.06) -17.369 (-3.61) Age of household head 0.531 (0.92) 5.266 (2.43) 1.542 (2.34) Age of household head squared -0.001 (-0.24) -0.040 (-1.8) -0.015 (-2.22) Highest education level (default is college) Primary school or lower 13.240 (2.95) 50.434 (2.4) 23.079 (3.11) Middle school 19.104 (5.99) 56.659 (3.58) 26.096 (4.34) High school 5.123 (1.62) 12.053 (0.95) 12.717 (2.39) Technical school 11.086 (3.23) 11.075 (0.88) 9.552 (1.62) College 3.974 (1.26) 3.447 (0.3) 11.013 (2.12) Sector (default is government) Agriculture -16.310 (-1.22) -25.590 (-2.23) 17.293 (1.76) Mining -14.586 (-3.24) 19.351 (1.13) -3.851 (-0.53) Manufacturing -9.231 (-2.59) 17.773 (1.28) -4.634 (-1.2) Utility -9.387 (-1.63) -10.816 (-0.42) 1.516 (0.13) Construction -6.394 (-1.18) 8.622 (0.63) -4.409 (-0.92) Geological prospecting and -27.422 (-2.62) 20.089 (0.92) -16.585 (-0.83) water conservancy Transportation and 6.368 (1.52) 16.525 (1.24) 1.644 (0.25) telecommunications Wholesale and retail -3.184 (-0.61) 5.664 (0.45) -1.983 (-0.4) Banking and finance -5.278 (-0.55) 3.888 (0.3) 9.491 (0.85) Real estate -11.708 (-1.71) 46.192 (1.35) 7.670 (0.37) Social services -5.542 (-1.02) -4.186 (-0.33) 0.504 (0.1) Health care -9.260 (-1.93) 0.683 (0.04) -1.049 (-0.17) Education -7.279 (-1.64) 7.649 (0.46) -5.219 (-0.87) Scientific research -20.982 (-4.06) 17.882 (1.14) -7.929 (-0.59) Other -7.784 (-1.42) -24.851 (-0.75) -7.012 (-0.73) Type of employer (default is state owned) Collective owned -1.927 (-0.76) 11.882 (0.54) -5.946 (-2.09) Foreign company -3.138 (-0.72) -10.988 (-1.22) 2.038 (0.31) Self-employed 4.278 (0.6) 9.448 (0.64) 10.582 (2.08) Privately owned business -9.587 (-1.41) -14.823 (-0.99) -4.601 (-0.57) Retirees reemployed -13.333 (-2.45) -35.591 (-1.82) -6.752 (-0.99) Retirees -15.569 (-3.66) -49.442 (-1.91) -12.218 (-1.95) Other -10.350 (-1.36) -6.568 (-0.34) -16.796 (-2.06) Occupation (default is retiree) Engineer and technician 10.244 (1.66) 3.479 (0.12) 10.179 (1.49) Officers 12.747 (2.07) 17.701 (0.64) 10.564 (1.53) Staff in commerce 11.742 (2.08) 18.553 (0.65) 12.734 (1.92) Staff in services 19.940 (2.54) 3.380 (0.11) 4.057 (0.5) Worker in manufacturing 17.484 (2.02) 13.151 (0.47) 13.810 (1.86) Worker in transportation 21.469 (3.59) 9.637 (0.34) 16.117 (2.35) and telecommunications Other 15.318 (2.05) 9.810 (0.27) -6.141 (-0.77) Constant -10.744 (-0.77) -164.442 (-2.43) -17.611 (-1.1) R-square 0.265 0.131 0.181 Note: Numbers in parentheses are t-statistics. Source: Authors' computations based on data from China National Bureau of Statistics 1999 Rural Household Survey and 1999 Urban Household Survey. Welfare Impacts of China's Accession to the WTO 277 TABLE 15.7 Regression Results for Percentage Gains in Urban Areas of Three Provinces, 2001­07 Variable Liaoning Guangdong Sichuan Log of household size 0.175 (3.54) -0.038 (-0.4) 0.036 (0.46) Single household head -0.022 (-0.36) -0.221 (-2.21) -0.259 (-3.07) Age of household head 0.000 (-0.01) 0.033 (2.55) 0.017 (1.53) Age of household head squared 0.000 (0.1) 0.000 (-2.12) 0.000 (-1.46) Highest education level (default is university) Primary school or lower 0.524 (6.43) 0.389 (3.7) 0.509 (5.15) Middle school 0.539 (10.41) 0.583 (7.25) 0.591 (8.27) High school 0.180 (3.56) 0.095 (1.46) 0.262 (3.83) Technical school 0.214 (4.04) 0.076 (1.22) 0.120 (1.79) College 0.054 (1.04) 0.015 (0.25) 0.125 (2.24) Sector (default is government) Agriculture -0.079 (-0.32) 0.166 (2.2) 0.338 (2.64) Mining 0.183 (1.11) 0.346 (3.38) -0.129 (-1.01) Manufacturing -0.015 (-0.27) 0.114 (1.41) -0.021 (-0.34) Utility -0.040 (-0.36) -0.144 (-1.18) -0.134 (-0.84) Construction 0.095 (0.91) 0.109 (1.19) 0.036 (0.51) Geological prospecting -0.407 (-3.06) 0.178 (1.03) -0.228 (-0.53) and water conservancy Transport and 0.206 (2.93) 0.060 (0.79) -0.036 (-0.4) telecommunications Wholesale and retail 0.060 (0.78) 0.081 (0.99) -0.015 (-0.18) Banking and finance -0.088 (-0.47) 0.049 (0.53) 0.013 (0.12) Real estate -0.108 (-0.91) 0.222 (1.16) 0.106 (0.29) Social services -0.090 (-1.09) 0.065 (0.69) 0.148 (1.37) Health care -0.088 (-1.1) 0.007 (0.06) -0.124 (-1.49) Education -0.057 (-0.75) 0.044 (0.44) -0.031 (-0.39) Scientific research -0.454 (-4.09) 0.126 (1.11) -0.082 (-0.73) Other 0.012 (0.14) 0.034 (0.25) -0.121 (-0.55) Type of employer (default is state owned) Collective owned 0.053 (1.16) 0.008 (0.08) 0.137 (1.73) Foreign company -0.046 (-0.54) -0.122 (-2.3) -0.193 (-2.08) Self-employed -0.069 (-0.59) -0.051 (-0.39) 0.317 (2.46) Privately owned business -0.182 (-1.65) -0.231 (-1.96) -0.037 (-0.22) Retirees reemployed -0.302 (-3.39) -0.242 (-1.41) -0.177 (-1.32) Retirees -0.341 (-4.2) -0.452 (-2.37) -0.359 (-3.42) Other -0.124 (-1.13) -0.187 (-1.24) -0.338 (-1.2) Occupation (default is retiree) Engineer and technician -0.015 (-0.14) -0.141 (-0.69) -0.036 (-0.29) Officers -0.044 (-0.43) -0.063 (-0.31) -0.045 (-0.36) Staff in commerce 0.012 (0.12) -0.036 (-0.17) 0.029 (0.24) Staff in services 0.437 (3.08) 0.019 (0.09) -0.011 (-0.08) Worker in manufacturing 0.118 (0.82) 0.025 (0.12) 0.091 (0.56) Worker in transport and 0.209 (2.02) -0.018 (-0.09) 0.130 (1.03) telecommunications Other 0.171 (1.33) -0.069 (-0.27) -0.636 (-4.2) Constant 0.172 (0.7) -0.623 (-1.68) -0.197 (-0.71) R-square 0.401 0.290 0.359 Note: Numbers in the parentheses are t-statistics. Source: Authors' computations based on data from China National Bureau of Statistics 1999 Rural Household Survey and 1999 Urban Household Survey. 278 China and the WTO TABLE 15.8 Average Impacts for Agricultural Households in Selected Counties, 2001­07 Gain Provincial Mean County Province Identifier Yuan Percent Yuan Percent Liaoning 210181 -73.72 -3.07 -32.34 -1.29 210212 -145.40 -2.99 210381 -172.01 -5.57 210921 -57.70 -5.21 211321 -45.58 -3.78 211322 -53.60 -3.23 Guangdong 440111 -107.31 -2.74 -29.34 -0.81 440126 -183.63 -2.64 440223 -102.33 -3.53 440523 -148.90 -2.55 440620 -227.23 -3.11 440621 -109.59 -2.64 441425 -316.49 -5.34 Sichuan 510121 -130.46 -2.86 -12.31 -0.67 510125 -63.19 -3.81 512425 -138.34 -5.71 512610 -52.23 -3.11 512825 -40.44 -2.80 513021 -93.02 -4.07 Note: A negative sign means a net loss. Agricultural household means that more than 75 percent of income is from agriculture. Source: Authors' computations based on data from China National Bureau of Statistics 1999 Rural Household Survey and 1999 Urban Household Survey. In urban areas the gains tend to be higher for mean household income, inequality, and poverty smaller households (except in Guangdong). As in incidence. There is, however, a sizable (and at least rural areas there is a U-shaped pattern (except for partly explicable) variance in impacts across house- Liaoning), with lowest gains at 66 years of age in hold characteristics. Rural families tend to lose; Guangdong and 51 in Sichuan. While there is no urban households tend to gain. Some provinces pattern in the relationship between education and experience larger impacts than others; highest welfare gains in rural areas, the gains in urban areas impacts occur in the northeast region of Inner tend to be larger for less educated households. Mongolia, Liaoning, Jilin, and Heilongjiang, where However, this may be biased by the fact that educa- rural households are more dependent on feed grain tion was used in identifying skilled labor (noting production (for which falling prices are expected that unskilled nonfarm wages are predicted to from WTO accession) than elsewhere in China. increase relative to skilled labor; see table 15.2). Within rural or urban areas of a given province, There are some signs of sectoral effects, though the gains from trade reform vary with observable only significantly so in Liaoning, with higher gains household characteristics. The most vulnerable for those with government jobs. Retirees tend to households tend to be in rural areas, dependent have lower gains than others. on agriculture, with relatively fewer workers and with weak economic links to the outside economy through migration. There are also some strong geo- Conclusion graphic concentrations of adverse impacts. For In the aggregate, the analysis finds that China's example, agricultural households in some counties trade reforms have had only a small impact on incur welfare losses of 3­5 percent of their incomes. Welfare Impacts of China's Accession to the WTO 279 Naturally, the approach taken here has limita- involuntary unemployment), for example. In prin- tions. For example, there may well be dynamic ciple, there are ways of dealing with these prob- gains from greater trade openness that are not lems by estimating complete demand and supply being captured by the model used to generate the systems that allow for rationing. This may prove a relative price impacts. Trade may facilitate learning fruitful avenue for future research, though it about new technologies and innovation, bringing should be noted that these methods generate their longer term gains in productivity. Trade reform own problems, such as those arising from incom- may also come with (and possibly help induce) plete data on price and wage levels at household other policy reforms, such as in factor markets. The level. approach here has attempted to capture only the Despite these limitations, the type of approach static welfare effects of WTO accession. followed here can usefully illuminate the range of A further limitation was the need to make lin- welfare impacts to be expected from economywide ear approximations in the neighborhood of an ini- reforms. By avoiding unnecessary aggregation of tial optimum for each household. In other appli- the primary household-level data, these relatively cations this could be deceptive if price or wage simple tools can also offer insights into the sorts of changes are large or if the household were initially policy responses that might be called for to com- out of equilibrium, due to rationing (including pensate losers from reform. 280 China and the WTO Appendix TABLE 15.A.1 Rural Gains and Losses by Province, 2001­07 Number of Original Post-WTO Share of Sampled Number of Income Income Gain or Change Losers Province Households Gainers (Yuan) (Yuan) Loss (Yuan) (Percent) (Percent) Beijing 750 381 4,221.05 4,210.08 -10.96 -0.26 49.20 Tianjin 595 219 3,401.71 3,380.48 -21.22 -0.62 63.19 Hebei 4,200 1,310 2,441.50 2,426.82 -14.68 -0.60 68.81 Shanxi 2,100 926 1,772.62 1,765.13 -7.49 -0.42 55.90 Inner Mongolia 2,198 206 2,055.49 2,011.26 -44.22 -2.15 90.63 Liaoning 1,886 353 2,501.98 2,469.64 -32.34 -1.29 81.28 Jilin 1,598 132 2,260.12 2,210.46 -49.66 -2.20 91.74 Heilongjiang 1,997 115 2,166.59 2,114.18 -52.41 -2.42 94.24 Shanghai 600 416 5,409.11 5,428.79 19.68 0.36 30.67 Jiangsu 3,400 1,209 3,495.20 3,486.78 -8.42 -0.24 64.44 Zhejiang 2,693 1,148 3,946.44 3,934.92 -11.52 -0.29 57.37 Anhui 3,095 676 1,900.76 1,885.79 -14.97 -0.79 78.16 Fujian 1,750 469 3,091.39 3,071.40 -19.99 -0.65 73.20 Jiangxi 2,450 553 2,129.45 2,117.26 -12.19 -0.57 77.43 Shandong 4,200 822 2,520.76 2,494.89 -25.87 -1.03 80.43 Henan 4,200 828 1,948.36 1,931.70 -16.66 -0.86 80.29 Hubei 3,188 755 2,212.71 2,200.04 -12.68 -0.57 76.32 Hunan 3,700 1,181 2,102.98 2,095.39 -7.60 -0.36 68.08 Guangdong 2,560 514 3,628.95 3,599.61 -29.34 -0.81 79.92 Guangxi 2,310 309 2,048.33 2,025.75 -22.58 -1.10 86.62 Hainan 718 28 2,086.40 2,057.85 -28.55 -1.37 96.10 Chongqing 1,500 404 1,736.63 1,730.20 -6.43 -0.37 73.07 Sichuan 3,998 879 1,843.23 1,830.92 -12.31 -0.67 78.01 Guizhou 2,240 417 1,363.07 1,354.03 -9.04 -0.66 81.38 Yunnan 2,397 399 1,438.34 1,421.34 -17.00 -1.18 83.35 Tibet 480 143 1,309.46 1,307.41 -2.05 -0.16 70.21 Shaanxi 2,217 446 1,456.48 1,442.09 -14.39 -0.99 79.88 Gansu 1,800 479 1,357.28 1,350.34 -6.95 -0.51 73.39 Qinghai 600 135 1,466.67 1,452.61 -14.06 -0.96 77.50 Ningxia 600 108 1,754.15 1,729.05 -25.11 -1.43 82.00 Xinjiang 1,495 312 1,471.11 1,447.57 -23.55 -1.60 79.13 Rural total 67,515 16,272 2,257.15 2,239.08 -18.07 -0.80 75.90 Source: Authors' computations based on data from Ianchovichina and Martin 2002 and China National Bureau of Statistics 1999 Rural Household Survey and 1999 Urban Household Survey. The ordering of provinces is the traditional administrative ordering as used in (for example) NBS 2000. Welfare Impacts of China's Accession to the WTO 281 TABLE 15.A.2 Urban Gains and Losses by Province, 2001-07 Number of Original Post-WTO Share of Sampled Number of Income Income Gain or Change Losers Province Households Gainers (Yuan) (Yuan) Loss (Yuan) (Percent) (Percent) Beijing 500 430 9,388.88 9,431.72 42.84 0.46 14.00 Tianjin 500 451 7,323.57 7,358.47 34.91 0.48 9.80 Hebei 650 591 5,673.46 5,702.35 28.89 0.51 9.08 Shanxi 650 598 4,519.20 4,549.94 30.74 0.68 8.00 Inner Mongolia 550 495 4,491.87 4,516.19 24.32 0.54 10.00 Liaoning 1000 916 5,257.42 5,285.65 28.23 0.54 8.40 Jilin 700 610 4,630.13 4,650.46 20.33 0.44 12.86 Heilongjiang 1000 887 4,798.92 4,820.50 21.58 0.45 11.30 Shanghai 500 458 10,927.18 10,984.16 56.98 0.52 8.40 Jiangsu 800 723 6,933.07 6,968.78 35.71 0.51 9.63 Zhejiang 550 498 9,044.40 9,098.28 53.87 0.60 9.45 Anhui 500 458 5,159.46 5,190.37 30.91 0.60 8.40 Fujian 550 516 7,521.52 7,569.70 48.18 0.64 6.18 Jiangxi 550 498 4,762.78 4,783.38 20.60 0.43 9.45 Shandong 650 602 5,689.90 5,720.69 30.78 0.54 7.38 Henan 600 565 4,689.43 4,717.89 28.46 0.61 5.83 Hubei 750 619 5,743.18 5,765.29 22.11 0.38 17.47 Hunan 700 612 5,727.42 5,750.43 23.00 0.40 12.57 Guangdong 600 490 10,871.06 10,903.85 32.79 0.30 18.33 Guangxi 600 496 6,011.10 6,033.40 22.30 0.37 17.33 Hainan 200 172 5,766.33 5,787.64 21.31 0.37 14.00 Chongqing 300 239 5,910.18 5,931.90 21.72 0.37 20.33 Sichuan 800 691 5,610.29 5,634.60 24.30 0.43 13.63 Guizhou 450 383 5,324.43 5,347.71 23.27 0.44 14.89 Yunnan 650 566 5,939.69 5,973.23 33.54 0.56 12.92 Tibet n.a. Shaanxi 500 427 4,768.99 4,788.25 19.26 0.40 14.60 Gansu 400 372 4,610.86 4,641.27 30.41 0.66 7.00 Qinghai 250 240 3,759.53 3,788.65 29.12 0.77 4.00 Ningxia 200 177 4,472.43 4,493.27 20.84 0.47 11.50 Xinjiang 250 214 5,277.25 5,295.94 18.69 0.35 14.40 Urban total 16,900 14,994 6,046.13 6,075.60 29.45 0.49 11.28 Source: Authors' computations based on data from Ianchovichina and Martin 2002 and China National Bureau of Statistics 1999 Rural Household Survey and 1999 Urban Household Survey. The ordering of provinces is the traditional administrative ordering as used in (for example) NBS 2000. Notes Decaluwe and Martens (1988) and Hertel (1997). For a useful overview of alternative approaches to assessing the welfare 1. Jalan and Ravallion (1998) report evidence of such churn- impacts of trade policies and examples from the literature, see ing using panel data for rural China. Baulch and Hoddinott McCulloch, Winters, and Cirera (2001). (2000) review evidence for a number of countries. 4. The only known example of this full integration is 2. For China's lagging poor areas see Jalan and Ravallion Cockburn (2002), who built a classic trade-focused CGE model (2002). onto the Nepal Living Standards Survey covering about 3,000 3. Examples of partial equilibrium analysis of the welfare households. distributional effects of price changes include King (1983), 5. In an antecedent to the approach taken here, Bourguignon, Deaton (1989), Ravallion and van de Walle (1991), and Robilliard, and Robinson (2003) also take price changes gener- Friedman and Levinsohn (2002). On applications to tax policy ated by a CGE model to survey data (for Indonesia). Method- reform, see Newbery and Stern (1987). On CGE models see ologically, the main difference is that Bourguignon, Robilliard, 282 China and the WTO and Robinson generate income impacts at the household level Cockburn, John. 2002. "Trade Liberalization and Poverty in from a microeconometric model of income determination, Nepal: A Computable General Equilibrium Micro Simula- whereas this study derives first-order welfare impacts analytically tion Analysis." Processed. Quebec: University of Laval. from a standard competitive farm-household model. Deaton, Angus. 1989. "Rice Prices and Income Distribution in 6. Hertel (1997) contains descriptions of the standard GTAP Thailand: A Non-Parametric Analysis." Economic Journal 99: model with applications. 1­37. 7. A full discussion of the assumptions of the general equilib- Decaluwe, B., and A. Martens. 1988."CGE Modeling and Devel- rium model and the results of its application to China's accession oping Economies: A Concise Empirical Survey of 73 Appli- to the WTO can be found in Ianchovichina and Martin (2002). cations to 26 Countries." Journal of Policy Modeling 10(4): 8. For food items, unit values can be calculated (expenditure 529­68. divided by quantity) from the survey data, but there is no such Dollar, David, and Aart Kraay. 2002. "Growth is Good for the option for food inputs to production, nonfood commodities Poor." Journal of Economic Growth 7(3): 195­225. consumed or used in production, or wages (the survey data do Edwards, Sebastian. 1997. "Trade Policy, Growth and Income not include labor supplies or quantities consumed of nonfood Distribution." American Economic Review 87(2): 205­10. goods, including production inputs). Friedman, Jed, and James Levinsohn. 2002. "The Distributional 9. Examples of this approach can be found in King (1983) Impacts of Indonesia's Financial Crisis on Household and Ravallion and van de Walle (1991). Welfare: A `Rapid Response' Methodology." World Bank 10. The full sample of the urban survey was about 40,000 Economic Review 16(3): 397­424. households, but until 2002 the central NBS office kept individ- Hertel, T., ed. 1997. Global Trade Analysis: Modeling and ual record data for only 16,900 households. Applications. Cambridge: Cambridge University Press. 11. For further discussion in the context of the Rural House- [www.gtap. org.] hold Survey, see Chen and Ravallion (1996). Ianchovichina, Elena, and Will Martin. 2002."Economic Impacts 12. By the International Labor Organization's definitions, of China's Accession to the WTO." Development Research skilled labor consists of managers and administrators, profes- Group, World Bank, Washington, D.C. sionals, and para-professionals, while unskilled labor consists of Jalan, Jyotsna, and Martin Ravallion. 1998."Transient Poverty in tradespeople, clerks, salespeople and personal service workers, Post-Reform Rural China." Journal of Comparative Econo- plant and machine operators and drivers, laborers, and related mics 26: 338­57. workers and farm workers. ------. 2002. "Geographic Poverty Traps? A Micro Model of Consumption Growth in Rural China." Journal of Applied Econometrics 17(4): 329­46. King, Mervyn A. 1983. "Welfare Analysis of Tax Reforms Using Household Level Data." Journal of Public Economics 21: References 183­214. The word processed describes informally reproduced works that Lundberg, Mattias, and Lyn Squire. 2003. "The Simultaneous may not be commonly available through libraries. Evolution of Growth and Inequality." Economic Journal 113: Barro, Robert. 2000."Inequality and Growth in a Panel of Coun- 326­44. tries." Journal of Economic Growth 5: 5­32. McCulloch, Neil, L. Alan Winters, and Xavier Cirera. 2001. Trade Baulch, Bob, and John Hoddinott. 2000. "Economic Mobility Liberalization and Poverty: A Handbook. London: Center for and Poverty Dynamics in Developing Countries." Journal of Economic Policy Research and Department for Interna- Development Studies 36(6): 1­24. tional Development. Bourguignon, Francois, and C. Morisson. 1990. "Income NBS (National Bureau of Statistics). 2000. China Statistical Year- Distribution, Development and Foreign Trade." European book. Beijing: China Statistics Press. Economic Review 34: 1113­32. Newbery, David, and Nicholas Stern, eds., 1987. The Theory Bourguignon, Francois, Anne-Sophie Robilliard, and Sherman of Taxation for Developing Countries. Oxford: Oxford Robinson. 2003. "Representative versus Real Households in University Press. the Macro-Economic Modeling of Inequality." Working Ravallion, Martin. 1990. "Rural Welfare Effects of Food Price Paper 2003-05. DELTA, Paris. Changes with Induced Wage Responses: Theory and Chen, Shaohua, and Martin Ravallion. 1996."Data in Transition: Evidence for Bangladesh." Oxford Economic Papers 42: Assessing Rural Living Standards in Southern China." China 574­85. Economic Review 7: 23­56. ------. 2001. "Growth, Inequality and Poverty: Looking ------. 2001. "How Did the World's Poor Fare in the 1990s?" Beyond Averages." World Development 29(11): 1803­15. Review of Income and Wealth 47(3): 283­300. Ravallion, Martin, and Dominique van de Walle. 1991. "The ------. 2003."Household Welfare Impacts of China's Accession Impact of Food Pricing Reforms on Poverty: A Demand to the World Trade Organization," Policy Research Working Consistent Welfare Analysis for Indonesia." Journal of Policy Paper 3040, World Bank, Washington, D.C. Modeling 13: 281­300. 16 Implications of WTO Accession for Poverty in China Thomas W. Hertel, Fan Zhai, and Zhi Wang After a decade and half of negotiations, the World likely to face substantial adjustments, because Trade Organization (WTO) successfully concluded tariffs remain high and the scale of production is negotiations on China's terms of entry to theWTO at far below international standards to date (François, the end of 2001. China's accession is expected to lend 2002). In addition, the imports of land-intensive a further boost to trade and foreign investment. agricultural products, including corn and soybeans, China's foreign trade quadrupled from 1990 to are expected to surge under the accession agree- 2000--growing much faster than world trade, which ment. The services sector is also expected to face only increased 87 percent during the same period. substantial expansion and restructuring. Such a By the year 2001,the volume of China's merchandise broad structural adjustment is likely to have an exports and imports had reached US$262 billion1 effect on the distribution of income in China, and US$244 billion, respectively, ranking it sixth in potentially depressing rural incomes in some areas world trade. China is also the largest recipient of relative to urban incomes. Similarly, it is anticipated foreign direct investment (FDI) among developing that coastal regions might benefit at the expense of countries; over US$320 billion flowed into its econ- the interior. Yet the rural­urban, interior­coastal omy during the last decade.Most studies predict that income disparities in China are already at record WTO accession will bring with it added benefits, levels (Yang and Zhou 1999). both for China as well as for most of its trading part- Kanbur and Zhang (2001) track regional inequal- ners (Ianchovichina, McDougall, and Hertel 1999; ity over the past 50 years and find that there are McKibbin and Tang, 2000; Wang 2001). three peaks over this five-decade period: the Great At the sectoral level, however, the effects of Famine of the 1950s, the Cultural Revolution of the accession are likely to be quite different across late 1960s and early 1970s, and the period of open- economies. China's trade surplus in apparel prod- ness and global integration of the 1990s. They also ucts is expected to increase even more with elimi- find that the interior­coastal disparities have espe- nation of the Multi-Fiber Arrangement (MFA) cially contributed to the recent growth in inequal- quotas under the Uruguay Round's Agreement ity as the Chinese economy has opened up. Chen on Textiles and Clothing, or ATC (Development and Wang (2001) attempt to discern the separate Research Center 1998; Wang 1999; Walmsley and impacts of growth and inequality on poverty in Hertel 2000). By contrast, the automobile sector is China since 1990. They find that the increase in 283 284 China and the WTO inequality has contributed to greater poverty over Bureau of Statistics aggregations for 2000 of the rural the past decade. However, the beneficial impact of and urban household surveys for three provinces: overall economic growth has dominated, and the Guangdong, Sichuan, and Liaoning. Together, these absolute number of people in poverty (by several provinces are fairly representative of the diversity different measures) has been reduced. within China as a whole. Guangdong represents the For China's WTO accession, the same two forces relatively wealthy coastal region, which is heavily are at work. On the one hand, further trade liberal- export-oriented; Sichuan represents the populous, ization will accelerate the country's integration into relatively poor inland region in which agriculture the world economy and enhance economic effi- plays a more important role in the economy; and ciency (Yang and Huang 1997; Wang and Zhai Liaoning has a typical "old industrial base," which is 1998). On the other hand, the experience of the last heavily urban and highly dependent on state-owned decade suggests that such liberalization might also enterprises. The latter are mainly concentrated in increase inequality (Kanbur and Zhang 2001). The highly protected, heavy industrial sectors. goal of this chapter is to shed more light on this Within these surveys, we have aggregated house- empirical question. Our approach is model-based. holds into several groups or "strata," according to We use an applied general equilibrium model their primary source of income. Recent analysis designed to capture the differential effects of acces- of trade and poverty by Hertel and others (2004) sion on each of China's economic sectors. Special suggests the merit of distinguishing those house- attention is paid to the disaggregation of house- holds that are specialized--that is, they receive holds in order to permit us to say something about 95 percent or more of their income from one source: the likely impacts of accession on income distribu- transfer payments, labor wages and salaries, or tion. In addition, we pay quite a bit of attention to self-employment income. Because of the limited the modeling of China's rural labor markets, which number of specialized, nonfarm households in the we see as the primary vehicle for poverty reduction survey, as well as the absence of transfer- and labor- in the wake of WTO accession. specialized households in the rural surveys, we end This chapter is organized as follows. The next up with two strata in the rural survey--agriculture- section provides an empirical overview of the specialized and diversified (all other)--and three Chinese economy--with a special emphasis on strata in the urban survey--transfer-specialized, trade and poverty. This overview sets the stage for labor-specialized, and diversified. Within each the computable general equilibrium (CGE) model stratum, we order households from poorest to rich- described in the section that follows. We then turn est, based on per capita income, and then group to a discussion of the baseline scenario, against them into 20 vingtiles, each containing 5 percent which China's accession will be evaluated, along of the stratum population. We then aggregate the with a description of the accession scenario itself. three rural surveys, and the three urban surveys, The next section uses simulation results to assess before incorporating them into the national social the impact of China's WTO accession on poverty, accounting matrix. In this way, we obtain a dis- and the final section offers some conclusions and aggregated representation of households in the suggestions for future research. national model, wherein the earnings and spending profiles reflect the diversity obtained from the three Overview of the Chinese Economy, provincial surveys. Poverty, and Trade Table 16.1 reports per capita income by stratum, location, and vingtile for the 100 representative This section provides an overview of the Chinese households in the model. Using the $1/day level economy. We start by looking at some stylized facts of poverty, Chen and Wang (2001) estimate that of household income and poverty and examining 24.9 percent of the rural population in China patterns of production and trade. and 0.5 percent of the urban population are in poverty. Using these figures as a guide, in table 16.1 Households and Poverty we place the rural poverty line in the fifth vingtile To facilitate analysis of the impacts of WTO acces- of the rural population and the urban poverty sion on poverty, we obtained from China's National line in first vingtile of the urban population. The Implications of WTO Accession for Poverty in China 285 TABLE 16.1 Per Capita Income by Location, Stratum, and Vingtile (1997 yuan) Vingtile Rural Households Urban Households (Poorest 1) Agric. Diverse Total Transfer Labor Diverse Total 1 845 889 874 2,903 2,135 2,454 2,351 2 1,049 998 1,006 3,995 3,151 3,054 3,212 3 1,156 1,162 1,161 4,674 3,790 3,703 3,827 4 1,303 1,301 1,301 5,273 3,987 4,250 4,216 5 1,433 1,432 1,432 5,595 4,513 4,452 4,608 6 1,755 1,551 1,568 6,280 4,763 4,528 4,769 7 1,675 1,679 1,678 6,594 5,237 4,884 5,155 8 1,822 1,811 1,812 7,794 5,692 5,370 5,588 9 1,947 1,944 1,944 8,643 6,096 5,786 6,045 10 2,099 2,095 2,096 8,142 6,694 6,334 6,564 11 2,240 2,252 2,251 8,220 6,866 6,482 6,718 12 2,415 2,411 2,411 8,946 7,420 6,901 7,192 13 2,602 2,595 2,595 10,807 7,686 7,532 7,671 14 2,835 2,818 2,819 12,973 8,432 7,974 8,303 15 3,031 3,069 3,066 10,601 9,120 8,526 8,799 16 3,344 3,353 3,352 12,925 9,709 8,727 9,130 17 3,708 3,717 3,717 -- 11,152 9,659 10,240 18 4,306 4,258 4,261 18,821 12,749 10,985 11,796 19 5,171 5,162 5,163 15,190 15,134 13,403 14,125 20 9,712 8,345 8,485 -- 21,997 19,659 20,522 Share of 7.35 62.73 70.08 1.44 12.01 16.47 29.92 population (%) Not available. Source: National Bureau of Statistics Rural and Urban Household Surveys 2000. average income in each vingtile is much higher for ruralpopulation,theaverageeducationalattainment the urban population, and even after adjusting of agriculture-specified households is lower than for differences in cost of living, a far greater per- that of other households. There is also a significant centage of the rural population falls below the gap in educational attainment between rural and poverty line. urban households. Thus the most vulnerable farm This preliminary analysis reveals that any policy households are likely to be those that are trapped that depresses incomes in agriculture is very likely in agricultural activity by virtue of their lack of to lead to a short-run increase in poverty. If oppor- education. tunities arise elsewhere in the rural economy, how- The impact of WTO accession on rural non- ever, it is possible that, after suitable adjustment, farm households and especially urban households many of these rural households could be made depends importantly on what happens to the man- better off. Such an adjustment will likely be most ufacturing and services sectors. Here, we expect a difficult for those households that currently do not mixed outcome, with profitability and employment participate in the nonfarm economy, either as in light manufactures rising, but the heavy industry workers or in a self-employed capacity. A critical and automobile sectors will be adversely affected. determinant of farm households' ability to partici- Households dependent on transfer payments are pate in the nonfarm economy will be their level also potentially vulnerable after WTO accession, of educational attainment. In general, the higher- because reduced tariffs may bring additional pres- income vingtiles in both rural and urban areas have sure to bear on state-run enterprises and on gov- a higher level of educational attainment. Within the ernment spending--and thus transfer payments. 286 China and the WTO Finally, by virtue of the way they have been defined, may be realized by means of WTO accession diversified households are likely to be little affected as the rules governing foreign investment in the by trade liberalization. Decreases in one source of services sector are liberalized. In contrast to serv- income will tend to be offset by increases elsewhere. ices, manufacturing's share of value-added is quite The fate of the labor-specialized households will substantial--33 percent versus a global average of depend on what happens to wages after accession. 19 percent based on GTAP data--placing China If educational attainment is used as a proxy for skill second, behind Thailand. Textiles, building materi- level, then the poor are most heavily reliant on un- als, and chemicals lead the way, followed by sectors skilled labor. Therefore, from a poverty perspective, related to metals, machinery, electronics, and asso- the key factor for labor-specialized households ciated products. is what happens to wages for the unskilled after An important factor in the growth of manufac- WTO accession. What are the likely factor market turing activity in the past decade has been the rapid effects of China's WTO accession? These effects will expansion of exports. Textiles, apparel, and leather depend on the pattern of production, trade, and products now account for more than one-quarter of protection in China, or the subject to which we total Chinese exports, and electronics account for next turn. another 11.5 percent. The export dependency of these sectors, as well as for crafts and toys and instruments, is high, with more than 30 percent of China's Production, Trade, and Protection their output directed to foreign markets. (Textiles Having provided an overview of the Chinese econ- are important intermediate inputs, and so the omy from the household perspective, we now offer export dependency ratio is somewhat less--about a perspective from the production point of view. 18 percent.) Much of this export growth has been table 16.2 has been assembled using the 1997 fueled by the export processing or export promo- Chinese social accounting matrix developed by the tion regime introduced in 1986 (Naughton 1996). Development Research Center (DRC) at the State The enterprises operated under this regime have Council. We used the most recent input­output attracted a great deal of foreign investment as a table. One of the most striking points about result of special preferences--most important, the table 16.2 is the very high share of employment in exemption from the tariffs on imported intermedi- agriculture. This sector accounted for more than ate inputs to be used in the production of exports. 55 percent of total employment in the Chinese By 1997 the export processing regime had economy in 1997, yet its contribution to economy- grown to account for 50 percent of China's total wide value-added (first column) is only about exports and 46 percent of its total imports. Nearly 22 percent. This finding is consistent with the 80 percent of China's primary iron and steel, low income levels observed in table 16.1 for electric machinery, electronics, and instrument agriculture-specialized households. Because this exports, and more than two-thirds of its leather, sector also relies almost entirely on unskilled labor, toys, and chemical fiber exports were "processing a reduction in the size of this sector will likely exports"--that is, assembling or transformation release additional unskilled labor into the rural-- of imported intermediate goods and re-export. and possibly urban--labor markets, thereby Export processing also constitutes a significant depressing wages. portion of China's exports of apparel, paper, and Table 16.2 also reveals a relatively underdevel- printing products. The high shares of processing oped services sector. For example, when compared exports in these sectors require a large volume of with the 65 other regions in the Global Trade Analy- raw materials, components, and semi-processed sis Project (GTAP) Version 5 database (Dimaranan products imported from abroad. Column 8 in and McDougall 2002), China ranks last in the share table 16.2 shows that in textiles, apparel, and of economy-wide value-added generated in the leather, ordinary imports constituted only 1­2 per- services sector, and this share is scarcely half the cent of total imports in 1997, and so almost all world average: 37 percent versus 72 percent. The imports were used for the production of process- share of services in total employment in China ing exports. Most imports of paper products, build- is even smaller, amounting to less than 27 percent ing materials, chemicals, basic metal and metal in 1997. This finding indicates the potential that products, machinery, and electronics were also Implications of WTO Accession for Poverty in China 287 (%) 0.5 0.2 0.0 1.0 4.1 1.0 3.4 0.2 9.6 5.3 0.0 0.0 0.0 1.2 9.0 1.9 9.9 9.1 3.7 2.5 0.8 5.1 4.7 4.0 3.8 (13) Post- ariffT 11.5 15.0 10.4 13.7 11.0 14.3 accession Rate (12) As cent 0.0 4.7 0.7 1.7 0.8 6.5 (%) 40.0 20.0 17.1 29.0 42.0 62.5 10.6 55.0 66.7 -- -- 31.0 31.2 93.6 31.7 15.9 39.9 21.6 17.4 28.2 32.3 55.2 27.8 66.1 100.0 Collected Nominal Per (%) (11) ariffT 0.4 0.2 0.0 0.6 5.7 8.3 0.7 2.1 0.5 1.7 3.3 1.0 0.0 0.0 0.9 5.3 4.4 9.5 3.7 0.2 0.7 0.3 2.5 3.1 1.0 4.8 3.0 7.2 1.0 24.0 10.6 Collected Rate (%) (10) ariffT 1.0 1.0 1.0 3.5 5.7 5.0 0.8 6.0 1.5 0.0 0.0 2.9 4.7 3.1 8.7 28.6 15.0 16.0 17.0 30.0 23.2 60.2 49.1 27.5 41.8 35.5 14.4 11.0 10.8 10.9 15.5 Nominal Rate of (9) Net Exports an)uY 0.0 6.3 7.7 2.6 3.8 6.8 2.5 1.5 8.6 4.9 0.5 7.9 0.8 7.5 3.3 8.5 11.3 12.8 20.6 14.4 10.3 42.4 61.4 22.8 29.2 87.6 19.6 77.8 17.4 104.5 156.2 (Billions y 6 1 2 1 7 (8) rdinar otalT (%) 92 87 21 17 25 33 47 77 12 63 71 43 37 36 46 94 36 18 45 24 20 33 28 62 31 75 O Imports/ Imports y -- (7) otalT 96 97 92 88 92 79 19 76 12 72 84 96 60 45 29 61 35 32 69 80 84 31 (%) 100 100 100 100 100 100 100 100 Ordinar Exports/ Exports (6) otalT (%) 0.9 0.0 7.9 0.0 2.3 3.6 8.9 0.9 1.2 1.2 3.1 0.0 1.1 4.3 4.3 1.6 3.3 9.7 3.1 3.2 2.0 5.7 8.3 5.9 6.6 14.4 18.4 37.0 32.6 13.1 40.0 Exports/ Outputs 1997 y (%) (5) 0.88 6.93 0.32 2.96 0.22 3.81 0.07 0.19 0.00 0.26 3.06 2.94 8.41 4.08 3.07 0.74 0.32 1.43 0.41 0.17 0.35 2.47 4.88 3.86 8.48 1.53 5.78 29.65 19.07 16.31 11.08 Ordinar Imports/ China, Domestic Use in (4) (%) 0.1 0.0 0.4 0.0 1.1 0.2 0.0 0.4 0.2 0.1 0.6 1.5 0.0 0.1 0.6 0.3 0.3 0.1 3.3 0.5 0.3 9.9 5.0 2.1 0.5 6.0 1.2 4.2 0.7 0.5 11.4 Exports Openness (3) (%) 0.2 0.9 0.1 0.6 0.5 0.5 0.3 0.1 0.0 0.0 0.1 3.6 1.1 0.5 0.8 1.2 1.1 0.2 1.0 0.1 0.2 6.8 0.7 1.8 0.9 2.9 1.0 3.1 0.2 2.1 11.8 Imports Market (2) (%) 7.5 4.1 2.6 2.0 2.0 0.0 8.2 3.2 2.2 1.2 0.2 0.1 0.2 0.8 0.1 0.2 0.1 0.6 0.3 0.1 1.5 0.6 0.3 0.4 0.6 0.3 0.2 1.0 0.2 0.1 23.8 and Employment (1) alue-V (%) 2.32 1.27 0.82 0.62 7.35 0.90 0.03 5.60 2.07 0.95 1.68 1.67 0.14 0.42 1.00 0.28 0.78 0.05 1.72 0.76 0.28 3.25 1.88 0.64 0.80 1.45 0.88 0.61 2.52 0.74 0.33 Added Structure gas crops mining Economic natural forage furniture mining ore and oil and food printing toys refining and fibers ore 16.2 nongrain y livestock agriculture oil mining mining mill and and ool Rice Wheat Corn Cotton Other Forestr W Other Fishing Other Coal Crude Ferrous Nonferrous Other egetableV Grain Sugar Processed Beverages obaccoT extilesT Apparel Leather Sawmills Paper Crafts Petroleum Chemicals Pharmaceuticals Chemical ABLET 288 China and the WTO (%) 7.2 9.3 4.2 2.9 5.6 6.2 5.0 3.3 6.7 1.6 4.0 0.0 0.0 0.0 0.1 0.0 0.0 0.0 1.7 0.0 4.0 (13) Post- ariffT 11.2 12.0 accession Rate corn, . only as cent 2.3 (12) (%) 10.1 17.3 24.7 12.7 18.3 30.7 18.4 64.3 23.2 17.3 17.8 18.4 -- -- -- -- -- -- -- -- -- 22.3 wheat, Collected Nominal Per rice, chandise of (%) mer (11) ariffT 2.0 3.6 2.0 0.9 2.4 4.2 2.6 1.3 3.1 2.1 2.3 0.9 0.0 0.0 0.0 0.1 0.0 0.0 0.0 1.7 0.0 2.7 32.6 for Collected Rate imports is The rate (%) (10) ariffT 8.1 7.1 5.6 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 19.8 20.8 13.1 13.7 14.1 50.7 17.9 11.8 12.5 38.9 11.2 tariff Nominal Rate price. of world average (9) Net Exports an)uY 7.6 5.8 9.6 3.8 2.6 8.7 4.2 2.7 1.7 1.5 at 42.8 23.9 17.5 13.1 36.4 41.0 82.9 42.2 25.2 12.3 38.9 38.8 409.5 The are (Billions tem. y sys exports 3 (8) rdinar otalT (%) 12 20 28 14 20 35 21 73 28 20 02 20 -- 32 100 100 100 100 100 100 100 100 and O Imports/ Imports y harmonized 9 imports (7) otalT 28 82 21 56 57 61 63 63 24 21 15 29 51 (%) 100 100 100 100 100 100 100 100 100 of the Ordinar Exports/ Exports of share level (6) otalT (%) 3.4 5.5 8.0 5.9 5.6 1.9 9.7 8.1 0.9 0.1 9.4 5.7 0.9 0.5 0.7 0.1 7.7 15.7 13.1 15.9 36.3 49.5 10.1 Exports/ Outputs sectoral y The eight-digit Council. (%) the (5) 1.49 0.60 4.12 6.63 4.36 4.11 9.43 4.69 0.16 0.00 0.28 1.83 1.27 1.31 1.22 5.17 0.44 0.45 3.59 11.16 23.65 13.27 24.51 price. at Ordinar Imports/ Domestic Use State, 1997 Center (4) (%) 4.2 2.1 1.8 1.2 4.1 1.8 1.1 0.4 1.6 5.5 2.7 0.9 0.2 0.1 2.8 0.7 0.7 0.1 4.5 0.3 0.0 for 11.5 100 domestic ch Exports at 1993­97. are of imports Resear y (3) (%) 2.1 0.8 3.8 2.6 2.6 5.6 8.0 1.1 2.6 4.0 2.7 0.4 0.0 0.4 0.7 0.2 1.2 0.4 2.9 0.2 0.2 13.2 100 Imports average ordinar the of exports/output are Development (2) (%) 0.6 2.1 0.7 0.3 0.7 1.0 0.6 0.4 0.4 0.6 0.4 0.2 0.6 0.5 5.8 3.1 0.3 9.0 0.5 1.6 4.0 1.8 100 oil total weight 1997, Employment and the on use vegetable matrix, (1) alue-V (%) 1.24 3.54 1.36 0.49 1.48 2.36 1.31 1.09 0.96 1.56 1.63 0.36 1.33 2.45 7.03 3.91 1.61 7.47 2.09 5.17 4.69 3.06 Added 100 and based forage, accounting calculated steel equipment y and vices imports/domestic (Continued) health social y are plastics and metals ser mill and and materials y rates iron communication vices equipment machiner cial grain 16.2 y Chinese products transport manufacturing ser administration available. Ordinar and tariff ce: Not Rubber Building Primar Nonferrous Metal Machiner Special Automobiles Other Electric Electronics Instruments Other Utilities Construction ransportationT Post Commer Finance Social Education Public otal/averageT -- Notes: cotton, The Sour ABLET Implications of WTO Accession for Poverty in China 289 used by foreign or joint venture companies as inter- in the six years after China's WTO accession. The mediate inputs to produce processing exports. sectoral distribution of tariff cuts is relatively even. The dominance of duty-exempt imports is also For most sectors, the tariff cuts range from 50 per- evident when one compares the nominal tariff cent to 80 percent. Some light manufacturing rate with the collected rate reported in table 16.2 sectors and capital goods sectors that historically (smuggling also plays a role here). The nominal have been highly protected--such as beverages, tariff for textile products is 27.5 percent, but the tobacco, automobiles, and electronics--will experi- collected tariff averages 0.2 percent. Similarly, dra- ence substantial tariff cuts. The dominant role matic differences appear in the apparel and leather played by the export processing sector will be very sectors as well as many of the other export-oriented important in determining the impact of these tariff activities. The lowest nominal/actual ratios appear cuts. In sectors such as textiles, where virtually all in the pharmaceutical and automobile industries imports come in under the export processing pro- where the nominal rate is only about twice as large visions, the impact of further tariff cuts on the cost as the actual rate. These sectors have relatively low of intermediate inputs will be negligible. The export/output ratios. importation of consumer goods,as well as imported Table 16.2 also reports the ratio of ordinary intermediates for domestic use, also will be affected imports to domestic use (column 5) as an indica- by the WTO accession tariff cuts. tion of overall import dependency. The sectors with To the extent that accession leads to a decline in the largest dependency on ordinary imports are agricultural, relative to nonagricultural, profitabil- energy and mining, machinery, special equipment, ity, we expect to see a rise in rural poverty and and electronics, all of which show import shares in increased pressure on the rural population to mi- excess of 10 percent of China's domestic use of grate to urban areas. However, a definitive answer to these commodities. the question posed by this chapter--Will China's An overall picture of China's trade is offered accession to the WTO increase poverty--and if so, by the sector-specific net exports also reported in among which groups?--requires use of a formal table 16.2 (value of exports minus value of imports model. We therefore turn next to a summary of the in billions of yuan). By and large, these exports single-country general equilibrium model used in reflect China's comparative advantage. China is a this paper. net exporter of unskilled labor-intensive manu- factures and a net importer of skilled labor and Description of the Model capital-intensive manufactures. The largest share of the trade surplus in China is attributable to apparel The CGE model of China used in this study is the and textiles. Within the agricultural sector, China is latest in a long line of models developed and main- a net importer of grains, but it has a trade surplus tained by the Development Research Center of the in other agricultural products. State Council in Beijing. Earlier versions have been China's nominal tariff structure is typical of that used to analyze the economy-wide implications of many developing countries in that it provides of China's accession to the WTO (Development high protection for the manufacturing sector, espe- Research Center 1998), the consequences that cially capital-intensive manufactures and final con- accession might have for urban unemployment sumption goods. Automobile imports face one of (Zhai and Wang 2002), and the income distribution the highest nominal tariff rates (50 percent), and consequences of trade and tax reform (Wang and the actual collection rate is also quite high (33 per- Zhai 1998). The CGE model has its intellectual cent). The tariff rates in other manufactures and in roots in the group of single-country, applied gen- the textile and apparel sector are also relatively eral equilibrium models used over the past two high, but their effects are limited because the share decades to analyze the impacts of trade policy of duty-paying imports (ordinary imports) is very reforms (Dervis, de Melo, and Robinson 1982; de small. Melo 1988; Shoven and Whalley 1992; de Melo and The final column in table 16.2 reports China's Tarr 1992). The China model began as a prototype postaccession tariff rates, by sector. The trade- CGE model developed for the Trade and Environ- weightedtariff ratewilldropbymorethan60percent ment Program of the Organisation for Economic 290 China and the WTO Co-operation and Development (OECD) Develop- Households consume goods and services ment Center (Beghin and others 1994). Since that according to a preference structure determined by time, however, significant modifications have been the extended linear expenditure system. Through made to capture the major features of the trading specification of a subsistence quantity of each good and tax system of the Chinese economy (Wang and or service, this expenditure function generates non- Zhai 1998; Zhai and Li 2000), to add a full-fledged homothetic demands, whereby the larger the rela- demographic module (Zhai and Wang 2002), and tive importance of subsistence consumption (e.g., to introduce disaggregated households and an importance would be high for rice and low for improved treatment of labor markets (this study). automobiles), the more income-inelastic is the Here, we focus on the main features of the model, household's demand for that good. especially those relevant for assessing the conse- The other important dimension of household quences of China's WTO accession for poverty. behavior is the supply of labor to off-farm activi- ties. Whereas in developed economies this supply is typically modeled as a function of the relative Modeling Household Behavior wages in the farm and nonfarm sectors, in China To come to grips with the poverty question, it is the off-farm labor supply decision is complicated critical that we disaggregate households to the by institutional factors that have been built into the maximum extent possible, subject to the limita- system to keep the agricultural population in place tions posed by survey sampling, computational (Zhao 1999b).For example,the Chinese government constraints, and human capacity for analysis. It is has sought to make it costly for people to leave particularly important to disaggregate households rural areas by tying incomes to daily participation along those dimensions most likely to be affected in collective work. More recently, the absence of by WTO accession. Thus, for example, one would well-defined land tenure has raised the opportunity not want to group together rural and urban house- cost of leaving the farm. Because households that holds,or agriculture-and-nonagriculture specialized cease to farm their land may lose their rights to it, households. Otherwise, one would be introducing they have a strong incentive to continue some level artificial diversification, thereby blurring the possi- of agricultural activity, even when profitability is ble consequences of WTO accession. Accordingly, quite low (Zhao 1999a). Low-skill farm households we have used the grouping of households in table are prevented from moving to urban areas through 16.1: 20 vingtiles 2 rural and 3 urban strata 40 denial of hukou (household registration), a govern- rural and 60 urban representative households. A ment measure discussed in more detail later in this total of 100 household groups are therefore used in chapter. Because the growth in rural, nonfarm the model. activities has been only modest, the factors just Each household is endowed with three types of described seriously limit the ability of households labor: unskilled,semiskilled,and skilled.These three to obtain off-farm work (Zhao 1999b).4 types are distinguished by educational attainment:2 Our approach to modeling the off-farm labor unskilled workers are illiterate or semi-literate, supply is to capitalize on the econometric work of semiskilled workers have a middle or high school Sicular and Zhao (2002). They report results from a education, and skilled workers have schooling be- household labor supply model estimated using labor yond high school.3 Households are also endowed survey data from the 1997 Chinese Health and with profits from family-owned agricultural and Nutrition Survey (CHNS) of nine central provinces. nonagricultural enterprises, property income, and This survey measures the labor supplied by mem- transfers. Agricultural profits represent returns to bers of each household to farm and nonfarm family labor, land, and capital. However, because activities. Sicular and Zhao estimate the implicit the land market is not well developed in China, the (shadow) wage for each individual in the sample if returns to land are effectively part of the return to he or she were to work in agriculture or nonagri- family labor (Yang 1997). If a household migrates cultural self-employment, and they also estimate from the rural area and ceases to farm its land, the the nonagriculture wage that this person could land cannot be sold or rented; Rather, it reverts obtain. They then estimate labor supply equations to the state. for self-employed agricultural labor, self-employed Implications of WTO Accession for Poverty in China 291 nonagricultural labor, and wage labor. From these For our modeling exercise, it is important to equations, it is possible to calculate elasticities obtain an estimate of the wage gap motivating the of labor transfer between the farm and nonfarm temporary migration of workers between the rural sectors. and urban sectors in China. Zhao (1999a) docu- Because of the variety of labor supply elastici- ments an average annual wage gap between rural ties produced in response to the three different and urban work of 2,387.6 yuan for unskilled rural wages, the authors obtain a variety of labor trans- workers of comparable background and ability in fer elasticities, depending on the "thought experi- Sichuan Province in 1995. Much of the wage gap ment" being conducted. These labor transfer stems from the social costs associated with migra- elasticities are asymmetric in their framework. For tion, including the disutility of being away from example, the responsiveness of labor transfers family, poor quality of housing, limited social serv- from agriculture to market wage employment ices for migrants, the danger of being robbed en based on a decline in the shadow wage of labor in route to and from the work location, and the gen- agriculture is 2.67, whereas the elasticity of labor eral uncertainty associated with being a nonregis- transfers in response to an increase in the market tered worker in an urban area (Zhao 1999a, 1999b). wage is only 0.6. Yet this response is symmetric in Although these transaction costs are not observ- our model, for which only relative prices matter, able, they clearly represent a very significant burden and so it is difficult to choose the correct parame- on the migrants and their families. ter for our analysis. For our base case, we adopt If there were no barriers to the movement of the "push" elasticity of labor transfer (2.67) (called labor between rural and urban areas, we would here the high labor transfer elasticity). However, expect real wages to be equalized for an individual for purposes of sensitivity analysis we also report worker with given characteristics. Shi, Sicular, and results from a simulation in which the "pull" elas- Zhao (2002) explore the question of rural­urban ticity of 0.6 is used (called here the low labor inequality in greater detail for nine different transfer elasticity). provinces using the CHNS. The authors conclude that the apparent labor market distortion is about 42 percent of the rural­urban labor income differential Modeling Household Migration and 48 percent of the hourly earnings differential.5 Despite the large income and poverty differential When applied to the average wage differential, this between rural and urban households, China has distortion amounts to an ad valorem "tax" on rural limited permanent migration through a combina- wages of 81 percent.6 tion of direct and indirect measures. First and We model these transaction costs as real costs foremost, households must have an appropriate that are assumed by the temporary migrants, recog- registration (hukou) to reside legally in an urban nizing that these migrants are heterogeneous and area. Without this registration, access to many of that the extent of the burden varies widely. Those the urban amenities, including housing and educa- migrants who are single and live close to the urban tion, is limited and quite expensive. Although area in which they are working are likely to experi- highly skilled persons and investors can purchase a ence minor inconvenience as a result of this tempo- "blue stamp hukou" (Chan and Zhang 1999), this rary migration. We expect them to be the first to avenue is not available to the vast majority of rural migrate, other things being equal, in response to residents. Because of these barriers to moving an higher urban wages. As for migrants who have large entire household to an urban area, rural­urban families and come from a great distance, their migration is largely a transitory phenomenon-- urban living conditions are often very poor, and it and one that is occurring on a massive scale. By is not uncommon for them to be robbed on the most estimates, the number of "floating workers" train when they are returning home from work. For (excluding commuters) exceeds 50 million, or such people, the decision to migrate temporarily is about 10 percent of China's rural labor force. If likely to be a marginal one--and one that they may commuters and mobile self-employed individuals not choose to repeat. With this heterogeneous pop- are included, the figure doubles to about 100 ulation in mind, we postulate a transaction cost million (Rozelle and others 1999). function that is increasing in the proportion of the 292 China and the WTO rural population engaged in temporary work. This high but less than infinite. Therefore, the terms of transaction cost function has a simple, constant trade for China are endogenous in our simulation. elasticity functional form, which begins at the ori- Production in each of the sectors of the econ- gin and reaches the observed wage gap (adjusted omy is modeled using nested constant elasticity of for transport and living costs) at the current level of substitution functions, and constant returns to temporary migration (about 70 million workers). scale are assumed. Sectors differentiate between We assume that further increases in temporary rural and urban labor that substitute imperfectly migration have only a modest impact on these between them. This is an indirect means of build- transaction costs.7 ing into the model a geographic flavor, because In our subsequent analysis, these transaction some sectors will be located largely in urban areas, costs associated with temporary migration will play while others will be predominantly in rural areas. an important role in the unskilled and semiskilled By limiting the substitutability of rural and urban labor market. To the extent that WTO accession labor in each sector, we are able to proxy the eco- increases the demand for unskilled labor in urban nomic effect of geographically distributed activity. areas, urban wages will rise and more rural labor Thus if WTO accession boosts the demand for will be drawn in. However, this supply of rural goods that are predominantly produced in urban labor will come at some cost--both in terms of areas, then urban wages will rise relative to rural higher transport and living costs for the worker, wages, and migration will be encouraged. Ideally, as well as indirect transaction costs--because the we would model the geographic distribution of additional workers are presumably being drawn industrial activity, but unfortunately the data do from a greater distance or from less favorable fam- not exist to support this type of split. ily or social circumstances. All commodity and nonlabor factor markets are assumed to clear through prices. With the excep- tion of the farm­nonfarm labor supply decision, Modeling Production, Exports and Imports labor is assumed to be perfectly mobile across sec- An important characteristic of our CGE model is tors. Capital is assumed to be partially mobile, the explicit treatment of two separate foreign trad- reflecting differences in the marketability of capital ing regimes. One is the export processing regime, goods across sectors. discussed earlier in the context of table 16.2. This The current version of the DRC's CGE model sector receives duty-free imports and is therefore has a simple recursive dynamic structure. Dynam- extremely open, with considerable foreign invest- ics in the model originate from the accumulation ment. The other sector is the ordinary trade regime. of productive factors and productivity changes. Since the 1990s, processing exports have grown The model is benchmarked on China's 1997 social rapidly as a result of their preferential treatment. accounting matrix and is solved for subsequent They now account for more than half of China's years from 1998 to 2007. We turn now to the details total exports. Obviously, any analysis of external of the baseline scenario. trade and the impact of changes in trade policy must have an explicit treatment of this dualistic Base Case Projections and WTO foreign trading regime in the model. Accession Simulation Trade is modeled using the Armington assump- China's WTO accession includes a complex package tion for import demand and a constant elasticity of trade and investment liberalization. In this chap- of transformation for export supply. Thus Chinese ter, however, only four aspects are considered: products are assumed to be differentiated from foreign products, and exports from China are (1) tariff reduction on industrial products treated as products different from those sold on the (2) agricultural trade liberalization (i.e., tariff domestic market. The small-country assumption is reduction for agricultural products), intro- assumed for imports, and so world import prices duction of the tariff-rate quota system for are exogenous in terms of foreign currency. Exports agricultural goods, and elimination of export are demanded according to constant-elasticity subsidies for corn and cotton demand curves, the price elasticities of which are (3) liberalization of direct trade in services Implications of WTO Accession for Poverty in China 293 (4) phaseout of MFA quotas on textile and and cooperating in dispute settlement. Because clothing. China's market accession commitments to WTO entry will be phased in over a seven-year transition Now that China is a member of the WTO, its period, a baseline from 1998 to 2007 is established exports of textiles and apparel products to North using our recursive dynamic model under a set of American and European markets are being sub- assumptions: China will continue its grain self- jected to accelerated MFA quota growth (from 2002 sufficiency policy, and the import quota of agricul- to 2006), similar to that affecting other developing tural goods will grow at 3 percent annually from countries that are WTO members. The remaining 2000 to 2007. This calibrated "benchmark" will quota restriction will be eliminated in 2007, according to the ATC.8 Therefore, the analysis at serve as a basis of comparison for counterfactual simulation conducted in the WTO accession exper- best captures only parts of the issue. It does not take iments. Major results from the baseline calibration into account other major aspects of WTO member- are summarized in table 16.3, and the key assump- ship, such as reducing barriers to foreign invest- tions are outlined in table 16.4. The baseline uses ment, protecting intellectual property rights, the economy-wide total factor productivity (TFP) securing market access, enforcing commitment, TABLE 16.3 Summary of Baseline Calibration 1997 2000 2001 2002 2003 2004 2005 2006 2007 Exogenous specified variables GDP growth rate (%) -- 8.1 7.3 7.6 7.8 7.8 7.8 7.8 7.8 Life expectancy (years) 70.3 71.8 72.0 72.2 73.2 73.4 73.6 73.8 74.0 Total fertility rate Urban 1.40 1.45 1.48 1.5 1.52 1.55 1.57 1.59 1.62 Rural 2.01 1.95 1.97 1.99 2.01 2.02 2.04 2.06 2.07 Calibrated results: Macroeconomic trends Growth rate (%) Total absorption -- 11.02 8.53 8.29 8.45 8.39 8.34 8.29 8.25 Labor force -- 1.23 1.21 1.17 1.15 1.17 1.16 1.15 1.15 Capital stock -- 10.08 10.51 10.45 10.28 10.18 10.08 9.98 9.88 Total Factor Productivity -- 2.87 1.74 2.16 2.43 2.44 2.49 2.53 2.56 Ratio to GDP ( %) Private consumption 44.3 47.0 47.3 47.4 47.4 47.4 47.5 47.6 47.6 Investment 32.7 34.1 35.2 35.9 36.7 37.4 38.1 38.8 39.4 Export 21.7 25.5 25.7 25.7 25.7 25.7 25.7 25.6 25.6 Import 16.3 22.9 23.5 23.7 23.8 24.0 24.1 24.2 24.2 Urban/rural per capita income 3.03 2.77 2.74 2.71 2.68 2.65 2.63 2.60 2.58 ratio Population and labor force Population (millions) 1,236.3 1,269.9 1,280.5 1,290.8 1,301.0 1,311.5 1,321.9 1,332.4 1,343.0 Urban 369.9 397.5 407.2 416.9 426.6 436.5 446.4 456.4 466.5 Rural 866.4 872.4 873.3 873.9 874.3 875.0 875.5 876.0 876.4 Labor force (millions) 640.7 665.6 673.6 681.5 689.4 697.4 705.5 713.6 721.8 Urban 181.1 196.9 202.5 208.2 213.8 219.5 225.3 231.2 237.2 Rural 459.6 468.7 471.1 473.4 475.6 477.9 480.2 482.4 484.6 Grain self-sufficiency rate (%) 99.5 98.5 98.4 98.4 98.3 98.2 98.1 98.0 98.0 Share of rural labor force (%) 71.7 70.4 69.9 69.5 69.0 68.5 68.1 67.6 67.2 Share of agricultural employment 72.0 71.9 71.6 71.6 71.2 71.0 70.9 70.7 70.7 in total rural labor force (%) Note: -- Not available. Source: Authors' calculations. 294 China and the WTO TABLE 16.4 Summary of Experiments Experiment Description Base Base case: -- Real GDP exogenous -- TFP growth rate -- Agricultural sector exogenous -- Other sectors endogenous -- Three percent growth rate of import quota for goods subjected to quantitative restriction (rice, wheat, corn, cotton, wool, vegetable oil, sugar, petroleum refining, automobiles) -- Exogenous export quota growth for textiles (5.7 percent) and apparel (6.0 percent)--annual average -- All tax rates fixed at their base year level -- Balance of payments gradually declines to 40 percent of its base year level in 2007 WTO WTO accession Tariff reduction -- An average 60 percent cut in 2000 tariff level from 2000 to 2007 Agricultural trade liberalization -- Quota restriction (10 billion yuan, 1997) Initial quota in 2000 Annual growth rate of quota (%) Rice 1.073 26.4 Wheat 1.252 10.5 Corn 0.374 18.0 Cotton 1.153 4.5 Wool 0.665 6.5 Vegetable oil 3.706 27.0 Sugar 0.408 7.6 -- Tariff cut for other agricultural goods -- Elimination of export subsidies for corn and cotton Phaseout of MFA -- Acceleration of MFA quota growth rate from 2001 to 2004 -- Zero export tax of textiles and apparel in 2005 Source: Protocol on the Accession of WTO; Authors' assumptions. variable as a residual and adjustment mechanism to also be reduced. Following François (2002), we match the pre-specified real gross domestic prod- model the impact of service sector liberalization as uct (GDP) growth rate under an exogenously spec- halving the barriers to services trade. We also intro- ified life expectation rate and total fertility rate for duce a 20 percent productivity boost for the auto- both the rural and urban population. mobile sector to reflect the efficiency gain from The WTO accession scenario is described in industrial restructuring and realization of economic detail in table 16.4. Import tariffs are reduced gradu- scale in this sector after China WTO accession. ally, and the sectoral reduction rates are aggregated from the Harmonized Commodity Description and Simulation Results Coding System tariff schedules for the period 2002­2007 and weighted by 1997 ordinary trade We now turn to the results of our analysis of WTO data.The quota control system is introduced for rice, accession, beginning with the economy-wide wheat, corn, cotton, wool, vegetable oil, and sugar. effects, then turning to the disaggregated house- Moreover,the tariffs for other agricultural goods will hold impacts. Implications of WTO Accession for Poverty in China 295 Economy-Wide Impacts of WTO Accession Chinese economy. To facilitate analysis, figure 16.1 reports the output changes in descending order, Our results, which are directed at the outcome in omitting the changes that are less than 2 percent in 2007 when China's accession to the WTO is com- absolute value. Clearly, the largest increases in out- plete, are reported in terms of deviations from the put stem from elimination of the quotas on textiles baseline in 2007. The first column of results in and apparel exports to North America and Europe; table 16.5 reports the macroeconomic impacts of these products, as well as the production of syn- accession on China's economy under the high thetic fibers and cotton, increase by substantial transfer elasticity assumption. Real GDP and wel- amounts. On the other end of the spectrum, the fare both increase by 0.65 percent and 0.73 percent, most heavily protected sectors, which have sizable respectively, while consumption increases by over trade exposure, experience declining output. These 1 percent and investment by only half as much. The sectors include wool, automobiles, vegetable oils, reduction in trade barriers gives a substantial boost sugar, machinery, nonferrous metal products, spe- to trade in China, with both exports and imports cial equipment, and wheat. rising by about 15 percent. With the exception of a few mining products As for changes in factor prices, skilled wages rise and transport services for which there is no cut in more than semiskilled wages, which, in turn, rise protection, import volumes increase for all sectors more than unskilled wages. The relatively greater in the economy. The largest increases are for food increase in skilled wages is fueled by the tendency and agricultural products (where there are deep for the manufacturing and service sectors to expand cuts in protection), textiles and apparel (where at the expense of agriculture. The manufacturing the demand for intermediate inputs increases and service sectors are relatively intensive in their strongly), as well as automobiles (where there are use of skilled and semiskilled labor, thereby boost- strong productivity gains from the rationalization ing wages for these factors relative to unskilled of production). Export volumes for most products wages. And then wages in the different markets are also increase, fueled by cheaper inputs and real linked. In particular, urban and rural nonfarm depreciation, as well as the elimination of textile wages are linked through the temporary migration and apparel quotas. Only cotton and corn, for of workers to the urban areas. Specifically, 1.46 mil- which export subsidies are removed, and wool, lion additional temporary migrants move to the which is used heavily in the expanding textile sec- city in response to the higher urban wages. As a tor, experience reductions in export volumes. consequence, the rural wage is once again equated Table 16.6 also reports the change in net exports with the urban wage, less the direct and indirect for each of the product categories. It is not surpris- costs associated with migration. With transaction ing that the largest increase in net exports is in the costs changing only modestly (­0.15 percent to apparel and textile sectors because of elimination of 0.63 percent), the boost in wages is relatively more the quotas under the Agreement on Textiles and important for rural households, whose wages Clothing. The net exports of apparel and textiles increase by a greater proportion because of the increase by 286 billion yuan and 61 billion yuan, (nearly fixed) transaction costs. respectively. Because chemical fibers and cotton are Returns to labor in agriculture are also linked to major intermediate inputs in the textile sector, their the rural nonfarm wages via off-farm migration. As net imports also increase sharply. In addition, discussed earlier,this decision is based on the combi- China's net imports of automobiles, chemicals, nation of agricultural wages and returns to land, electronics, machinery, and commercial services becausemigrantsrisklosingtherighttothelandthey also increase strongly, offsetting the increasing leave. Because the returns to land fall, the returns to trade surplus from the textile and apparel sectors. remaining in farming rise very little, thereby leading Finally, table 16.6 reports the percentage change to additional off-farm migration (1.89 million). in commodity prices relative to the Consumer Price Index in China. Wool, vegetable oil, and automo- Sector Impacts biles experience substantial relative price declines, Table 16.6 reports the impact of accession on out- because imports have a high market penetration put, imports, exports, and relative prices in the ratio and their rates of import protection are cut 296 China and the WTO TABLE 16.5 Implications of China's WTO Accession, 2007 (percentage change) High Labor Low Labor Transfer Elasticity Transfer Elasticity Macroeconomic variables Welfare (Equivalent Variation) 0.73 0.62 GDP 0.65 0.54 Consumption 1.02 0.94 Investment 0.46 0.39 Exports 14.98 14.35 Imports 14.23 13.64 Factor prices Returns to agricultural land 3.54 1.65 Capital returns 0.56 0.35 Unskilled wages Urban 0.30 0.24 Rural nonagricultural 0.34 0.23 Agricultural 0.76 0.61 without land return 0.13 0.08 Semiskilled wages Urban 0.37 0.69 Rural nonagricultural 0.26 0.64 Agricultural 0.28 0.54 without land return 0.66 0.00 Skilled wages Urban 1.11 0.89 Rural nonagricultural 1.11 0.89 Agricultural 0.42 0.21 without land return 1.10 0.21 Inequality measurementa Urban/rural income ratio 0.0107 0.0167 Gini 0.0013 0.0012 Urban 0.0004 0.0002 Rural 0.0001 0.0003 Transaction costs Unskilled 0.15 0.31 Semiskilled 0.63 0.84 Labor migration (millions) Agricultural-nonagricultural 1.89 0.54 Rural-urban 1.46 0.45 Unskilled 0.35 0.13 Semiskilled 1.03 0.29 Skilled 0.07 0.03 Labor migration (%) Agricultural-nonagricultural 2.42 1.55 Rural-urban 2.05 1.07 Unskilled 1.20 0.62 Semiskilled 2.86 1.75 Skilled 1.28 0.59 aChange of original value, not percentage change. Source: Authors' model simulation results. Implications of WTO Accession for Poverty in China 297 TABLE 16.6 Sector Impacts of China's WTO Accession in 2007: Percentage Deviation from Baseline Net Trade Prices (Relative (Billions of to Consumer Output Import Export Yuan) Price Index) Rice 0.8 151.7 2.3 4.5 0.4 Wheat 4.2 50.9 -- 8.9 1.8 Corn 3.7 181.1 66.5 6.3 0.2 Cotton 3.2 78.6 23.8 8.2 0.5 Other nongrain crops 1.0 53.9 2.4 7.8 0.3 Forestry 2.9 16.9 2.2 2.9 0.3 Wool 32.9 24.8 19.2 1.8 14.6 Other livestock 1.3 1.6 3.5 0.3 0.2 Fishing 0.5 3.3 2.1 0.1 0.6 Other agriculture 0.1 5.7 6.2 0.1 0.2 Coal mining 0.8 1.8 1.3 0.1 0.6 Crude oil and natural gas 0.3 0.3 5.7 1.9 0.4 Ferrous ore mining 2.3 3.6 3.6 1.3 0.3 Nonferrous ore mining 4.1 4.1 0.5 0.7 0.6 Other mining 0.4 1.3 3.5 0.1 0.1 Vegetable oil 14.6 167.2 170.3 20.1 7.9 Grain mill and forage 0.8 5.1 27.9 0.9 0.6 Sugar 10.9 180.1 295.5 2.9 4.5 Processed food 1.3 17.0 10.2 5.9 0.3 Beverages 0.6 130.0 11.2 5.0 0.7 Tobacco 2.1 81.4 6.6 6.1 1.3 Textiles 17.9 62.0 45.3 60.6 1.3 Apparel 38.6 54.8 102.2 286.2 2.7 Leather 1.3 19.7 1.4 6.4 0.6 Sawmills and furniture 1.0 14.9 3.3 3.5 0.4 Paper and printing 3.3 13.6 5.5 16.3 0.5 Crafts and toys 1.1 6.5 2.7 3.4 0.2 Petroleum refining 0.0 1.0 4.8 4.8 0.6 Chemicals 1.8 7.9 3.0 26.1 0.7 Pharmaceuticals 0.3 14.1 5.2 0.3 0.3 Chemical fibers 8.0 48.0 6.8 28.7 1.1 Rubber and plastics 2.1 10.7 1.3 5.3 0.4 Building materials 0.4 13.2 2.7 2.5 0.1 Primary iron and steel 3.8 2.4 1.3 3.8 0.1 Nonferrous metals 4.8 1.7 0.1 2.1 0.1 Metal products 2.4 10.1 1.5 11.4 0.4 Machinery 5.1 9.1 0.1 21.8 0.7 Special equipment 4.6 7.0 1.1 22.1 1.6 Automobiles 19.7 83.5 37.9 50.1 11.3 Other transport equipment 0.9 0.1 2.2 1.3 0.1 Electric machinery 2.6 13.4 1.0 21.0 0.8 Electronics 3.8 6.3 0.3 29.2 2.1 Instruments 3.3 6.9 0.3 7.1 2.0 Other manufacturing 1.7 21.6 0.2 2.4 0.5 Utilities 0.6 16.5 2.9 0.4 0.1 Construction 0.3 10.3 3.1 2.6 0.2 Transportation 0.3 1.5 3.2 4.7 0.1 298 China and the WTO TABLE 16.6 (Continued) Net Trade Prices (Relative (Billions of to Consumer Output Import Export Yuan) Price Index) Post and communication 0.5 4.0 4.3 2.1 0.4 Commercial sources 1.7 21.7 2.4 15.8 0.2 Finance 0.5 17.4 1.7 2.7 0.1 Social services 1.0 14.9 2.1 13.2 0.8 Education and health 0.3 18.6 0.9 2.9 0.5 Public administration 0.4 18.0 1.4 1.9 0.3 -- Not available. Source: Authors' model simulation results. FIGURE 16.1 Change in Sector Output due to WTO Accession, China Output (% change) 50.0 40.0 30.0 20.0 10.0 0.0 10.0 20.0 30.0 40.0 oils Corn Sugar Wool ApparelTextilesfibersCotton Tobaccoplasticsmining steelminingWheat metals products Forestryprinting and and ore machineryand InstrumentsElectronicsore equipment Machinery Automobiles Vegetable Chemical Metal Iron RubberFerrous Electric Paper SpecialNonferrous Nonferrous Sector Source: Authors' model simulation results. sharply. For most other sectors the price changes enterprises. Government transfers are assumed to are quite modest. be constant in real terms and are indexed by the Consumer Price Index. The smallest welfare increases for the other Household Impacts (nontransfer) households in table 16.7 are associ- Table 16.7 reports the household impacts in 2007 of ated with agriculture-specialized households. They China's accession to the WTO. The first point to are adversely affected by the decline in returns to note in table 16.7 is that WTO accession benefits agricultural land, but rising returns to labor still all households, except those reliant on transfers. The boost the average returns to farming overall as latter group, which receives much of its transfer some individuals leave the sector. Thus, even the income from state-run enterprises, is adversely agriculture-specialized households benefit from affected by the decline in rents accruing to these WTO accession. Figure 16.2 contrasts the gains for Implications of WTO Accession for Poverty in China 299 TABLE 16.7 Household Impacts of China's WTO Accession, 2007 (Equivalent Variation as percentage of household income) Urban Rural Vingtile Transfer- Labor- Agriculture- (Poorest 1) Specialized Specialized Diversified Specialized Diversified High labor transfer elasticity 1 0.274 1.217 1.044 0.446 0.534 2 0.548 1.395 1.247 0.426 0.571 3 0.450 1.533 1.279 0.343 0.606 4 0.534 1.477 1.103 0.295 0.591 5 0.584 1.451 1.252 0.366 0.609 6 0.464 1.571 1.048 0.355 0.636 7 0.397 1.631 1.315 0.320 0.658 8 0.287 1.607 1.395 0.522 0.636 9 0.381 1.706 1.295 0.370 0.623 10 0.550 1.844 1.308 0.409 0.671 11 0.457 1.709 1.269 0.316 0.650 12 0.397 1.731 1.271 0.367 0.648 13 0.356 1.766 1.267 0.345 0.635 14 0.263 1.719 1.224 0.421 0.683 15 0.270 1.826 1.339 0.329 0.678 16 0.190 1.768 1.204 0.654 0.664 17 -- 1.631 1.102 0.782 0.749 18 0.052 1.682 1.230 0.629 0.642 19 0.354 1.774 1.006 0.391 0.618 20 -- 1.743 0.506 0.469 0.619 Low labor transfer elasticity 1 0.270 1.260 1.103 0.129 0.373 2 0.566 1.418 1.272 0.090 0.411 3 0.463 1.453 1.279 0.020 0.440 4 0.553 1.471 1.128 0.029 0.435 5 0.596 1.393 1.212 0.037 0.464 6 0.466 1.470 1.065 0.025 0.501 7 0.437 1.520 1.264 0.011 0.531 8 0.291 1.526 1.337 0.188 0.510 9 0.393 1.570 1.239 0.042 0.519 10 0.566 1.689 1.217 0.064 0.573 11 0.470 1.566 1.183 0.014 0.559 12 0.398 1.570 1.179 0.021 0.564 13 0.369 1.570 1.186 0.021 0.550 14 0.326 1.562 1.123 0.068 0.588 15 0.294 1.647 1.233 0.023 0.601 16 0.185 1.547 1.126 0.253 0.596 17 -- 1.488 0.996 0.410 0.672 18 0.073 1.490 1.082 0.309 0.598 19 0.356 1.552 0.921 0.075 0.561 20 -- 1.484 0.477 0.146 0.519 Note: -- Not available. Source: Authors' model simulation results. 300 China and the WTO FIGURE 16.2 Impact of WTO Accession on Rural Households EV as percent of income 0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0.0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 Vingtile Ag-specialized Diversified Source: Authors' model simulation results. FIGURE 16.3 Impact of WTO Access on Urban Households EV as percent of income 2.0 1.6 1.2 0.8 0.4 0.0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 Vingtile Labor-specialized Diversified Source: Authors' model simulation results. rural households specialized in farming with those diversified households who also have significant that have diversified income sources. For the latter income from capital. Because the latter rise more group, income rises more because of the increased modestly, the highest-income, diversified house- role of nonfarm wage earnings. The largest holds benefit proportionately less than the other increases in welfare after WTO accession accrue to nontransfer-specialized, urban household groups. the urban households--especially the wealthier, labor-specialized households. They benefit from Sensitivity to the Elasticity of Labor the fact that, of all factor prices, skilled wages Transfer out of Agriculture increase the most after WTO accession. Figure 16.3 contrasts these changes with the changes in EV As noted earlier, Sicular and Zhao (2002) provide (relative to the initial expenditure) for the urban several estimates of the labor transfer elasticity out Implications of WTO Accession for Poverty in China 301 of agriculture. In our base case results, we use the diverse provinces in China: Guangdong, Liaoning, elasticity that relates the relative supply of labor to and Sichuan. Using these rural and urban surveys, agriculture and nonagriculture (market wage) we estimated earnings and spending patterns for 20 activities to a change in the relative wage rates-- income vingtiles across five representative house- owing to a perturbation in agricultural returns. hold groups in China. These groups were divided Our rationale for selecting this elasticity is that the based on their location (rural versus urban) and largest relative factor price effect of WTO accession their primary source of earnings, yielding five cate- is the downward pressure on returns to agriculture. gories: agriculture-specialized, rural-diversified, This "push" elasticity is 2.67. It is considerably urban labor-specialized, urban transfer-specialized, higher than the "pull" elasticity engendered by an and urban-diversified. By embedding these disag- increase in the market wage for labor, which is esti- gregated households in a national CGE model of mated to be 0.6. For purposes of sensitivity analy- China, we are able to identify the disaggregated sis, we turn now to the second set of results in household impacts of WTO accession. tables 16.5 and 16.7, which utilize the low labor We find that WTO accession benefits all house- transfer elasticity between agriculture and nonagri- hold groups--with the exception of the urban culture. The second column of table 16.5 reveals transfer-specialized households, which experience that the aggregate gains from WTO accession are modest losses because of the decline in rents accru- somewhat smaller in the low transfer elasticity ing to state-run enterprises. The biggest gains go to case--although not dramatically so. urban labor-specialized households because of the The most important difference between the strong rise in wages for skilled and semiskilled high and low labor transfer elasticity results is the labor in urban areas. The smallest gains go to the impact of accession on the agriculture-specialized rural agriculture-specialized households, which households (table 16.7). In the case of reduced off- face declining returns to agriculture-specific factors farm mobility of labor, the gains of agriculture- and which have limited labor mobility out of farm- specialized households are much smaller than those ing. Indeed, some of these households could actu- arising with a high off-farm labor transfer elasticity. ally lose from accession under low values of the In fact, a few of these farm household vingtiles now farm­nonfarm labor transfer elasticity. Because lose from WTO accession, because they are not able these households are also the poorest in China, to take advantage of the improved opportunities WTO accession runs the risk of widening income outside of the farm sector and fall prey to depressed disparities. Other studies that have abstracted from factor returns in agriculture. This finding confirms intersector labor mobility altogether have come to those of Chen and Ravallion (2003)--see Chap- the same conclusion (Chen and Ravallion 2003). ter 15 in this volume.By not modeling factor market The critical importance of labor markets to the adjustment at all, Chen and Ravallion have effec- rural impact of China's accession to the WTO sug- tively assumed an even smaller labor transfer elas- gests the potential for complementary labor market ticity out of agriculture--zero. In their chapter, they reforms to offset some of these adverse impacts. In conclude that the poor agricultural households will a companion paper to this chapter (Zhai, Hertel, lose from WTO accession. This qualitative differ- and Wang 2003), we find that reforms aimed at ence in findings on rural incidence highlights the facilitating an improved flow of labor out of agri- importance of explicitly modeling labor markets in culture and between the rural and urban markets, order to obtain an accurate assessment of the impact can result in significant gains for rural households. of WTO accession on farm households in China. In that study, we explore the following implications: · reforming agricultural land markets to permit arms-length land rental in all rural areas, Conclusions and Policy thereby facilitating the permanent movement of Implications labor out of farming The aim of this study has been to quantify the · enhancing off-farm labor mobility impact of China's accession to the WTO on · abolishing the hukou system, thereby reducing income distribution and poverty. Toward this end, the transaction costs imposed on rural­urban we have drawn on household surveys from three migrants. 302 China and the WTO When combined, these reforms reduce the esti- Chan, Kam Wing, and Li Zhang. 1999. "The Hukou System and mated 2007 urban­rural income ratio from 2.58 (in Rural-Urban Migration in China: Processes and Changes." China Quarterly 160 (December): 818­55. the absence of WTO accession) to 2.09. When com- Chen, Shaohua, and Martin Ravallion. 2003."Welfare Impacts of bined with WTO accession, the 2007 urban­rural China's Accession to the WTO." World Bank, Washington, income ratio is still reduced to 2.12 as a conse- D.C. Processed. Chen, Shaohua, and Yan Wang. 2001. "China's Growth and quence of these three factor market reforms. Poverty Reduction: Recent Trends between 1990 and 1999." Although these reform scenarios are perhaps more World Bank, Washington, D.C. Processed. dramatic than those likely to be undertaken in the de Melo, J. 1988. "Computable General Equilibrium Models for Trade Policy Analysis in Developing Countries: A Survey." near term, even a small movement in this direction Journal of Policy Modeling 10: 469­504. would clearly go a long way toward offsetting the de Melo, J., and David Tarr. 1992. A General Equilibrium potentially adverse consequences of WTO accession Analysis of US Foreign Trade Policy. Cambridge, Mass.: MIT on rural­urban inequality. Press. Dervis, K., J. de Melo, and Sherman Robinson. 1982. General Equilibrium Models for Development Policy. Cambridge: Cambridge University Press. Notes Development Research Center. 1998. The Global and Domestic Impact of China Joining the World Trade Organization, 1. All dollar amounts are current U.S. dollars. Research Paper, Development Research Center of the State 2. We would have preferred to base this split on occupa- Council, PRC, Beijing. tion--that is, what they actually do--rather than potential as Dimaranan, Betina V., and Robert A. McDougall. 2002. Global determined by education. However, the rural household survey Trade, Assistance, and Production: The GTAP 5 Data Base. does not support this split. West Lafayette, Indiana: Center for Global Trade Analysis, 3. Because the rural survey reports only the highest educa- Purdue University. tional attainment of the household, attainment by worker is Francois, Joseph F., and Spinanger, Dean. 2004. "Regulated Effi- unknown. This factor biases the skill level of rural households ciency, WTO Accession, and the Motor Vehicle Sector in upward, but, because the vast majority of rural households are China." In D. Bhattasali, S. Li, and W. Martin, eds., China and unskilled, this bias is less of a problem in practice. the WTO: Accession, Policy Reforms, and Poverty Reduction 4. However, as noted by Parish, Zhe, and Li (1995), the rural Strategies. Washington, D.C.: World Bank and Oxford labor market is looking more like a market all the time. University Press. 5. Other, unobserved factors inducing this rural­urban wage Ianchovichina, Elena, R. McDougall, and T. Hertel. 1999."China differential are likely. If so, estimation of the labor market distor- 2005 Revisited: The Implications of International Capital tion via subtraction of known factors is biased in the direction Mobility." Global Trade Analysis Project, Purdue University, of overstating the hukou-related distortion. Therefore, it is use- West Lafayette, Indiana. Processed. ful to estimate as well the direct impact of household registra- Hertel, Thomas W., Maros Ivanic, Paul V. Preckel, and John tion status on the observed wage difference among households. Cranfield. 2004. "The Earnings Effects of Multilateral Shi (2002) takes this approach to the problem, using the same Trade Liberalization: Implications for Poverty." World Bank CHNS data set. He finds that only 28 percent of the rural­urban Economic Review. Forthcoming. wage difference can be explained directly via the coefficient on Kanbur, R., and Zhang, Xiaobo. 2001. Fifty years of regional the hukou registration variable. This figure is quite a bit less than inequality in China: A journey through revolution, reform the 48 percent left unexplained via the subtraction approach of and openness. Working Paper 2001-04, Department of Shi, Sicular, and Zhao (2002). Applied Economics and Management, Cornell University, 6. See Zhai, Hertel, and Wang (2003) for a detailed descrip- Ithaca, New York. tion of how this ad valorem distortion is obtained. McKibbin, Warwick, and K. K. Tang. 2000. "Trade and Financial 7. We assume that a doubling of temporary migration would Reform in China: Impacts on the World Economy." World only increase the marginal cost of migration by 10 percent. Economy 23(8): 979­1003. 8. On January 1, 1995, the ATC entered into force and Naughton, Barry. 1996. "China's Emergence and Prospects as a replaced the old Multi-Fiber Arrangement. The ATC provides Trading Nation." Brookings Paper on Economic Activity 2, for elimination of the quotas and the complete integration of Washington, D.C. textiles and apparel into the WTO regime over a 10-year transi- Parish, W. L., X. Zhe, and F. Li. 1995. "Nonfarm Work and tion period, ending on January 1, 2005. All WTO countries are Marketization of the Chinese Countryside." China Quarterly subject to ATC disciplines, and only WTO members are eligible 143: 697­730. for ATC benefits. Rozelle, Scott, Guo Li, Minggao Shen, Amelia Hughart, and John Giles. 1999. "Leaving China's Farms: Survey Results of New Paths and Remaining Hurdles to Rural Migration." China References Quarterly 158 (June): 367­93. Shi, Xinzheng. 2002. Empirical research on urban-rural income The word processed describes informally reproduced works that differentials: The case of China. unpublished manuscript, may not be commonly available through libraries. CCER. Beijing: Beijing University. Beghin, John, Sébastien Dessus, David Roland-Holst, and Shi, Xinzheng, Terry Sicular, and Yaohui Zhao. 2002. "Analyzing Dominique van der Mensbrugghe. 1994. "Prototype CGE Urban-Rural Income Inequality in China." Paper presented Model for the Trade and the Environment Programme- at International Symposium on Equity and Social Justice in Technical Specification." OECD Development Centre, Paris. Transitional China," Beijing, July 11­12. Implications of WTO Accession for Poverty in China 303 Shoven J. B., and J. Whalley. 1992. Applied General Equilibrium Yang, Y., and Y. Huang. 1997. "The Impact of Trade Liberaliza- Analysis. Cambridge: Cambridge University Press. tion on Income Distribution in China." China Economy Sicular, Terry, and Yaohui Zhao. 2004. "Earnings and Labor Working Paper 97/1. Research School of Pacific and Asia Mobility in Rural China: Implications for China's Accession Studies, Australia National University, Canberra. to the WTO." In D. Bhattasali, S. Li, and W. Martin, eds., Zhai, Fan, and Shantong Li. 2000."The Implications of Accession China and the WTO: Accession Policy Reform, and Poverty to WTO on China's Economy." Paper presented at Third Reduction Strategies. Washington, D.C.: World Bank and Annual Conference on Global Economic Analysis, Oxford University Press. Melbourne, Australia, June 27­30. Walmsley, Terrie, and Thomas Hertel. 2000. "China's Accession Zhai, Fan, and Zhi Wang. 2002. "WTO Accession, Rural Labour to the WTO: Timing Is Everything." Paper presented at Migration and Urban Unemployment in China." Urban International Conference on Global Economic Transforma- Studies 39, no.12. tion after the Asian Economic Crisis, Hong Kong. Zhai, Fan, Thomas W. Hertel, and Zhi Wang. 2003. "Labor Mar- Wang, Zhi. 1999. "Impact of China's WTO Entry on Labor ket Distortions, Rural-Urban Inequality and the Opening of Intensive Export Market--A Recursive Dynamic CGE China's Economy." GTAP working paper 27. Center for Analysis." World Economy 22 (3): 379­405. Global Trade Analysis, Purdue University, West Lafayette, ------. 2001. "The Impact of China's WTO Accession on Trade Indiana. Available at http://www.gtap.agecon.purdue.edu/ and Economic Relations Cross the Taiwan Strait." Economics resources/res_display.asp?RecordID 1323. of Transition 9 (3). Zhao, Yaohui. 1999a. "Labor Migration and Earnings Differ- Wang, Zhi, and Fan Zhai. 1998. "Tariff Reduction, Tax Replace- ences: The Case of Rural China." Economic Development and ment and Implication for Income Distribution in China." Cultural Change, 47(4): 767­82. Journal of Comparative Economics 26: 358­87. ------. 1999b. "Leaving the Countryside: Rural-to-Urban WTO (World Trade Organization). Protocol on the Accession of Migration Decisions in China." American Economic Review the People's Republic of China. Geneva: WTO. (May 1999): 281­86. Yang, Dennis T. 1997. "China's Land Arrangements and Rural Labor Mobility." China Economic Review 8 (2): 101­16. Yang, Dennis Tao, and Hao Zhou. 1999. "Rural-Urban Disparity and Sectoral Labour Allocation in China." Journal of Devel- opment Studies 35 (3): 105­33. 17 Coping with and Adapting to Job Losses and Declines in Farm Earnings in China Athar Hussain The knock-on effects of the conditions of China's losses, there is a social dimension to be considered accession to the World Trade Organization (WTO) whenever the interpersonal distribution of losses on the labor market (including on farmers) are a and gains do not coincide. In the context of China's major issue in analysis of the social impacts of WTO accession, this might happen in the following WTO membership, in particular its impact on four ways: poverty and income equality. The nature of the labor market effects can be expected to differ in the 1. Because of a reduction in the prices of imported short and the long run. In the short run, some of commodities, a large number of consumers gain, the effects will be adverse and will take two forms: while a comparatively small number of work- (1) job shedding in particular industries, and (2) a ers lose jobs or farmers experience a decline in reduction in the earnings of some farmers, depend- earnings. ing on the region and the commodities produced. 2. Loss of jobs or decline in farm earnings precedes Correlatively, the beneficial labor market effects job creation or increases in farm earnings due to will accrue in the form of expanded employment in trade liberalization. the sectors opened wider to foreign direct invest- 3. Those losing jobs or experiencing an income ment (FDI). Additional beneficial effects would decline differ from those recruited into new jobs stem from job creation and the rise in earnings or experiencing an income rise. from the lower barriers facing Chinese exports. 4. Regions reaping gains from WTO accession dif- Broadly, the consensus is that product and labor fer from regions affected by the adverse effects of markets effects all included, gains from WTO WTO accession. membership will far outweigh losses. Moreover, given the increased efficiency flowing from WTO All four scenarios involve noncoincidence accession in the long run the positive labor market between the distribution of gains and losses and impact (changes in jobs and earnings) would ex- thus create a potential issue for social protection. ceed the negative impact. The first concerns the distribution of gains between As for the social impact of WTO accession, the consumers and producers (employees and farm- distribution of gains and losses is as important as the ers), and it arises from the simple fact that, for balance between the two. Even when gains exceed many consumer goods, the number of consumers 305 306 China and the WTO far exceeds the number of producers, including numerous subeconomies that are loosely connected workers employed directly and indirectly. The rather than fully integrated. Regional differences in implication is that adverse effects in the form of job factor endowments, labor productivity, or scale losses or decline in earnings produced by WTO economies within China are arguably as pro- accession are likely to be far more concentrated nounced as those between other countries. One than the favorable effects associated with a fall in implication is that the impact of WTO accession prices dispersed over a large number of consumers. is likely to vary regionally. Moreover, the impact Such patterns of distribution of losses and gains are on the Chinese economy of WTO accession will likely to arise for consumer goods such as passenger depend crucially on what happens to internal trade vehicles, making the adverse effects more visible barriers in the years after accession. A lowering of than the favorable ones. such barriers induced by China's membership in The second is particularly important because the WTO may have the effect of a geographic redis- China had to start lowering barriers to foreign tribution of production within the Chinese eco- imports from the date of accession, December 11, nomy as well as between China and the rest of 2001, but its exports will benefit from reduced bar- the world. It is conceivable that the impact of the riers in other WTO member countries only after a removal of internal trade barriers may dominate time lag, in some cases a long and uncertain one. In the impact of lower international trade barriers. short, whereas China's WTO accession involves an Empirical evidence on the seriousness of losses obligation to lower barriers to imports, it provides in the Chinese context is not available. However, the no more than a distant prospect of lower barriers evidence from the United States provides a bench- on exports. The abolition of the Multi-Fiber mark for estimating losses suffered by adversely Arrangement, from which China is expected to be a affected workers.The re-employment rates of work- major beneficiary, is scheduled to take place in ers displaced by competition from imports are typ- 2005, and Chinese textiles may continue to face spe- ically low. The reason is not the nature of import cial textile safeguards until 2008. Similarly, barriers competition. Rather, the explanation lies in the on foreign farm imports began to fall immediately mismatch between the characteristics of laid-off upon China's accession to the WTO, but it is uncer- workers and those of workers hired for new jobs. tain when China's farm exports will face lower bar- Workers to be laid off are not chosen randomly but riers; it is contingent on the outcomes of bilateral selectively. In the United States, workers losing jobs trade negotiations and of the Doha Round of trade are likely to be older and to have lower educational talks, which will take some years to conclude. attainment. In contrast, new jobs tend to demand In the third and fourth scenarios in the list, a younger and better-educated workers. The patterns mismatch occurs between the distributions of of selection of employees for redundancy and of losses and gains over individuals and regions. In the new recruits are aimed at raising the efficiency of first of two examples of likely outcomes of WTO the U.S. labor force, and these patterns are likely to accession, employment expands in the textile and apply in China as well. Furthermore, the new jobs clothing industry and the number of jobs in the that laid-off workers find tend to be lower paying vehicle and motor parts industries falls. Those get- than the previous jobs. ting jobs in the textile and clothing industries will In the United States, workers losing jobs because naturally not all be just those losing jobs in the of competition from imports may experience a vehicle industry, even when both sets of industries long spell of unemployment and a significant drop are located in the same place. The second example in wages when they find a new job (Kletzer 2001). concerns agriculture, where grain farmers, espe- Arguably, in terms of spells of unemployment, the cially those heavily dependent on corn, wheat, and losses suffered by Chinese workers losing jobs coarse grain, will likely be losers. These losers will because of WTO accession are likely to be no less be concentrated in the Northeast and Northwest. than those experienced by U.S. workers unem- The gainers will consist of producers of export ployed because of competition from imports. crops and animal products. Most of the gainers will Indeed, they may be higher for three reasons. First, be in the coastal provinces of the East and the China, as a developing economy with a large pool Southeast. The Chinese economy is made up of of surplus labor, would have more job seekers Coping with and Adapting to Job Losses and Declines in Farm Earnings in China 307 per vacancy than the United States. Added to this, this impact is intertwined with those of the two workers losing jobs because of WTO accession transitions under way in the Chinese economy: the would be competing with the large numbers of transition from a planned to a market economy workers laid off from the state sector. Second, bar- and the process of economic development that is riers to labor mobility in China serve to prolong seeing a transfer of labor out of farming. Both of unemployment between jobs. Third, the problem these transitions involve job displacement and a of a lower re-employment rate among older work- redeployment of labor on a large scale. The labor ers, who are likely to account for a substantial share recruitment practices of the planned economy left of employees selected to be laid off, is in the behind a legacy of surplus labor in the state sector, Chinese context compounded further by the fact which, forced by competition, has been shedding they that they have much lower educational attain- this excess baggage. In just six years, from 1996 to ment than that of younger workers. Because of 2002, the state sector (including enterprises, public the adverse impacts of the Cultural Revolution institutions, and the government) has lost about (1966­78) on secondary and higher education, a 40 million jobs, or 37 percent of the total in 1996, large proportion of birth cohorts from 1955 to and the trend is likely to continue for some years to 1965 (ages 37­47 in 2002) missed out entirely on come. The argument is that job losses or job dis- post-basic education and occupational training. As placement arising from WTO membership would a result, the difference in educational attainment simply be a part of the same phenomenon on a between younger and older workers in China is larger scale. Similarly, a reduction in income among particularly wide. sections of the farming population, which is likely A preliminary hypothesis is that China's WTO to be one of the effects of WTO membership, has membership is highly unlikely to have an identifi- for other reasons been a common occurrence, espe- able impact on the wages of skilled and unskilled cially since the late 1990s. employees--an issue that dominates the discus- China's accession to the WTO in December sion of the impact of imports from developing 2001, although a historical landmark, was just one economies on U.S. workers (Collins 1998). Like- event in a series of step-by-step openings of the wise, a rise in the demand for unskilled labor Chinese economy to international trade and for- because of a lowering of impediments to Chinese eign direct investment since 1979, referred to as the exports, such as textiles and clothing, is highly Open Door policy. From a historical perspective, unlikely to translate into a rise in the wages of the effects directly attributable to WTO accession unskilled workers, as predicted by the standard may be seen as one chapter in the extensive story international trade theory that proceeds on the (i.e., ramifications) of the Open Door policy. Given assumption of full employment. In China, a devel- the long time it has taken China to gain entry to the oping economy, the supply of unskilled workers far WTO, the event has been "expected" and "prepared exceeds demand, and a rise in demand induced by for," at least to a degree. Some of the effects of WTO WTO membership is likely to be small relative to accession have preceded the cause itself. This is par- the excess supply. In fact, the wages of professional ticularly true about the restructuring of industries and skilled employees have been rising more than over the last few years and perhaps also about the those of unskilled workers in China, and the trend decision to convert marginal farming land into is likely to continue. This pattern runs contrary to forests. This situation is akin to the observation the expected change in relative wages, but it has that part of the relative decline in the wages of little to do with international trade. Rather, it stems unskilled works observed in the United States may largely to a correction of the highly compressed be attributable to employers securing restraint in wage structure inherited from the planning period. wage demands by using the threat of develop- This chapter proceeds on the assumption that the ing country imports--that is, the expectation of impact of WTO entry on industry and services will heightened competition may have the same effect as principally be on jobs and that on farming will be the actual event. Most of the likely impact of WTO on earnings. membership on the Chinese economy is attributa- For anyone studying the labor market impact of ble to sharper competition in product and services WTO membership, the principal problem is that markets. In this respect, increased competition 308 China and the WTO between domestic producers, which has risen fall in labor earnings, such as old-age pensions and sharply since the mid-1990s, has the same effect as health insurance. The category "social protection" that produced by heightened competition between covers various schemes differentiated by the target domestic and foreign producers after WTO entry. of assistance and the conditions attached to the All this discussion leads to one implication and provision of assistance, such as the requisite contri- raises a question. The implication is that the adverse bution record or insufficient means. These schemes labor market impacts of WTO membership would do not prevent job losses or declines in earnings; be neither novel nor out of scale with what the they operate after the event and provide no more Chinese economy has already experienced in recent than partial compensation. Full compensation is years, or more specifically since the second half of ruled out in order to maintain economic incentives the 1990s. This implication suggests that observers to remain employed and because of budgetary con- keep an open mind about the magnitude of the straints. The category "adaptation" covers a diverse effects, at least at the national level. One important range of schemes with the common feature of facil- question is whether separate or special policies are itating individuals, localities, or industries to adapt needed to deal with the adverse impacts of WTO to a new trade environment through labor rede- entry. Arguably, the policy responses to job losses or ployment and modification of the composition of falls in farm incomes should be the same regardless economic activities. Most of the schemes falling of their proximate causes. A justification for special under this category have not been introduced to measures is that the adverse impact of the pre- deal with the effects of WTO entry, but they lend scribed reduction in trade barriers after WTO entry themselves to serving that purpose. A notable may be concentrated over a short time period and exception is the restructuring of various industries fall heavily on particular regions, such as the rural over the last few years in the expectation of WTO areas of the already impoverished Northwest. accession, such as those in the textile, steel, and vehicle industries. Some of the schemes, such as public works programs, contain elements of both Social Protection and Adaptation social protection and adaptation. Schematically, the current social security schemes Two gaps in the schemes and measures and policy measures that may be used to respond to (table 17.1) that would provide protection and the adverse impacts of WTO accession on the labor facilitate adaptation to the adverse impacts of WTO market (i.e., job losses and lower farm earnings in entry stand out. First, the schemes for households/ China) fall into four categories, depending on the individuals that fall under "social protection" vary policy aim and the target (table 17.1). greatly between rural and urban inhabitants, which Excluded from table 17.1 are the social security leaves some groups such as migrants unprotected. schemes that are not directly related to a job loss or Second, among the schemes for localities/industries TABLE 17.1 Classification of Schemes and Measures Aim Target Social Protection Adaptation Households/individuals · Unemployment insurance · Job search facilities · Poverty relief · Retraining · Assisted migration Localities/industries · Schemes to assist poor · Development schemes for rural localities lagging regions · Public works programs · Assisted industrial restructuring · Assisted conversion of marginal forest land into forests Source: The authors. Coping with and Adapting to Job Losses and Declines in Farm Earnings in China 309 that fall under "adaptation," there are no assistance Instruments for Social Protection measures targeted at towns and cities adversely Social protection schemes are intended to reduce affected by industrial restructuring. These meas- the incidence of poverty and economic inequality. ures assume a special importance in the Chinese They provide a partial compensation for a job loss context because of impediments to labor migra- or fall in income, but on their own they do not tion. The problem of towns and cities caught in an facilitate a redeployment of labor, which is the ulti- unemployment-and-stagnation trap has come to mate aim. Following the usual convention, such the fore with the massive labor redundancies and schemes fall into two categories: social insurance layoffs in the state sector since 1996. The problem is and social assistance (social safety nets). particularly serious in medium and small towns China, which has both types of schemes, has a heavily dependent on a few state enterprises that more extensive social insurance system than those have been forced to shed labor because of financial found in most developing economies. This system, difficulties. In such towns, job losses in a few enter- then, provides the Chinese government with two prises accounting for a substantial percentage of types of measures to compensate, partially, the local employment can generate significant negative losers from trade liberalization. The differences pecuniary externalities. Such externalities arise between the two sets of schemes arise from two whenever there is a combination of imperfect com- related features: first, the conditions attached to the petition and increasing returns to scale. In the provision of benefits (entitlement qualification) Chinese context, the negative regional impact of and, second, the method of financing. Under social job losses is further exacerbated by the arrange- insurance schemes, benefit payment is jointly con- ment whereby urban income maintenance schemes ditional on the contribution record and on the are largely financed at the city level. Cities with a occurrence of specific contingencies, such as unem- higher poverty or unemployment rate also tend to ployment. The first condition makes social insur- have strained public finances and therefore must ance similar to commercial insurance and limits its reduce coverage or benefit levels. As described later coverage of the population. In China, as in many in this chapter, there is a rural assistance scheme economies, social insurance is confined to the labor targeted at poor localities, but that scheme is of force in formal wage employment, and, unlike in little or no use as a cushion against the adverse other economies, is restricted to registered urban impacts of increased agricultural imports on farm residents.1 Under formal wage employment, it is incomes. easy to collect contributions in the form of a pay- The framework employed in this chapter to roll tax and define and verify the state of unem- analyze social protection measures is structured ployment, both of which raise difficulties in around the following questions. the case of informal wage employment or self- employment. The protective cover provided by · Do the beneficiaries (individuals/industries/ social insurance in China is subject to two limita- localities) of a social protection measure overlap tions. First, coverage is in practice limited to the with those adversely affected by WTO entry? If labor force in urban areas in formal wage employ- yes, how significant is the overlap, and how ment, which represents 14 percent (105.6 million) effective is the targeting? out of the total labor force of 737.4 million · How does the measure address the adverse (National Bureau of Statistics 2003a).2 Second, the impacts (e.g., assistance or facilitating adapta- contribution condition imposes a limit on unem- tion)? ployment compensation. In most schemes, the · How is the measure financed, and is the duration of compensation is linked to the duration financing adequate to meet the measure's of previous contributions and is subject to a maxi- objectives? mum, which in China is two years. The implication · What effects does the measure have on incentive is that unemployment insurance cannot be the sole and efficiency? In terms of efficiency, would an mainstay of a comprehensive social protection alternative measure achieve the desired effect at system; it has to include a needs-based income a lower cost? maintenance scheme. Financed by contributions, 310 China and the WTO social insurance resembles commercial insurance, the means test and some supplementary condition but it differs in two respects, which help to extend are satisfied.3 China has several social assistance its role in mitigating risks: schemes, but each covers only a particular section of the population, which leaves some glaring blind 1. Participation is invariably mandatory for a des- spots in coverage (this is discussed later in this ignated section of the population, such as urban chapter). Table 17.2 presents the salient differences wage employees, which avoids adverse selection between social insurance, with reference to unem- whereby low-risk groups opt out, thus raising ployment insurance, and social assistance. the percentage of higher risk groups in the cov- Though overlapping in aims, social assistance ered population. and social insurance schemes are as much com- 2. The relationship between benefit and contribu- plements as they are substitutes. Developed and tion in social insurance is typically looser than comprehensive social security systems invariably in commercial insurance and involves redistri- include both types of schemes. Does a low-income bution in favor of vulnerable groups. developing economy such as China need both types of schemes, especially social insurance that covers In contrast to social insurance, social assistance only a minority of the labor force, and usually that is noncontributory and thus far removed from in regular wage employment? A low income does commercial insurance. A key distinguishing feature not constitute a barrier against the institution and is that it can, in principle, provide protection to functioning of social insurance, because benefits everyone in need in the relevant population. Nor- are tied to contributions that are usually a percent- mally, social assistance is available to all people age of payroll. The principal advantage of social whose resources fall under a specified minimum-- insurance is that, when well designed, it generates for example, an income that falls under the poverty most if not all of the funds needed for its opera- line. Furthermore, the benefit is available as long as tion, which matters when possibilities for raising TABLE 17.2 Differences between Social Insurance and Social Assistance Social Insurance Social Assistance (Safety Nets) Financing · Self-financing, at least partially if · Must be financed from general not totally government revenue, which poses · Easy to combine with commercial problems in low-income localities insurance to cover additional risks with a tight budget Benefits · Can be tailored to serve a range of · Usually set a level just sufficient to objectives avert hardship, in some cases only · Typically, benefits more generous severe hardship than those under social assistance, but only available for a limited duration Coverage · Limited to those who can make · Restricted to those with insufficient the requisite contributions. By income and savings to cover basic design, such schemes cannot needs provide universal coverage · In principle, can cover the whole population, depending on the budgetary constraint Impact on poverty and · Provides protection against · Can serve as social safety net to economic inequality poverty to participants and those ensure protection against poverty with an adequate contribution · Limited reduction in economic record inequality constrained by · Reduction in inequality limited by targeting coverage Source: The author. Coping with and Adapting to Job Losses and Declines in Farm Earnings in China 311 revenue through alternative taxes are limited. The them in providing income security for the rural contribution condition that limits the coverage of population. Three features characterize Chinese social insurance also makes it attractive from a income maintenance schemes. The first is the stark fiscal point of view. In contrast, social assistance dichotomy between the urban and rural popula- schemes are, by design, not self-financing. The tion. At present, no scheme covers both the popula- principal argument in favor of introducing and tions, which are defined differently in China than extending unemployment insurance is that it pro- elsewhere.4 Second, the schemes aim to do no more vides a first line of defense against poverty and than alleviate severe poverty. In the two urban reduces the task of poverty alleviation left for social schemes, the benefits are set at a flat rate either just assistance schemes. The latter is of particular above or at the local poverty line. The two rural importance when the coverage of social assistance assistance schemes are narrowly targeted, though is limited by budgetary stringency, as it is in China in many cases inaccurately. The third feature is and other developing economies. a high degree of decentralization. With the excep- tion of the "Assistance to Poor Rural Localities" program, all schemes are in the first instance Overview of Chinese Income Maintenance Schemes financed by low tiers of government: cities for Table 17.3 presents an overview of various income the two urban schemes and villages for the "Assis- maintenance schemes in China relevant to coping tance to Poor Rural Households" program. Higher with job losses or declines in income stemming government tiers contribute toward the cost of the from WTO accession, excluding old-age pensions schemes. But there is as yet no settled formula for and disability compensation. cost sharing, and lower tiers of government end up The land plots allocated to households are bearing a share of costs that is too high to be equi- included as an income maintenance instrument in table and consistent with the sustainability of the table 17.3 because of the importance accorded to schemes. TABLE 17.3 Income Maintenance Schemes in China Urban Schemes Rural Schemes Unemployment Minimum Living Assistance to Poor Assistance to Poor Allocation of a Insurance Standard Localities Households Land Plot Coverage Urban, formal Registered urban Designated rural Usually restricted Long-standing wage employees residents localities to very poor inhabitants of households and the locality highly variable across localities Benefits · Cash allowance · Cash allowance · Investment in Benefits in cash and A plot of land · Health insurance · Health insurance social services, kind infrastructure, · Free or to cover large and economic subsidized medical costs activities services · Subsidies targeted at households Financing Contributions Central and Central and Locally financed and residual territorial provincial government governments governments financing Adminis- City labor and City civil affairs Civil affairs bureaus Villages Villages tering social security bureaus agency bureaus Source: The author. 312 China and the WTO BOX 17.1 Advantages and Disadvantages of Decentralization Decentralization in the Chinese social security differences. Furthermore, such variations may system offers both advantages and disadvan- impede the transferability of social security tages. entitlements (such as unemployment bene- fits) and thus impede labor mobility. Non- Advantages: transferability is the principal barrier in · Decentralization allows flexibility to account bringing migrants under the social insurance for local variations and gives local govern- cover. ments the freedom to take initiatives. In fact, · Decentralization accentuates disparities in some local government initiatives have later social security provisions. Localities with a been adopted as national policy. For exam- higher poverty level or a higher percentage of ple, the Minimum Living Standard Insurance the needy in the population also tend to have (MLSI) scheme for urban areas, now a nation- more strained public finances and thus are wide scheme, was first implemented in a few forced to reduce coverage or the benefit coastal cities. A uniform system for all of level. China runs the risk of being determined by the lowest common denominators. Planned Reform Disadvantages: The pilot program for improving the urban · Municipalities and the tiers for the organiza- social security system, which is currently being tion of rural social security are, in many tested in Liaoning Province, recognizes the instances, too small to provide sufficient risk problems created by decentralization. One of its pooling for maintaining the financial integrity aims is to upgrade the level of budgeting for of schemes. Decentralization is associated social insurance from the municipal to the with variations in standards of provision that provincial level in selected localities, starting may go beyond the socially acceptable with the old-age pension scheme. The rural­urban segmentation that casts a long palpable as that facing their counterparts in urban shadow on the Chinese social security system has enterprises; the labor retrenchment in TVEs in several facets. The urban and rural schemes not recent years is evidence. Additionally, many of the only are mutually exclusive, but also are structured TVE employees are just as exposed to the impact of differently. For the rural population, there are no international trade as their urban counterparts. equivalents of unemployment insurance (UI) and This being the case, the direct impact of WTO entry the urban Minimum Living Standard Insurance on rural China would not be confined to farming-- (MLSI), although the latter has begun to change. it also would extend to rural industry. The impact The explanation for confining UI to urban em- of WTO entry on industry would straddle both ployees is not that the rural labor force, being self- urban and rural areas. employed, does not face the risk of unemployment. The MLSI, a social assistance scheme based on Although a majority of the rural labor force is means, marks another major difference between engaged in the cultivation of its own land plots, a social protection in urban and in rural areas. The substantial percentage is employed in farming and scheme, which was introduced in 1997 and was by nonfarming activities, and most of these are wage- the middle of 2000 established and functioning in employed. Over a third (35 percent) of the 490 mil- all towns and cities, aims to provide assistance to all lion strong rural labor force is engaged full time in registered urban residents with a household per nonfarming activities, which says that in China capita income below the local poverty line. The rural is far from synonymous with agricultural. number of beneficiaries under the scheme has risen Leaving aside household units, town and village sharply since inception, indicating improved cover- enterprises (TVEs) employ 133 million workers, or age. The scheme marks a major change in the his- 27 percent of the rural labor force and far more tory of social security in China, in that until its than the total employment in the state sector introduction all social assistance was confined to a (National Bureau of Statistics 2003a). The risk of small section of the population characterized by unemployment facing TVE employees is just as the "three nos": no ability to work, no savings, and Coping with and Adapting to Job Losses and Declines in Farm Earnings in China 313 no relatives to depend on. Under the planned eco- any, in cushioning the adverse impacts of WTO nomy system, those physically able to work were membership on income from farming. In general, expected to work and were provided with a job, the poor rural localities are remote and close to which the government was able to do by virtue of autarky. Market transactions with the economy its control of the economy. The introduction of the outside the locality are sporadic and of low value. MLSI is underpinned by the government's recogni- China's accession to the WTO would have little tion that it cannot provide a job guarantee to all in direct impact on such localities and their inhabi- a market economy. With the combination of MLSI tants. However, the program could be adapted to and UI, the urban population has all the ingredi- cope with the adverse impacts of WTO accession ents of a comprehensive insurance against poverty, on rural areas by redefining its scope. including that caused by the adverse impacts of Also in rural areas is a wide range of assistance WTO accession. As outlined later in this section, schemes that are targeting poor households. The both schemes suffer from deficiencies, but those oldest of these is the so-called Five Guarantees stem largely from problems in the design and Scheme,covering food,clothing,shelter,health care, implementation of the existing schemes. The lack and funerals. The recipients of assistance under the of appropriate schemes does not create gaps in scheme have been mostly elderly persons without social protection for urban households, but it cer- means and relatives.8 Therefore, the scheme is of tainly does for the rural population. little or no relevance in providing assistance to Although the rural population benefits from rural inhabitants adversely affected by WTO entry. numerous antipoverty schemes targeted at poor Other rural assistance schemes cover a wider cate- localities and at poor households (table 17.3), it gory of the poor than just the indigent elderly. does not enjoy the same degree of protection as the These schemes have two features in common. First, urban population. The program offering assistance they are highly decentralized, mostly organized and to poor localities is a big one that dates back to financed at the grass-roots level of villages. Second, 1994, when the government launched the so-called associated with this high degree of decentralization, the "8-7" Plan.5 The plan targeted the 592 poorest the structure of the scheme, funding, eligibility for counties (designated as national poor counties) assistance, and details of assistance vary widely. In selected out of 2,000-plus rural counties. These contrast to the urban schemes (UI and MLSI), the counties, many of which were in the western rural schemes lack the underpinning of a national region, contained, it was estimated, more than framework and are, as a result, far more variable in 70 percent of the 80 million people who remained coverage and benefits. In localities, the spectrum poor. These counties were generally remote, char- runs from well-functioning schemes similar to the acterized by backward socioeconomic conditions6 urban MLSI to no social protection scheme at all. and poor natural endowments--that is, they were Decentralization has two related consequences, areas in which geographic targeting is appropriate. both of which erode the degree of protection pro- The plan was financed entirely by the central gov- vided by the rural schemes targeting households. ernment and backed by a multitiered administra- First, localities with a higher incidence of poverty tive structure from the central government down to tend to have poorly financed and organized the counties. The principal instruments of assis- schemes, exactly the opposite of what is needed. tance were tax reduction, financial grants, and a Second, decentralized schemes are not able to cope socioeconomic development program. The princi- well with correlated risks that adversely affect a pal thrust was toward building infrastructure and substantial percentage of households in the locality. developing the local economy rather than toward Competition from imports is an example of such a providing poor households with assistance. The risk when households in a locality have similar or program came to an end in 2000 and was replaced identical cropping pattern. The second feature of with a new one with similar aims and mode of rural assistance schemes is that they are generally operation, but targeted differently. Putting aside the designed to provide assistance to those in extreme issue of the appropriate territorial units for target- poverty only. The rural population is expected to ing,7 these programs, though valuable in reducing rely on kin and community in case of all but dire severe rural poverty, are of little consequence, if contingencies. In sum, the rural areas lack schemes 314 China and the WTO that can directly address the adverse impacts of rural­urban migration. This trend may continue WTO accession in the form of unemployment and for some time. a drop in farm earnings. After a sharp drop in the 1980s, the percentage gap between the rural and urban income per capita has been widening particularly rapidly in recent Land Plot as Insurance years. The implication is that the level of economic One of the two rationales for the sparse and highly security provided by land plots, as measured by selective social protection cover in rural areas is average farm income per rural inhabitant, has been that each rural household has been assigned a plot falling relative to per capita income. Added to this, of agricultural land that serves as a floor to house- the level of income security land plots provide to hold income, a feature that is particular to China rural households has been highly variable across and a few other economies. The other is the high localities. The distribution of land to rural house- cost of introducing a social security regime compa- holds takes place at the grass-roots level of villages. rable to the one in urban areas. In rural areas, there As a result, the average area of land plot per house- is a limited capacity to collect taxes and social secu- hold varies across villages even in close vicinity. rity contributions, which is a common problem in Land plots have served to reduce the incidence of developing economies. Agricultural land in China, rural poverty but they have not been sufficient to compared with most developing economies, is very prevent poverty, not even extreme poverty. Most of equally distributed. The removal of the adverse the rural population with a per capita income of economic incentives that marred the collective less than 635 yuan per year--the official poverty economy and the provision of each rural house- line for rural inhabitants--has land. hold with a piece of land have together contributed It is also important to point out that not all to the dramatic drop in the (official) headcount of rural households have a land plot, and the percent- the poor, from 250 million in 1978 to 70 million in age without land has been rising. Two examples are 1994, on the eve of the 8-7 Plan.9 Landlessness and described here--one from a prosperous region and rural poverty do not go together in China, in another from a poor region. In rural areas of marked contrast to the prevalent pattern in devel- southern Jiangsu, one of the richest rural areas in oping economies. However, in the present context China which covers 12 counties11 known as Sunnan the pertinent question is how much of a cushion do and has highly developed TVEs, a survey con- land plots provide against job losses in rural areas ducted in 1998 revealed that about 10.3 percent of and a decline in farm income in general and those households no longer possessed land. Reasons arising from China's accession to the WTO. included the conversion of agricultural land to Controlling for the impact of input and output nonfarm uses and the relinquishment of land by prices on farm earnings, income per capita from a households in order to opt out of liability for the land plot depends crucially on the land area per agricultural tax. Furthermore, among the 89.7 per- rural inhabitant, land quality and location, and cent of households with a land plot, another 3 per- technology. The constraint on income per capita cent planned at the time of the survey to give up imposed by the land­rural population ratio can be their land plot in the near future. However, even in pushed back through investment in transport and poorer rural regions not all rural households have a other infrastructure and through technological land plot. A survey of 1,500 households conducted upgrading, but it cannot be eliminated. The land­ in 1999 and drawn from 15 nationally designated rural population ratio acts as a limit on the income poor counties in Gansu and Inner Mongolia found security provided by land. Because of China's huge that about 3 percent of rural households for vari- population, the cultivated area per rural inhabitant ous reasons had no land. Most of the rural house- is slightly less than a fifth of hectare (0.19).10 The holds without land had in the past given up land cultivated land per rural person has been falling voluntarily and for good economic reasons. The steadily since the distribution of land to households important point is that on giving up land the in the early 1980s because of the combination of households are not included in another income rising population and various impediments to maintenance scheme. Coping with and Adapting to Job Losses and Declines in Farm Earnings in China 315 How much of a cushion do land plots provide anomalies for the incidence of the adverse effects of against a decline in household income from, for China's accession to the WTO. example, a job loss from a TVE or a fall in grain Urban is both a spatial and a demographic cate- prices because of competition from imports? Land gory. According to the usual international usage, plots are not transferred to rural households in urban in its spatial sense denotes localities with a perpetuity but are leased for a definite period on high population density and heavily dependent on a renewable basis.12 In principle, the land lease can industry and services as sources of income. In its be sold or transferred, freely, but, in practice, it is demographic sense, urban denotes the population subject to severe restrictions. One serious conse- resident in an urban locality--that is, the spatial quence of the limited tradability of land plots is division urban­rural coincides with the demo- that they cannot be used as collateral for loans. graphic division of the population into urban and Land plots provide rural households with the rural, and the two are interchangeable. In contrast, option of an income generated by deploying house- in China the spatial and the demographic divisions hold labor on their own land plot. It is this income do not coincide, and the urban and rural popula- that determines the value of insurance provided by tions, as officially designated, are intermixed and land plots. In the event of the loss of a job in the not mutually exclusive as they usually are. This nonfarming sector, households have the option of situation creates anomalous divisions in the popu- compensating for the loss of income by redeploying lation and is associated with arbitrary exclusions or the now unemployed labor to their own land plot.13 inclusions in the coverage of the social security The important feature of this scenario is that schemes. Chinese cities are divided into urban income obtainable from the land plot remains districts (shi qu) and rural counties (xian).Similarly, unaffected by the adverse shock. However, when the population is classified by personal registration the decline in household income is the result of status as agricultural (nongye) and nonagricultural competition from cheaper agricultural imports, (fei nongye). The spatial and the demographic divi- household land plots provide no cushion against sions overlap, but only partially, producing the the fall in income. Such competition reduces the following anomalies (National Bureau of Statistics insurance value of land. Because the competitive 2001): threat posed by agricultural imports varies across commodities, the extent of reduction would de- · Of the 296 million people with nonagricultural pend on the cropping pattern and the possibility registration (thus officially urban) living within of changing it. More specifically, in terms of the the boundaries of 663 cities and towns in 2000,14 impact of WTO accession on farming, the North- 42.7 percent (126 million) were resident in rural east and the Northwest would be losers and the East counties, rather than in urban districts, and and Southeast would be gainers. were covered by social security schemes con- fined to the urban population. · Just as anomalous, 40.6 percent (116 million) of Anomalies in the Classification long-term residents in urban districts carried of the Population agricultural registration and were thus regarded As noted earlier, under the current social security as part of the rural population, even though arrangements any assistance that a person may most of them no longer had any relationship receive after a job loss or reduction in earnings, for with farming.15 They were still classified as example, stemming from WTO entry would de- agricultural because the occupational and resi- pend crucially on whether the person is living in an dential shift from farming to industry and serv- urban or rural area. The categories urban and rural ices does not automatically lead to a change in are defined differently in China than elsewhere. registration from agricultural to nonagricultural. This section will outline the anomalies created by These long-term inhabitants of urban districts the combination of the control on population are excluded from social security schemes for migration and the rules governing household regis- the urban population on the grounds of their tration and will point to the implications of these registration status. 316 China and the WTO Thus the official designation urban in China between the two ratios in China is wide even for a does not denote the usual place of residence and developing economy. What factors explain the very work, such as urban districts of cities. Rather, it low relative labor productivity in farming? Two carries the label nonagricultural (fei nongye) on the explanatory factors are possible: low land produc- personal register (hu kou). Urban is a personal tivity and a high labor-land ratio. The first does not attribute or status rather than a spatial designa- have much explanatory power in the Chinese con- tion. A striking example is provided by a compara- text. As in other East Asian economies, farming in tively common anomaly in China where some China is intensive, and productivity per unit of members of a household in an urban district are land (e.g. hectare) is high by international stan- registered as nonagricultural and therefore regarded dards. There is scope for raising land productivity as urban, while other members of the same house- further through the adoption of best-practice tech- hold are regarded as rural on the grounds of their niques, while keeping the labor­land ratio constant. registration as agricultural. Thus a household But the principal explanation of the low labor pro- receiving the MLSI allowance may get less than the ductivity in farming lies with three impediments to amount needed to bring its members up to the the transfer of labor out of farming to nonfarming poverty line, as prescribed, because some of its activities. The first impediment to rural­urban members are not registered urban residents and so migration is the household registration system, do not qualify. A frequent cause of such anomalies which has diminished in rigor but not disappeared. is marriage across the registration line. Similarly, The second impediment is the land tenure system migrants do not qualify for the MLSI allowance. in which rural households receive little or no com- Urban registration, if not inherited, is difficult to pensation for giving up user rights on the allocated acquire, especially for large cities such as Beijing land. This arrangement is in effect tantamount to a and Shanghai. tax on leaving farming completely. The third is the The control on labor migration and the rules bias in favor of industry, which has tended to raise governing assignment of the labels nonagricultural the investment cost of transferring labor into non- and agricultural to individuals produce two oddi- farming activities and has therefore led to a higher ties at the aggregate levels of the labor force and the land­labor ratio than it would otherwise be. population. These oddities are important because An important feature of the administrative con- they affect the pattern of incidence of impacts of trol on rural­urban migration is that the impedi- WTO entry and also the coverage of social protec- ment to the transfer of labor out of farming varies tion schemes. The first oddity is that whereas the according to locality. It does not pose a serious gross domestic product (GDP) ratio of farming is a impediment to the transfer of labor out of farming mere 15.8 percent,16 similar to that in middle- in localities with favorable conditions for the devel- income economies, the percentage of the labor opment of TVEs, such as a relatively high per capita force reported to be engaged in farming is almost income and educational level and good transport three times higher, 47 percent, which is similar to links. In such localities, most of which are in a few that in low-income economies. Put together, the coastal provinces, controls on migration to cities two percentages imply that the output per capita in have, arguably, spurred the growth of TVEs. In these farming, given by the ratio of its GDP share to its areas, the result has been a substantial weakening of employment share, is less than a third of the aver- the link between labor outflow from farming and age, or a mere 21 percent of that in the nonfarming out-migration from the countryside. However, in sector.17 A GDP share of farming that is a small localities where conditions for the growth and pro- fraction (say less than half) of its labor force share liferation of TVEs are comparatively unfavorable, is a common feature of developing economies and the control on labor migration bears down heavily corresponds to a wide gap between nonfarm (or and thus keeps a higher percentage of labor in farm- urban) and farm (rural) per capita incomes in ing. Such localities, which are poorer than those those economies. But with a GDP share of farming with developed TVEs, tend to be found in the inte- close to one-third of its labor force share, which rior provinces. There, migration control functions implies that the urban GDP per capita is about four as a high barrier to the redeployment of labor out and a half times the rural GDP per capita,18 the gap of farming. Given that, on average, income from Coping with and Adapting to Job Losses and Declines in Farm Earnings in China 317 farming is a small fraction of that from nonfarming working in the same employment unit may have dif- activities, any measure that keeps a higher percent- ferent social security cover simply because of their age of labor in farming in effect creates a trap that different designations in the household register. perpetuates low incomes and poverty. What are the implications for the incidence of the Closing Loopholes in Social adverse impacts of WTO membership? First, the Protection larger the percentage of the labor force engaged in farming, the larger is the percentage of the popula- What feasible options are there for reducing, if not tion exposed to the risk of falling income due to abolishing, the arbitrary anomalies in the coverage competition from agricultural imports and there- of social security schemes? An integrated social fore the more widespread are the adverse impacts. protection system that abolishes the urban­rural Second, among the localities reliant on crops espe- distinction as a consideration for the provision of cially vulnerable to imports, the adverse impacts assistance would eliminate the arbitrary anomalies would be variable. In particular, the impacts would altogether and immediately. Such an integrated sys- be more severe in localities with less-diversified tem, however, although highly desirable on equity economies and those with an already low per capita grounds, is not feasible for the foreseeable future income. The two tend to be correlated because less- for financial reasons. The tax system such a system diversified rural localities tend to be more depend- requires is a decade into the future. A piecemeal ent on farming activities that, on average, generate a and gradual elimination of the anomalies is the lower income per capita than do nonfarming activi- only possibility. Assuming that the fiscal constraint ties. However, low income per capita is also impor- remains binding, piecemeal reforms along the fol- tant in its own right because that heightens the lowing three lines appear feasible: vulnerability to falling into poverty after even a minor decline in income. On these criteria, the 1. extending contributory social insurance schemes Northwest is particularly vulnerable. The region is to regular wage employees regardless of personal reliant on crops at a disadvantage against imports, registration status, which is losing significance such as corn, wheat, and coarse grain; it has a very because of changes in the household registration low income relative to the national average, and its system rural economies are less diversified. Moreover, the 2. broadening the coverage of means-tested social shortage of water in the region limits the possibilities assistance schemes targeting poor households, of changing the cropping pattern. The fact that the including the introduction of an integrated so- reduction in barriers to agricultural imports would cial assistance scheme covering urban and rural be uneven and heavily concentrated in some regions populations in selected localities with healthy that are already poor suggests the need for introduc- public finances ing assistance schemes that target particular regions. 3. overhauling the scheme, financed by the central The second oddity about the Chinese economy is government, to assist poor rural localities in the wide divergence between the sectoral distribu- order to extend coverage to localities seriously tion of the labor force and the division of the popu- harmed by competition from agricultural im- lation into agricultural and nonagricultural (or ports, and introducing a similar scheme to assist urban and rural). The percentage of the labor force towns and cities caught in a spiral of decline or engaged in farming is 47 percent, but the percentage in a stagnation trap. of the population/labor force designated as rural is 64 percent, 17 percentage points higher. The diver- The first would involve little, if any, extra expen- gence has widened over the reform period because diture, because employees brought under the social of the rapid growth of TVEs and the correlated insurance net will be making advance contribu- expansion of nonfarm employment in rural areas. tions toward benefits. Such an extension will be The divergence matters because the coverage of largely limited to regular wage employees because social security schemes is determined by the sta- of difficulty in collecting contributions and defin- tus division nonagricultural­agricultural, not the ing unemployment. The main groups brought sectoral distribution. As a result, two individuals under the umbrella of unemployment insurance 318 China and the WTO would include migrants in regular employment long-term migrants. Correlated with this, migrants and regular wage employees of TVEs. Some cities differ in their official status. A small percentage of and provinces already include these groups in so- migrants receive a new household registration and cial insurance. For example, Guangdong Province become indistinguishable from the local popula- extends social insurance to migrants. tion. A proportion of them, usually long-term The second recommendation covers both the migrants, have temporary household registration. MLSI for the registered population and the assort- One set of social security schemes that covers all ment of social assistance schemes in rural areas. A migrants regardless of differences among them is change in the coverage of MLSI to include all long- neither appropriate nor feasible. That would term urban residents regardless of their registration require an integrated social security system cover- status will remove glaring anomalies in the way the ing both the urban and rural population, which, as scheme currently functions. The suggested change argued earlier, is not realizable is the foreseeable in the coverage will mean a greater expenditure, but future. There is no doubt that for the purposes of it is likely to be only a small fraction of the current social security long-term immigrants should be expenditure on the scheme. In rural areas, the pri- regarded as part of the urban population. However, ority is to institute a national framework for the for short-term migrants, it is far from obvious provision of social assistance to poor households in whether they should be regarded as part of order to set a minimum standard and put in place a the urban or the rural population. This decision mechanism to ensure that at least the minimum is takes on a special importance in China, because, reached. Some provinces are taking steps to abolish according to the current official practice, the the urban­rural segmentation in social security. For urban poverty line, 1,800 yuan per person per year, example, Zhejiang Province has decided to extend exceeds the rural poverty line, 635 yuan, by a very the MLSI scheme, which had been confined to its wide margin.20 Therefore, a person who is poor urban population, to its rural population as well. according to the urban poverty line would be well Similarly, Shanghai is planning to introduce an off by the rural one. integrated social security system for the whole The disadvantageous position of migrants is population--both urban and rural. associated with the control on population migra- The third recommendation is based on the tion, especially between rural and urban areas. De- assumption that an assistance scheme should target pending on the locality, immigrants are restricted particular localities, thereby supplementing the to jobs and occupations that are not attractive to social assistance scheme that targets poor house- local residents. Furthermore, immigrants do not holds. As noted earlier, the current scheme to assist have the same access to public facilities, such as poor localities is of little use, if any, in assisting housing and schools, as local residents. With a few rural localities adversely affected by WTO acces- notable exceptions, migrants in cities are excluded sion. Furthermore, there is a strong case for intro- from social insurance.21 The notable exceptions ducing a similar scheme for towns and cities caught are Guangdong Province and, to a lesser degree, between the pincers of high unemployment and Shanghai. The former treats migrants on a par with poverty rates, on the one hand, and strained public the local population for the purposes of social finances, on the other.19 insurance, and the latter has begun to include The inclusion of migrants in social insurance migrants in particular social insurance schemes. and MLSI is of particular importance, because The exclusion of migrants in cities from social labor migration is one of the principal ways of insurance is not restricted to migrants from rural adapting to changes in the structure of the econ- areas, but also extends to migrants from other cities omy, including that induced by trade liberalization. because entitlements to social insurance benefits In the Chinese context, migrants form a heteroge- are not portable across cities. This arrangement neous group, but they differ in their length of stay. reduces the benefits migrants, including those from Some are visitors or short-term migrants, and some another city, can derive from participating in social have resided in China for a substantial length of time. insurance schemes. Six months is the cut-off period used in Chinese The official control on migration remains, but statistics to distinguish between short-term and its rigor has diminished considerably. According a Coping with and Adapting to Job Losses and Declines in Farm Earnings in China 319 recent State Council circular, migration to county migrants because of the absence of social insurance and some small or medium-size towns has been schemes in rural areas. completely freed, in principle if not yet in practice. The realistic options for extending social insur- Moreover, the last few years have seen the introduc- ance to migrants have to take into account two tion of assisted migration out of rural localities that considerations. First, restrictions on rural­urban have little or no prospect of emerging from poverty. migration, though considerably relaxed, are likely However, while the focus of government policy and to stay for some time, especially for large cities. Sec- discussion has largely been rural­urban migration, ond, urban and rural social security systems are urban­urban migration is assuming increased im- likely to remain segmented for some time. It would portance with the rise in urban unemployment and then be inappropriate to apply one formula to all layoffs. migrants. It is important to distinguish between Two aspects of the participation of migrants in social insurance schemes that (1) cover all migrants social insurance schemes are the collection of con- and (2) are confined to particular categories of tributions and the disbursement of benefits. The migrants, such as those from other cities. The first problems raised by the first are, in principle, the set of schemes covers contingencies that seriously same for migrants and residents. The only differ- reduce a person's capacity to work such as physical ence is that, compared with residents, migrants are injury and major illnesses. The implication is that more likely to be in jobs outside the coverage of the priority should be to extend disability compen- social insurance schemes, such as casual or self- sation and insurance against large health care employment. The extension of social insurance expenditure to all migrants. In these two sets of schemes to the labor force in the informal sector, schemes, benefits are not deferred but run concur- which includes not only migrants but also increas- rently with the payment of contributions. The issue ing numbers of local residents, would pave the way of the portability of benefits does not arise. The for wider coverage of migrants. But the aim should second group of schemes includes old-age pensions be not only collecting contributions but also dis- and unemployment compensation. These schemes bursing benefits in return for contributions. Unlike involve deferred benefits and raise the issue of collection, disbursement raises issues that can be portability, which can be solved through a clearing solved only through some fundamental changes in mechanism. These are suitable for extension to the organization of social insurance. Aside from city­city migrants, because all cities are covered by immigrants who receive a change in residence sta- similar social insurance schemes. tus, the rest are by definition temporary residents. This definition poses a particular problem for Education and Training social insurance schemes, such as old-age pensions and unemployment insurance, in which contribu- Between 1978 and 2000, the portion of the labor tions and benefits are separated in time. The prob- force employed in the primary sector, most of lem arises because current social insurance schemes which is farming, fell at an average rate of over are largely organized and financed at the level of 1 percentage point per year. This rate, although cities. Arrangements for the transferability of bene- twice that in the planned economy period fits between cities even in the same province are (1952­78), would have been even higher except for nonexistent or rudimentary. An integration of three factors: (1) the household registration system, social insurance at the provincial level would make (2) the land tenure system that provides little or no benefits transferable within a province, and it compensation for giving up use rights on allocated would go a long way toward addressing the prob- land, and, (3) the bias in favor of industry, which lem of transferability because most migration is has tended to raise the cost of transferring labor confined to one province. Such integration is from farming to nonfarming activities. However, as planned, but it will take some time to implement. noted earlier in this chapter, all these factors are The portability of social insurance benefits also undergoing radical changes. Taking all the changes makes sense between urban areas that have com- together, the implication is that the transfer of parable social insurance schemes, but it begs the labor out of farming over the 20 years to 2020 question of relevance in the case of rural­urban would be at least as rapid as that seen over the 320 China and the WTO 20 years to 2000. Added to this, driven by a more education in many rural localities falls far short of open international trade regime spurred by WTO the prescribed nine years of schooling--in fact, membership and economic growth, the Chinese about 26 percent of primary schools have a cycle of economy is likely to experience a more rapid redis- less than six years. Moreover, a significant percent- tribution of labor between occupations and sectors. age of children in poor localities drop out before In the Chinese context, the methods for adapting to completing the full cycle. The tenth five-year plan the changes and reducing their cost in terms po- (2000­04) envisages that, by the end of 2005, at verty and increased economic inequality fall into least 90 percent of those leaving school will have three broad categories: had nine years of education. The target date for achieving universal basic education (UBE) is set for 1. improving the employability of the labor force 2010. However, according to the government's plan, by raising the general level of education and the UBE does not mean a uniform length of school- skills, especially in the rural population ing for all students. Depending on local financial 2. facilitating the growth and proliferation of pri- resources and organizational capability, the UBE vate enterprises will actually mean less than nine years of schooling 3. encouraging the adaptability of existing enter- in some localities. For the poorest 5 percent of the prises. population, all in the western provinces, the UBE will mean no more than three to four years of The remainder of this section focuses on the first schooling. point, education and skills. The literacy rate in With respect to post­basic schooling, the period China is substantially higher than that in many since 1995 has witnessed a sharp rise in the gross larger developing economies, especially low-income enrollment rates in secondary schools (including ones. However, the literacy rate is no longer an vocational and regular schools) and in tertiary appropriate indicator of human development, nor institutions. This trend is expected to continue. The are low-income economies relevant comparators in tenth five-year plan (2000­04) envisages that the field of human development for China. Thanks almost 100 percent of children in cities and towns to the high growth rate of its per capita income after will go on to senior secondary and vocational 1978, China graduated to lower­middle-income schools after nine years of basic education. It is status at the end of 2000. The appropriate reference expected that by 2005 the percentage of the relevant group for setting targets for human development age group in a tertiary institution would rise to and social security is not the present-day low- 15 percent, from less than 10 percent in 2000. These income economies but the middle-income ones. trends will help to raise the average number of In terms of the average number of years of basic years of education to the level in East Asia, but they schooling for the current cohort of children, China also will widen the gap between urban and rural lags behind many Asian countries, such as the areas, which has some adverse implications for the Democratic People's Republic of Korea, Malaysia, future. Thailand, and Sri Lanka, as well as the Republic of What factors should be taken into account in Korea, Singapore, and Japan. Furthermore, as China deciding the appropriate general level of education narrows the technology gap with developed econ- for the population? Broadly, educational attain- omies, the educational attainment of the labor force ment influences labor productivity directly and the is assuming growing importance in determining the productivity of capital indirectly. An important growth rate of the economy. feature of investment in formal education is that In China, nine years of schooling for all children much of it precedes workers' entry into the labor has been a policy goal since 1986, but it has not yet force, and additional educational investment after been attained. The target is largely achieved in the entry in the labor market is, for most individuals, urban areas, but is far from realized in rural areas, small. Moreover, learning through the work experi- which is home to 64 percent of the population and ence is crucially dependent on educational attain- an even higher percentage of those in the 15­16 age ment prior to entering the labor force. The cohort. In the western and central regions, basic implication is that investment in education should Coping with and Adapting to Job Losses and Declines in Farm Earnings in China 321 be decided with reference to the likely course over trade to GNP ratio of about 45 percent, the Chinese the whole working life and to some sections of the economy was on the eve of its WTO accession as population, especially poor rural communities, open as any large economy.22 because of the high cost of education relative to On the social side, China has also been remark- current income. ably successful in reducing poverty. The official A major determinant of labor market events headcount of the rural poor dropped from about over an individual's working life is the changes in 250 million in 1978 (31 percent of the rural popu- the skills or the education level needed to perform lation) to 32 million at the end of 2000 (3.5 percent work. Assuming a working life of between 40 and of the rural population). The headcount of the 50 years, most of the current cohort of school-age poor and the poverty rate are highly sensitive to the children in rural areas either would not be poverty line. Both of these figures are much higher employed in farming or would switch to a job in when the $1-a-day poverty line is used, but the industry or services fairly early on during the reduction in poverty still remains dramatic. Unfa- course of their working life. The implication is that vorable developments over the last decade or so the requisite level of education for rural children is include a sharp rise in interpersonal inequality in not needed currently to get by with life in rural both urban and rural areas and a widening urban- areas and work in farming, but that it would be rural disparity. Interprovincial inequality remains needed in the future to live in an urban environ- large, though falling in recent years. The urban ment and take on jobs in industry or services. unemployment rate has been rising since the mid- Viewed this way, the wide gap in educational attain- dle of the 1990s and is high by historical standards. ment between urban and rural areas (especially the In recent years, rural household incomes have risen backward ones) is a serious deficiency that should very slowly and have even fallen in some localities. be corrected quickly through a concerted policy. All the studies of the impacts of China's acces- Comparatively low educational attainment among sion to the WTO agree that the overall benefits out- children in backward rural areas will in time act as weigh the losses. Although in the aggregate the an impediment to their out-migration or reduce population would benefit, losses and gains would their chances of getting or holding jobs when they not be evenly distributed, and there is little doubt migrate. Either way, comparatively low educational that there would be a significant number of losers, attainment among a group of children or a time lag at least in the short run. These will include workers in achieving the target of nine years of basic educa- displaced by competition from imported manufac- tion for all has adverse implications for reducing tured goods and farmers suffering a decline in poverty or income inequality in future. The policy income because of cheaper agricultural imports. implication is the importance of reducing the cost Although it is difficult to foretell the losses pre- of basic education relative to average income for cisely, evidence from the United States help in families in rural communities. forming a rough idea of the magnitude. That evi- dence suggests that workers losing jobs because of competition from imports on average endure a sig- Conclusions nificant spell of unemployment and also experience On the eve of its entry into the WTO, China pre- a decline in earnings when they find another job. sented a mixed picture of notable successes and Arguably, the losses suffered by Chinese workers some adverse developments. The Open Door who lose jobs because of WTO entry would be no policy, first introduced in 1979, has been a remark- less than those experienced by U.S. workers unem- able success. It has brought huge benefits and a few ployed as a result of competition from imports. problems. China has faced no major balance of This point raises the issue of compensation for payments or currency crises. In the 24 years from workers and farmers adversely affected by WTO 1978 to 2002, China's exports have grown at an entry. What should be the magnitude of compen- average annual rate of 15 percent (in dollar terms), sation relative to losses suffered? And what schemes which points to its ability to adapt to and thrive in or programs should provide the compensation? the face of foreign competition. With a foreign Full compensation for losses is ruled out on two 322 China and the WTO grounds. First, it would be very costly to verify the poverty and cannot protect households from the true magnitude of losses, because full compensa- lower earnings that might result from competi- tion would provide a huge incentive to exaggerate tion by agricultural imports. losses. Second, even when losses can be verified accurately, the opportunity cost of compensation The crucial question is how effective would the may be prohibitive. On the grounds of feasibility, existing social security schemes be in preventing then, any compensation has to be partial, which workers who have lost jobs and farmers who have raises the issue of the goal of compensation. The experienced a decline in income (including their minimum goal for any compensation has to be the families) from falling into poverty, which is the prevention of poverty, a term that can be defined minimum expected of any system of social protec- in a number of ways. For the present purposes, tion? The answer depends crucially on whether the poverty is defined as "inadequacy of command over person is, officially, urban or rural. The (registered) economic resources," with "inadequacy" calibrated urban population is covered by unemployment to take account of personal or household character- insurance, subject to the payment of the requisite istics such as old age, ill health, or schooling age. A contribution, and the Minimum Living Standard broader aim of compensation is to avoid relative Scheme. Together, these schemes provide a fair poverty, which may be defined as a percentage (e.g., degree of protection against poverty, although 50 percent) of average or median income per capita. there are serious problems in the coverage of the These goals remain valid regardless of the proxi- health care and schooling costs of low-income mate cause of poverty. Adverse events such as a job urban households. In contrast, the provisions for loss or a decline in wages or in farm earnings can be income maintenance of the rural population are produced by a wide range of causes, both internal sparse and patchy. The employees of TVEs are and external. There is no general justification for excluded from unemployment insurance despite special measures to deal with the adverse impacts of the fact that they number 128 million, which is China's WTO membership. However, supplemen- about a quarter of the rural labor force. The rural tary measures may be needed if the existing ones social assistance schemes, which have goals similar cannot cope with adverse impacts concentrated to those of the MLSI, vary widely in coverage, ben- over a short time period or in particular regions, efits, and operation. They do provide a social safety such as the rural areas of the already impoverished net, but one that is ridden with tears. As presently Northwest. constituted, the scheme to assist poor rural locali- China has a range of schemes for income main- ties is not relevant because it is directed mostly at tenance and poverty relief, and most of these share remote localities with little connection to the out- three characteristics: side world, although the scheme can be altered to cover rural localities seriously harmed by competi- · a stark dichotomy between the urban and rural tion from cheaper imports. However, there is at population present no scheme to assist towns and cities caught · a goal of doing little or no more than alleviating in a high unemployment and stagnation trap. absolute poverty, a feature that applies especially Assuming that the aim is to close the major gaps to the rural schemes in social protection against the threat of poverty, · a high degree of decentralization in financing the priorities in urban and rural areas are different. and organization, which goes beyond the level In urban areas, the priorities are to alter the cover- consistent with equity and maintaining varia- age of existing schemes, to institute a financing tion in provision within socially acceptable arrangement that ensures an equitable distribution bounds. The income maintenance schemes for of burden across government tiers, to make the the rural population are supplemented with a contributory benefits portable, and to eliminate scheme that provides each rural household with inadequacies in benefits. As for altering the cover- a land plot. The household land plots have age of existing schemes, all the major anomalies can played a crucial role in reducing poverty and in be eliminated by extending the schemes to include keeping rural poverty comparatively low, but all long-term residents, regardless of their registra- they do not guarantee full protection against tion status. Coping with and Adapting to Job Losses and Declines in Farm Earnings in China 323 In contrast to its urban counterpart, the rural 6. The estimates of the percentage of the rural poor living population needs additional income protection in the selected counties vary over time. Recent estimates were 50 percent. schemes. First, given the sizable labor force in TVEs, 7. The program previously targeted the county level. It is there is a strong argument for extending UI to wage generally accepted that counties are too large and differentiated employees in TVEs. Second, China needs a social for targeting. The proposal is to lower the level of targeting to the village, which would reduce the numbers of false negatives assistance scheme that targets poor rural house- and positives. holds, similar to the MLSI, backed by a national 8. In the past, the Five Guarantees Scheme also applied to framework to ensure a minimum level of protec- urban areas, but since 1997 the urban Minimum Living Stan- dard Scheme (MLSS) has superseded it. tion and equitable distribution of financial burden 9. As one might expect, the headcount of the poor is highly across government tiers. Third, the assistance sensitive to shifts in poverty. However, the observation that rural scheme for rural localities needs to be recast so that poverty fell dramatically during the first 10 years of economic it covers a wider range of disadvantaged localities. reforms remains valid even when a poverty line higher than the official one is used. There is a broad consensus on the value of rais- 10. The actual figure depends on the definition of the land ing the general education level and investing in area (numerator) and of the denominator. Here, the land area skills, especially of the rural population, in order to is taken to be the sown area in 2000, and the total rural popula- tion serves as the denominator. The area of land per person is mitigate the adverse impacts of WTO entry on higher if the labor force employed in farming is used as the labor and thus the population. However, these goals denominator. are not panaceas and they have their limitations. As 11. These 12 counties include 6 within the administrative boundaries of Suzhou City and 3 each within the boundaries of pointed out earlier, most investment in education Wuxi and Changzhou City. and much of the acquisition of skills take place 12. In principle, land plots are leased for 30 years, but the before workers enter the labor market. Therefore, actual duration in many localities is shorter than 30 years. raising the level of basic education has little rele- Furthermore, at any particular point in time the remaining lease is generally a lot less than 30 years. vance for improving the employability of those 13. In considering the implications of the redeployment of who are already in the labor force, especially older labor to a household land plot as a way of coping with a job loss, workers. Raising the level of basic education should one should also take into account any hired labor displaced as a result. be viewed as a long-term measure that would be 14. About 86 percent of China's 1.266 billion population in complemented by short-term measures, such as 2000 was living within the boundaries of cities and towns. retraining adapted to the immediate circumstances 15. This situation arises largely from the expansion of city districts and the associated redesignation of some rural counties of adversely affected workers. as urban districts. 16. This is the GDP share of the primary sector, which also Notes includes activities such as quarrying. 17. If the share of farming in GDP is and in employment is 1. That is, those with urban registration. Some rural social , then labor productivity in farming relative to the average, insurance schemes are in place, but none provides an unemploy- which by definition is 1, is given by (/). Similarly, the ratio of ment benefit. labor productivity in farming with respect to that in nonfarming 2. According to the regulation, social insurance applies in activities is (/) * {(1 - )/(1 - )}. principle to the whole of the urban labor force of 247.8 million 18. This is based on the assumption that nonfarming activi- (34 percent of the total), but the actual coverage falls well short ties are confined to urban areas and farming activities to rural of that prescribed because of difficulties in collecting contribu- areas. The first part of the assumption is not strictly true for tions. Some cities extend unemployment insurance to migrant China. As a result, the urban­rural gap would be less than four workers in regular wage employment. and a half times. 3. The supplementary condition is usually the obligation to 19. The latter matters because social insurance and social take offered employment or participate in a training program. assistance schemes are budgeted at the city level. The plan is to 4. This situation has begun to change. At the end of 2003, 11 raise the budgeting to the provincial level. However, the plan has provinces had some form of a minimum living standard scheme yet to be fully implemented. for the rural population, which creates the possibility of institut- 20. The rural and urban lines, especially the latter, can vary ing one social safety net covering the whole population. On a across localities. However, the fact of a wide gap between the similar note, Shanghai is planning to set up a unified social secu- urban and rural poverty remains valid. rity system covering both the urban and the rural population. 21. Social insurance covers the old-age pension, health care 5. The "8" in the "8-7 Plan" refers to the eighth five-year plan, insurance, unemployment benefits, disability compensation, which was under implementation in 1994, and the "7" refers to and maternity benefits. the seven years from 1994 to 2000. The plan expired at the end of 22. The ratio depends crucially on how GDP is measured. the ninth five-year plan and has since then been replaced with However, the general proposition about the openness of the the 2005­2010 Rural Poverty Reduction and Development Plan Chinese economy remains valid regardless of how GDP is with the same aim. measured. 324 China and the WTO References NBS (National Bureau of Statistics). 2001. China Towns and Cities Statistics Yearbook 2002. Beijing: China Statistics Press. Collins, Susan M., ed. 1998. Imports and Exports and the ------. 2003a. China Labor Statistical Yearbook 2003. Beijing: American Worker. Washington, D.C.: Brookings Institution China Statistics Press. Press. ------. 2003b. Statistical Yearbook of China 2003. Beijing: China Ehrenberg, R. G. 1994. Labor Markets and Integrating National Statistics Press. Economies. Washington D.C.: Brookings Institution Press. Wood, A. 1994. North-South Trade, Employment and Inequality-- Fishlow, A., and K. Parker, ed. 1999. Growing Apart--The Causes Changing Fortunes in a Skill-Driven World. Oxford: Oxford and Consequence of Wage Inequality. New York: Council of University Press. Foreign Relations. Kletzer, L. G. 2001. Job Loss from Imports: Measuring the Costs. Washington D.C.: Institute for International Economics. Index accession of China, 1, 21­28 gradual opening and, 82­85 approval and acceptance, 24 inland areas and, 90­94 background prior to, 21­24, 264 integration tests, 91­93, 93t6.7 bilateral negotiations, 22­23 price transmission coefficients, 93­94, 93t6.8 economic impacts of, 99­236 income of farmers, 5 See also economic impacts of accession compared to urban income, 1, 12, 107, 108t7.8, 213 future implications of, 15­16, 26­27 decrease in, 110, 111t7.10, 220 household and poverty impacts of, 237­324 education and training, effect of, 319­21 See also poverty impacts of WTO accession; welfare increases under Five-Year Plan, 69­70 impacts of WTO accession need to increase, 71­73 multilateral steps of, 23­24 poverty level of. See poverty impacts of accession policy reforms and, 19­97 regional declines in, 72, 72t5.3 See also specific sectors (e.g., agriculture) sources of, 71, 71t5.2 Protocol and Working Party Report, 24, 107 spending reductions due to, 72­73, 72t5.4, 73t5.5 WTO's basic principles and, 24­26 infrastructure improvements and, 113 Administration for Industry and Commerce (AIC), 61 land tenure system, 113, 314­15, 323nn12 & 13 Agreement on Government Procurement, 26 local advantages in, 76 Agreement on Textiles and Clothing. See textiles nominal rates of protection (NRPs) and. See nominal Agreement on Trade-Related Aspects of Intellectual rates of protection (NRPs) Property Rights. See TRIPS optimization of varieties and quality of products, 75 agriculture Outline for the Development of Agricultural Science accession impacts on, 5­13, 101­16, 219 and Technology in the Next 10 Years, 74 background developments since late 1970s, 102­5 poverty of agriculture-based regions, 112 cap on domestic support for, 24, 26 prices, compared to nonfarm products, 109­11, China National Rural Survey (CNRS) on, 92 110t7.9 comparison of income, employment, and food output processing, expansion of, 75­76 non-accession vs. post-accession, 112, 113t7.11 production increases, 69, 70t5.1, 103t7.1 regional, 107, 108t7.7, 115n10 research and development in, 113 Decisions of the Communist Party of China Central restructuring of, 5, 73­78, 101 Committee on Major Issues of Agriculture and self-sufficiency, 104, 112, 114, 114t7.12 Rural Work, 73 social assistance and. See social assistance decollectivization of, 81 spending on food, 75, 75t5.6 development and policy, 5, 69­79, 106­7 tariffs and quotas, 79, 107, 115n2 environmental issues and, 78 taxation and, 84, 85t6.3, 95n1 exodus to nonagricultural jobs. See labor mobility township and village enterprises (TVEs) and, 76, 102 Five-Year Plan for the Development of the National trade and, 79, 83, 83t6.2 Economy and Society (ninth), 69­71 liberalization, 84­85, 106 government subsidies for, 84­85, 85t6.3, 113­14, rates of protection, 83, 83t6.1 215, 221 sanitary and phytosanitary measures, 115n1 "green barriers," 79 transient employment for farmers, 76­77 GTAP (Global Trade Analysis Project) model, AIC. See Administration for Industry and Commerce application of, 14, 102, 107­9, 115n5 air transport, 152­53 hukou system, effect of. See hukou system for urban See also logistics sector residence antidumping and safeguards, 3, 15, 29­47 import protection measures, 6­7, 6t1.1 current situation under, 3, 8, 30­33 incentives and distortion, 5­7, 81­97 measures in force against China, 22, 32, estimates of nominal rates of protection, 85­90, 32t3.3, 46n2 87t6.4­88t6.5, 95 sensitivity to "antidumping-intensive" sectors, 35 325 326 Index antidumping and safeguards (continued) China Health and Nutrition Survey (CHNS) targets and measures in force, 30­32, 31t3.2, data used in income and labor study, 242­43, 33, 46n2 244t14.1, 259n4 users and measures in force, 30­32, 30t3.1, China Insurance Regulatory Commission (CIRC), 183 33, 46n2 China Mobile Communications Corporation, 159, dispute settlement use of, 3 162, 163 EC and U.S. echoing cases, 36, 37t3.5 China National Rural Survey (CNRS), 92 enforcement challenges for China, 39­40 China Netcom, 162, 170 measures in force (as of June 2003), 39, 40t3.6 China Netcom Communications Group Corporation, 170 exposure to foreign antidumping, 33­36 China Satellite (ChinaSat), 159 Chinese policies to minimize, 35­36 China Securities Regulatory Commission (CSRC), 183 Harmonized Tariff System (HTS) and, 33­35, China Telecom 34t3.4, 43 break up of monopoly of, 158­60 linking liberalization to better treatment of exports by lowering of long-distance rates and, 162 antidumping users, 42­43 mobile market and, 179n17 negotiations on contingent protection, 41­45 monopoly of, 157, 158 non-market economy (NME) status, use of, 29, profitability of, 162 42­43, 42t3.7, 46n4 restructuring as of 2002, 170 regulations China United Telecommunications Corporation. adoption of, 26, 38­39 See Unicom stricter rules, proposal for, 4, 29, 43­44 CHNS. See China Health and Nutrition Survey (CHNS) retaliatory approach, Chinese use of, 3­4, 39, 46 data used in income and labor study Russia and, 46n1 CIRC (China Insurance Regulatory Commission), 183 safeguard, defined, 29 clothing. See textiles sectors that are "antidumping-intensive," 33­35, CNRS (China National Rural Survey), 92 34t3.4, 46n2 competition steel safeguard measures, 27 government procurement, fair competition in, 26 transitional product-specific safeguard (TPS), 29, intellectual property rights and, 57, 62, 66 44­45, 46n2, 216 in logistics sector, 151 apparel. See textiles in services sector, 132 Argentina and antidumping, 4, 30 in telecommunications, 171­72 AT&T, 178n11 growth and competition in subsectors, 162­64 automobiles. See motor vehicle sector reforms, effect of, 173 averaging tariffs, 35 computable general equilibrium (CGE) model, application of bankruptcy, 27 economic impacts of accession, 211, 289­92 banks. See financial services motor vehicle sector, 197 beverages welfare impacts, 262 import increases, 224 computer industry, 121, 123t8.2 price reduction, 219 See also technology tariff reduction tariff reductions for, 8, 15, 215 software protection, 55, 57 trademark protection, 55 construction and engineering services, 121, 124t8.2, 127 bilateral negotiations for accession, 22­23 consumer information, adequate provision of, 9, 132­33 Brazil cotton antidumping and, 30 export subsidies for, 84­85 intellectual property rights and, 55 NRP for, 88, 94 protection for, 88t6.5, 89 cable and convergence issues, 173 tariff reductions for, 15 Central Working Committee on Rural Affairs, 73, 74 courier services, 155 cereals, 23, 26 cross-border supply, 118 CGE. See computable general equilibrium (CGE) model, application of developing countries Chile and universal access antidumping and, 30, 46n3 in services sector, 133 China's accession with provisions for, 23­24, 26 in telecommunications, 175b10.3 preferential treatment for, 2, 4 China Customs data, 212 dispute settlement, use of, 3, 27 Index 327 distribution services, 124t8.2, 127, 155n3 Fair Trade Administration for Import Doha Development Agenda, 7, 15, 25, 29 and Export (FTA), 39 drought, 71 farmers. See agriculture drugs. See pharmaceuticals FDI (foreign direct investment). See foreign dual pricing practices, 2, 25 investments financial services, 10­11, 181­90 earnings. See income levels accession impacts on, 15 E-commerce, 169, 172­73, 178n14 compared to other East Asian countries, 189n9 See also Internet foreign banks, 182, 185, 189n12 economic impacts of accession, 1, 5­13, 211­36 GATS commitments and, 125t8.2, 127, 138n16 See also specific economic sectors (e.g., agriculture, initial conditions of, 182­85 financial services, etc.) issues needing to be addressed, 187­89 assessment of, 219­23, 295, 296t16.5 nonperforming loans (NPLs), 184 on China, 213­15, 219­21 policy banks, 185 on China's trading partners, 215­17, 221­23, 224 preaccession situation, 183­85, 183t11.1 complementary policy reforms, 223­24 reform, 182­83 experimental design for evaluation of, 217­19 sequencing of, 134­36 changes in China's key economic indicators, rural system, 184 220t13.5, 231t13.A.5­233t13.A.7 state banks, 10, 184­85 projected growth, 217, 218t13.4 survival of, 187­88, 188b11.1, 189n8 export processing arrangements, 212 WTO commitments in, 186­87, 186t11.2 intellectual property rights and, 54­60 First Automobile Works (FAW), 193, 197 labor market policies. See labor market Five Guarantees Scheme, 313, 323n8 labor mobility and. See labor mobility Five-Year Plan for the Development of the National market access and (1997), 286, 287t16.2, 289 Economy and Society (ninth), 69­71 methodology for analysis, 212­13 fixed-line sector, 162 rural development. See rural economy food sectoral effects, 109­11, 110t7.9, 219­21 See also agriculture; specific products See also specific sectors (e.g., agriculture) consumption trends, 96n4 skill levels and. See labor market import increases, 224 trade reform, 81, 213­17 population growth and, 70 See also trade reform tariff reductions for, 215 Edge2net Inc., 178n11 trademark protection, 55 education and training trade structure of, 104, 105t7.5 cross-border delivery of, 124t8.2, 127 food processing, 75­76 Cultural Revolution, effect of, 207 foreign exchange market, 27 for farmers, 319­21 foreign investments income level and, 246, 247 accession impacts on, 226 on intellectual property rights, 49, 61, 65 growth in, 283 for logistics services, 148 in intellectual property rights, 56 nonagricultural vs. agricultural work and, 259, 285 in services sector, 120, 131, 215 skill level increases as result of, 223­24 in telecommunications, 10, 158, 167­68 8-7 Plan, 313, 323n5 forestry, 78 electronics. See technology tariff reduction forwarding, 154­55 El Salvador, 22 See also logistics sector employment. See labor market; labor mobility franchising. See distribution services engineering services. See construction freight services. See logistics sector and engineering services FTA (Fair Trade Administration for Import and environmental issues and agriculture, 78 Export), 39 European Community antidumping GATS (General Agreement on Trade in Services) "echoing" against China, 36, 37t3.5 China's commitments to, 118­28, 120t8.1 position on, 29, 46n5 comparison over time, 121, 122t8.2 bilateral agreement with, 22 liberalization on faster schedule, 117 European Union patent protection, 66n6 by subsector, 119­28, 122t8.2, 137t8.A.1 exports. See trade reform See also specific subsectors (e.g., financial services) 328 Index GATS (continued) 8-7 Plan and, 313, 323n5 country commitments by mode, 118­19, 119f 8.1 factors affecting, 239 logistics sector and, 10 farmers. See agriculture WTO reference paper on basic communications, gain/loss after accession, 269, 270f15.3­272f 15.8, 273 10, 167, 174 gap between poorer and richer households, 240, 247, responding to, 168­69 248t14.4, 258 summary of, 177­78 income generation functions, determination of, GATT (General Agreement on Trade and Tariffs) 243­48, 245t14.2­246t14.3, 258 China and, 2, 12, 21 land plot as insurance, 314­15, 323nn12 & 13 principles of. See specific principle maintenance schemes, 311­14 (e.g., nondiscrimination) Minimum Living Standard Insurance (MLSI), GATT Working Party on China's Status, 21, 22 312­13, 318 GDP. See gross domestic product (GDP) growth per capita income by stratum, location, and vingtile, Global Trade Analysis Project. See GTAP (Global Trade 284­85, 285t16.1 Analysis Project) model, application of survey sample data for study of, 242­43, 244t14.1 government procurement, fair competition in, 26 types of labor income, 265, 301 government subsides. See agriculture wages and shadow wages, 240, 251­55, grain production and consumption 253t14.7­255t14.9, 259 See also specific types (e.g., wheat) India demand for and adjustment in quantity, 73­74 antidumping and, 4, 30 exports, 83 effect of China's accession on, 223 factors in successful harvests, 70­71 telecom infrastructure in, 132, 138nn8 & 12 local comparative advantages and, 76 Industrial and Commercial Bank of China (ICBC), 182 local grain bureaus, 96n2 industrial products, 7­8, 15 NRP for, 87, 87t6.4 nontariff barriers and, 226 reduction in government subsidies for, 113­14 tariff reductions for, 107, 215 self-sufficiency and, 104, 114, 114t7.12 wholesale price reductions for, 219 variances from year to year, 104, 104t7.4 infrastructure improvements Great Wall Telecom Corporation, 178n7 for agriculture, 113 gross domestic product (GDP) growth, 81 for e-commerce and Internet, 173 and rural poverty, 242 for logistics services, 148 GTAP (Global Trade Analysis Project) inland areas model, application of agricultural prices and, 90­94 agriculture, 14, 102, 107­9, 115n5 services sector in, 131, 138n10 defined, 234n5 inspection system, 24 economic impacts, 211, 212­13, 217 institutional implications of accession, 26­27 motor vehicle sector, 197 insurance. See financial services welfare impacts, 264­65 intellectual property rights, 49­67 See also TRIPS (Agreement on Trade-Related Aspects Harmonized Tariff System (HTS) and antidumping of Intellectual Property Rights) measures, 33­35, 34t3.4, 43, 46n2 China's record on, 4, 57­59, 58t4.2 Hong Kong competition and, 57, 62, 66 antidumping and, 36 copyright, 54, 55, 57, 58­59 Trade Development Council, 149 ineffective enforcement, 61 universal access, 175b10.3 reform, 50, 53, 58 household impacts of WTO accession, 13­15, 237­324 costs associated with, 53, 56 See also poverty impacts of accession; welfare impacts current situation in China, 60­65 of accession interview findings, 60­62 income levels. See income levels database protection, 59 hukou system for urban residence, 109, 213, 291 defensive measures used, 62 abolition of, 12­13, 14, 16, 224, 301 economic development and, 54­60 constraints on economy, 4, 56­57 ICBC (Industrial and Commercial Bank of China), 182 stimulation of economy, 4, 55­56 income levels, 239­60 education and training programs in, 49, 61, 65 See also poverty impacts of accession; welfare impacts enforcement, 49, 55, 59­60 of accession ineffective, 60­61, 65 Index 329 foreign companies and, 61­62 estimation of labor supply functions, 255­58, foreign direct investment in, 56 256t14.10 government support for inventions, 54 estimation of production functions, 250­51, licensing contracts, 56 251t14.5­251t14.6 local protectionism and, 50, 61 estimation of wages and shadow wages, 251­55 nature of, 53­54 time allocation by rural household for patents, 54, 55, 62­65 labor, 249, 259 biological, 59 policies, 212­13 increase in, 62­65, 62t4.3­64t4.7 effect of reforms, 223, 224, 225t13.8 ineffective, 61 skill levels reform, 50 accession impacts on, 307 regional statistics, 63­65, 64t4.7 increase in, after accession, 223­24, 225t13.8 software coverage, 57­58 skilled, defined, 282n12 pirating of recorded entertainment and skilled vs. unskilled in services sector, 138n13 software, 56, 60, 61 urban skilled and semiskilled labor, 301 plant varieties protection, 50, 54, 59 social protection and adaptation in, 308­14 regulatory changes due to WTO commitments, See also social assistance 50­53, 65­66 classification of schemes and measures, 308, trademarks, 54, 55, 62­65 308t17.1 counterfeit, 56 decentralization and, 312b17.1, 313 enforcement weakness, 61 differences between social insurance and social increase in number of, 64­65 assistance, 310, 310t17.2 ineffective enforcement, 60 instruments for, 309­11 reform, 50, 52­53, 59 types of labor, 265, 295 trade secrets, 54, 56, 59, 62 unemployment, 76, 225, 235n25, 306­7, 312, 317­18 types of, 54­55 labor mobility, 13, 15, 239­60 Internet barriers to, 212­13, 223, 224, 235n28 administration of, 160 farmers and, 76­77, 220, 223, 224, 235n23 competition in services for, 158, 159 hukou system, effect of. See hukou system for urban development of, 172­73 residence increased use of, 163­64, 164f10.2, 178n9 limited nature of, in rural environment, 240 regulations regarding, 169 return to agriculture, 295 IPRs. See intellectual property rights rural-urban migration, 241, 291­92, 295, 316 irrigation, 71 land agricultural plot as insurance, 314­15, 323nn12 & 13 Japan tenure system, 113, 265, 314­15 antidumping and, 36 licensing. See intellectual property rights China's accession, effect on, 224 local protectionism intellectual property rights and, 55 intellectual property rights and, 50, 61 Jiang Zemin, 74 logistics sector and, 147, 148b9.1 Jia Yu Freight Group Company, 148b9.1 motor vehicle sector and, 209n4 Jitong, 158, 159, 162, 170 logistics sector, 9­10, 141­56 joint ventures for services, 120, 131 accession implications for, 15, 145­48 direct effects of, 145­47, 145t9.1­146t9.3 labor market, 13 indirect effects of, 147­48, 147t9.4 See also labor mobility by subsector, 152­55 accession impacts on, 259, 305­8 See also specific subsector (e.g., air transport) agriculture. See agriculture competition in, 151 background of, 239 constraints on development of, 143­44 defection of employees with technical cost savings in third-party sector, 148­50, 149t9.5 knowledge, 61 definition of logistics, 142 displaced workers as result of accession, 306 demand for 3PL (specialized) services, 144, 146, 152 income and. See income levels evolution of, 142­43 level and pattern of labor supply, empirical study of, freight traffic by mode (1970-2000), 143, 143f 9.1 248­58 GATS commitments and, 126t8.2, 128 empirical specification, 249­50 human resources, 148 330 Index logistics sector (continued) local protectionism, 209n4 infrastructure and facilities/equipment, 148 modeling framework for assessing accession impact, in inland areas, 131, 138n10 197­206 less developed areas, implications for, 150­51, 152 CGE model used, 197 local protectionism and, 147, 148b9.1 data used, 197­98, 198t12.5 separation of government administration from experiments and results, 201­6, 202t12.7­203t12.8 enterprises, 147­48 export impacts, 205­6, 205t12.10­206t12.11 long-distance service. See telecommunications scale and average cost, 200, 201f12.2 Long Yongtu, 29, 43 structure of model, 198­200 value-added results, 204, 204t12.9 maize price reduction, 219 See also agriculture production border prices and, 93­94, 93t6.8 comparison with U.S., 200, 201f12.3 export subsidies for, 84­85 location of plants, 208f12.5 NRP for, 83, 88, 94 by plant (1995-2002), 194, 195t12.2 rate of protection, 7 since open-door policy adopted, 193b12.1, tariff reductions for, 15 194f 12.1 variation in price between imported and domestic, regional and sectoral breakdown of model, 88, 88t6.5 198t12.5 manufacturing. See industrial products restructuring after accession, 115n7, 221 market access, 2­3, 25 scale and average cost, 201f12.2 See also tariffs smuggling, 209n3 economic structure and (1997), 286, 287t16.2, 289 summary of developments, 193b12.1 GATS obligations, 118, 121, 130­31 tariff reductions for, 215 Marrakesh Agreement, 21 tariffs on vehicles, 26, 196, 196t12.4, 199, 199t12.6 meat exporters, 85 MPT. See Ministry of Posts and Telecommunications medicines. See pharmaceuticals Multi-Fiber Arrangement (MFA), 2, 3, 33, 283, 306 mergers of telecommunications companies, 170 multilateral steps of accession, 23­24 Mexico and antidumping, 4, 30 MFA. See Multi-Fiber Arrangement National Conference on Agricultural Science and MFN. See most favored nation (MFN) status Technology, 74 MII. See Ministry of Information Industry National Conference on Rural Work, 73 Minimum Living Standard Insurance (MLSI), National Industrial and Commercial Administration, 53 312­13, 318 national treatment and GATS obligations, 118, Ministry of Foreign Trade and Economic Cooperation 121, 129 (MOFTEC), 39 NME. See non-market economy (NME) status, use of Ministry of Information Industry (MII), 158­61 nominal rates of protection (NRPs) complaints against telecom companies to, 165 agriculture and, 85­90, 87t6.4­88t6.5, 106, September 2001 regulations, designation of, 168 106t7.6, 234n3 universal service fund, 175 greater protection cuts, 111­12 Ministry of Posts and Telecommunications (MPT), removing protection, 112 157­58, 161 challenges and issues in measuring, 95 MLSI. See Minimum Living Standard Insurance nonagricultural employment. See rural economy mobile phones. See telecommunications nondiscrimination, 2, 24, 25 MOFTEC (Ministry of Foreign Trade and Economic GATS commitments and, 129­30 Cooperation), 39 non-market economy (NME) status, use of, 29, 42­43, most favored nation (MFN) status 42t3.7, 46n4 accession and, 2, 25 nontariff barriers (NTBs), 82, 83, 84, 85 GATS obligations, 118, 129 import coverage, 213­14, 214t13.1, non-market economy (NME) status and, 43 227t13.A.1­229t13.A.3 U.S. treatment of China, 22­23 manufacturing sector and, 226 motor vehicle sector, 11, 15, 191­210 NRPs. See nominal rates of protection comparison with other countries, 191, 192t12.1 import and domestic shipment values, 194, 196t12.3 Open Door policy, 307, 321 import increases, 224 motor vehicle sector production after, 193b12.1, import share of total auto parts, 203, 203f 12.4 194f 12.1 Index 331 Organisation for Economic Cooperation and pricing Development (OECD) countries agriculture, 78 agricultural reform and, 101, 114 compared to nonfarm products, 109­11, 110t7.9 packaging services, 155 inland areas and, 90­94 Patent Cooperation Treaty, 56, 66n8 dual pricing practices, 2, 25 patents. See intellectual property rights medicines. See pharmaceuticals People's Bank of China (PBC), 136, 181, 182 reductions after accession, 219­20, 221t13.6, Peru and universal access, 175b10.3 230t13.A.4 pharmaceuticals telecommunications, 161­62 patent protection, 56­57 welfare impacts of accession and price changes, pricing of, 5, 56, 66 265­66, 266t15.1­267t15.2 trademark protection, 55 private enterprises, 241, 259n2 trade secret protection, 59 professional services, 121, 122t8.2 Philippines and intellectual property rights, 55 Protocol and Working Party Report, 24, 107 pirating of copyrighted material. See intellectual property rights Railcom, 159 plant varieties. See intellectual property rights rail transport, 154 Poland and telecommunications, 138n11 See also logistics sector policy reforms of WTO accession, 2­5, 19­97 regional comparisons poverty impacts of accession, 13­15, 16, 27, 237­324 income gap, 248 See also welfare impacts of accession income of farmers, 72, 72t5.3 agricultural income and, 285, 289 logistic services, 150­51 agriculture-based regions, 112, 242 patents, 63­65, 64t4.7 background, 283­84 rural income, employment, and food output, 107, base case projections and accession simulation, 108t7.7, 115n10 292­94, 293t16.3 research and development simulation results, 294­301 in agriculture, 113 summary of, 294t16.4 intellectual property rights for. See intellectual CGE model used in study of, 289­92 property rights economy-wide impacts, 259n1, 295 retaliatory antidumping, 3­4, 39, 46 8-7 Plan and, 313 rice Five Guarantees Scheme and, 313, 323n8 border prices and, 93t6.8, 94 gross domestic product (GDP) growth and, 242 import from Thailand, 89­90 household impacts, 242, 284­86, 298­300, 299t16.7 NRP for, 83, 88t6.5 rural households, 300f16.2 rate of protection, 7 urban households, 300f16.3 subsidies for, 85 land plot as insurance, 314­15, 323nn12 & 13 road transport, 153­54 Minimum Living Standard Insurance (MLSI), See also logistics sector 312­13, 318 rural economy model used in study of See also agriculture; township and village computable general equilibrium (CGE) model, enterprises (TVEs) application of, 289­90, 289­92 8-7 Plan and, 313, 323n5 household behavior, 290­91 income levels, 239­60 household migration, 291­92 labor market in. See labor market production, exports, and imports, 292 land plot as insurance, 314­15, 323nn12 & 13 per capita income by stratum, location, and vingtile, multiple occupations of workers, 242 284­85, 285t16.1 nonagriculture employment, 241­42 production, trade, and protection, 286­89 regional comparison of income, employment, and rural nonfarm poverty, 112 food output, 107, 108t7.7, 115n10 sector impacts, 295­98, 297t16.6, 298f16.1 restructuring of, 74­78 social assistance and. See social assistance small town development, 76­77 special measures following accession, 322­23 social assistance and income maintenance schemes, predictability. See transparency and predictability 309­14 preferential treatment for developing countries, telecommunications, 158 2, 4, 26 Russia and antidumping, 46n1 332 Index safeguards. See antidumping and safeguards special measures following accession, 322­23 satellite communications, 159 types of, 309­14 schooling. See education and training South Africa SCIO. See State Council Information Office antidumping and, 4, 30 securities. See financial services telecommunications, 134 services sector, 2, 8­11, 117­39 universal access, 175b10.3 adjustment costs, 133­34 soybeans business scope for service, 120 NRP for, 83, 88, 88t6.5, 94 comparison with other WTO members, 118­19, tariff reductions for, 115n2 119f 8.1 value-added tax on, 84 competition in, 132 variation in price between imported and domestic, computer and related services, 121, 123t8.2 89, 89t6.6 construction and engineering services, 121, State Council Information Office (SCIO), 132, 160, 169 124t8.2, 127 State Economic and Trade Commission (SETC), 39, 158 consumer information, adequate provision as priority, State Information Leading Group (SILG), 160 9, 132­33 State Market Administration Bureau (SMAB), 91 distribution services, 124t8.2, 127 steel safeguard measures, 27 educational services. See education and training stock exchanges, 182 financial services. See financial services See also financial services form of establishment, 120, 131 subcontracting in logistics sector, 148­50, 149t9.5 GATS framework and commitments, 118­28 subsidies for agriculture. See agriculture, subheading: See also GATS (General Agreement on Trade government subsidies for in Services) sugar geographic scope of service, 120 NRP for, 88, 94 grandfather provisions for, 127­28 price reduction, 220 import increases, 224 protection for, 88t6.5, 89 liberalization after accession, 115n7, 221 tariff reductions for, 15 eliminating barriers to entry, 2, 26, 130­31, 215 eliminating discretion, 128­29 Taiwan eliminating discrimination, 129­30 China's accession, effect on, 217, 221, 224, 225, 234n1 eliminating foreign ownership restrictions, 131, 215 counterfeit goods, regulation of, 66n5 eliminating regional restrictions, 9, 131­32 recognition of, 22 nontariff barriers, reduction of, 215 Tariff Commission under the State Council (TCSC), 39 priorities for reform, 132­36 tariffs professional services, 121, 122t8.2 agricultural products and, 15, 79, 107, 115n2 regulatory requirements for, 120, 132­36 antidumping and, 35 telecommunications. See telecommunications averaging, 35 universal service as priority, 9, 133 motor vehicles and, 26, 196, 196t12.4 SETC. See State Economic and Trade Commission reduction, 2­3, 25, 26, 27­28, 107 Shanghai Banking Association, 181 See also specific products Shenzhen China Motion Telecom, 178n11 by China's trading partners after accession, 217 small town development, 76­77 comparison of rates pre- and post-accession, 214, See also township and village enterprises (TVEs) 216t13.3 social assistance, 14­15, 305­24 experimental design to evaluate accession impact classification of population and, 315­17 on, 217, 219 classification of schemes and measures, 308, for years 1992-2001, 214­15, 215t13.2 308t17.1 structure of, 289 closing loopholes in, 317­19 taxation. See value-added tax decentralization of, 312b17.1, 313 TCSC (Tariff Commission under the State Council), 39 differences from social insurance, 310, 310t17.2, technology tariff reduction, 8, 217 323n21 telecommunications, 10, 132, 157­79 8-7 Plan, 313, 323n5 accession impacts on, 15 Five Guarantees Scheme, 313, 323n8 allocation of scarce resources, 168, 178 migrants and, 318­19, 323n2 break up of monopoly, 158­60, 159t10.1, 160t10.2 Minimum Living Standard Insurance (MLSI), cable and convergence issues, 173 312­13, 318 competition in, 164­66, 171­72, 173 Index 333 complaints against companies, 165 rural workforce employed by, 312 defining coverage of, 167, 167b10.2 for telecommunications, 158 E-commerce, 169 TPS. See transitional product-specific safeguard fixed-line sector, 158, 159, 162, 163f10.1 trade expansion, 1, 12 foreign investment in, 10, 158, 167­68 trademarks. See intellectual property rights GATS commitments and, 121, 123t8.2 trade reform, 1, 213­17 growth and competition in subsectors, 162­64 agriculture. See agriculture in inland areas, 131 China's policy changes, 213­15, 219­21 Internet. See Internet export processing arrangements, analysis of, 212 issues not yet addressed, 177 export shares, 206, 207t12.12 long-distance rates, 161­62 global trade, effect on, 224 mergers and consolidation, 170 nontariff barriers. See nontariff barriers (NTBs) mobile phones, 159, 162­63, 163f 10.1, 171­72 tariff reduction. See tariffs new entrants into, 171­72 three-year transition period for full liberalization, 24 preaccession situation, 157­61 trade partners' policies, 215­17, 221­23, 224 pricing welfare impacts of, 262­64 decline, 161­62 Trade-Related Aspects of Intellectual Property Rights setting of, 158 Agreement. See TRIPS profitability, 162 trade secrets. See intellectual property rights quality of service, 164­66 training. See education and training reforms, 165t10.3, 176 transitional product-specific safeguard (TPS), 29, 44­45, direct impact of, 170­76 46n2, 216 new law, 169, 177 transparency and predictability, 2, 3, 25­26, 118 prior to accession, 161­66 transport of goods. See logistics sector sequence of, 134, 135f 8.2 TRIPS (Agreement on Trade-Related Aspects of regulatory framework for, 166­70, 166b10.1 Intellectual Property Rights), 2, 4­5, 15 September 2001 regulations, 167­69 Chinese implementation of, 26, 49, 57 restructuring, 168, 170, 170f 10.3 substantive requirements of, 50­53, 51t4.1 revenue share by company (2001-2002), 172f 10.4 TVEs. See township and village enterprises satellite communications, 159 third-generation (3G) technology, 163, 171 undistorted trade, 2, 3, 26 township and village enterprises (TVEs) for, 158 See also antidumping universal service to poor and isolated areas, 161, 168, unemployment, 76, 225, 235n25, 306­7, 312, 317­18 174­76, 175b10.3, 178, 179n19 Unicom (China United Telecommunications value-added services, 166, 166b10.1, 176, Corporation), 158, 159, 162, 163, 178n5 178n12 Union for the Protection of New Varieties of Plants WTO commitments on, 166­67, 166b10.1 (UPOV), 50 WTO reference paper on basic communications United States (GATS), 10, 167, 174 antidumping and safeguards responding to, 168­69 "echoing" against China, 36, 37t3.5 summary of, 177­78 position on, 46n5 textiles steel safeguard measures, 27 antidumping measures and, 33 bilateral agreement with China, 22 import increases, 224 universal services quotas, 2, 3, 11, 115n7, 215, 219, 221 as priority, 9, 133 tariff reduction, 8 telecommunications, 161, 174­76, 175b10.3, 178, trade expansion after accession, 12, 215­16, 219 179n19 tobacco urban areas, migration to. See labor mobility import increases, 224 urban income. See income levels price reduction, 219 Uruguay Round restricted to state trading, 23 accession process and, 22 tariff reductions for, 8, 15, 215 antidumping and, 30 township and village enterprises (TVEs) preferential treatment for developing countries for agriculture, 76, 102 and, 26 growth of, 316 textiles and, 215. See textiles nonagricultural employment in, 241 Zero-for-Zero program, 217 334 Index value-added tax GTAP (Global Trade Analysis Project) model, agriculture and, 84, 85t6.3, 95n1 application of, 264­65 experimental design to evaluate accession impact household income, defined for, 265 and, 219 labor and labor earnings, defined for, 265 export processors and, 234nn8 & 9 land prices, effect of, 265 Vietnam, effect of China's accession on, 222t13.7, 223 measured impacts, 265­73 VoIP (Voice over Internet Protocol), 158, 164 poverty incidence, 268f15.1­269f15.2 price changes, 265­66, 266t15.1­267t15.2 wages. See income levels summary statistics on, 268t15.3 warehousing and storage, 154 sampling weights for study of, 264 See also logistics sector survey data for study of, 264 water conservation, 78 trade reform and, 262­64 water transport, 153 wheat, 83, 86­88, 87t6.4, 89­90 See also logistics sector World Intellectual Property Organization welfare impacts of accession, 261­82 (WIPO), 53 See also poverty impacts of accession Copyright Treaty and Treaty on Performances and approach to measuring, 262­64 Phonograms, 59 computable general equilibrium (CGE) model, WTO membership, 21­22 application of, 262 See also accession of China gains and losses, 221, 222t13.7 WTO Working Party on the Accession of China, 21, 22, by agricultural households, 275, 278t15.8 23, 24 explanation of, 273­78 income levels, 269, 270f15.3­272f15.8, 273 Zhu Rhongji, 74 by rural areas of provinces, 274, 274t15.4­275t15.5, 280t15.A.1 by urban areas of provinces, 274, 276t15.6­277t15.7, 281t15.A.2 D.C. STD #411 PAID PRSRT U.S.POSTAGE PERMIT WASHINGTON, . 20433 N.W D.C. Bank Street, orld W H ashington, The 1818 W U.S.A. by and areas, key analysis SUlatoT CODE Visa SOURCE DCWTO bank order. Bank, U.S. in a your on elopment on TEGIES dev ytQ$ atotbuSl ** $S $ H & S needed UlatoT of World with Mastercard the charges) drawn card) check only: order. STRA amount from the card only 800-645-7247 Drawing on your U.S. economic reforms ecirP. the or 2$ credit in Express all directly funds send the purchase books@worldbank.org / appears it for in at U.S. economy? # Please policy its continued REDUCTIONYT the kcotS* 76651D5 and at lautca,desu as( in us era .noita ordering include (required American When Bank. customers +1-703-661-1580 E-mail mrofni ruoy hi payable date no. Please www.worldbank.org/publications w my: World ensure no. check me. phone: looks to, hti ees siredro drac .hsac Check enclosed. The Bill By Online: Questions? development people w Charge Card Expiration Name Signature is make to: Institutional POVER its book this, property AND impact noitacilbupoc )X-7665-3128-0 dlro To NBSI( esaelP.seirtnuoc esahcrup segrahc,setatS dnes on detin tidercro kcehC.knaB ton Order: U tne W od choice eht myap eht of A e yb ot WTO O: elsewhere intellectual reference the .seireS .segap afI.redrorep edistuo USA WTO 005 mosrof . deinap esaelP.sdrac the WT REFORM, to and and tne %57 00.8$S desserdda U tiderc and The China THE services . mpoleve .4002 ot pu era yreviledlia metirep mocca 20172-0960, Code yluJ VA - accession D 00.6$S eb China tsu sredrorof m elbatpecca of Publications POLICY s from erty O d ylppa U ylno copy T na .sserP segrahc W e tuoba Zip/Postal Bank BANK my Herndon, AND pov gnildna H send 960, China' in eht darT stnuocsid .stnuocsid/gro.knabdlro mriaroF.degrahc DOHTE slaudividni detpecca World +1-703-661-1501 WORLD Publications scholars agriculture, will dna as knaB ytisrevin dna eblli M Box U w sulpredrorep a TNE morf sdrac please mail: fax: d print) By P.O. By CHINA ACCESSION, YES, G sredr How leading such reductions eltiT nihC lroW drofx cihpargoe O * w.snoitacilbup//:ptth gnippihS** gnippihs 00.7$S U MYAP O tiderC rotubirtsidlacol Please( Name Title Organization Address City State Country Phone Fax E-mail T R A D E A N D D E V E L O P M E N T S E R I E S " T his book fills a large gap in the analysis of China's WTO accession. Evidence-based investigations of the likely effect of further opening of China's economy, following the steady globalization that had occurred since 1979, have been largely unavailable. Moreover, the strategic dimensions of the policy changes implemented by China's far-seeing leaders have not been addressed adequately until now, and not as thoroughly as in this book. This collection of first-rate technical studies, drawn from good understanding of Chinese and international economic institutions, and tracing the likely effects of China's globalization on different segments of its economy -- especially its poor -- is essential reading for students of public policy." Long Yongtu, Secretary General, BOAO Forum for Asia "This book provides the most comprehensive and up-to-date evaluation of China's entry into the WTO. It provides a succinct summary of China's commitments to the WTO and a detailed analysis of the economic consequences of the fulfillment of these obligations. A particular strength of the volume is its treatment of the relationship between trade liberalization and the distribution of income in China." Nicholas R. Lardy, Senior Fellow, Institute for International Economics 3 TMxHSKIMBy356678zv,:&:<:':( THE WORLD BANK 0-8213-5667-4